EXPERIMENTAL AND STATISTICAL INVESTIGATION OF … · EXPERIMENTAL AND STATISTICAL INVESTIGATION OF...
Transcript of EXPERIMENTAL AND STATISTICAL INVESTIGATION OF … · EXPERIMENTAL AND STATISTICAL INVESTIGATION OF...
EXPERIMENTAL AND STATISTICAL
INVESTIGATION OF AUSTRALIAN NATIVE
PLANTS FOR SECOND-GENERATION
BIODIESEL PRODUCTION
Md Jahirul Islam
B.Sc., M.Sc.
Supervisors: Dr Wijitha Senadeera, A/Prof Richard Brown, Prof Zoran Ristovski, A/Prof Ian O’Hara
A thesis submitted in fulfilment of the requirements for the degree of
Doctor of Philosophy (PhD)
School of Chemistry, Physics and Mechanical Engineering
Faculty of Science and Engineering
Queensland University of Technology
2015
Keywords
Alternative energy, Artificial neural networks (ANN), Australian native plants, Biofuel,
Biodiesel, First-generation biodiesel, Multi criteria decision analysis (MCDA),
PROMETHEE-GAIA, Renewable energy, Second-Generation biodiesel, Fuel properties,
Principle component analysis (PCA), Response surface methodology (RSM), Analysis of
variance (ANOVA), Beauty leaf (Calophyllum inophyllum), Candle nut (Aleurites
Moluccana), Blue berry lily (Dianella Caerula), Queen palm (Syagrus Romanzoffiana),
Castor (Ricinus Communis), Bidwilli (Brachychiton Bidwilli), Karanja (Pongamia Pinnata),
Whitewood (Atalaya Hemiglauca), Cordyline (Cordyline Manners – Suttoniae), Flame tree
(Brachychiton Acerifolius), Chinese rain (Koelreuteria Formosana). Fatty acid methyl ester
(FAME), Chemical composition, Bio-oil extraction, physico-chemical properties, Fuel,
Transesterification , Pre-estrification, Nitrogen oxides (NOX), Diesel particulate matter
(DPM), Diesel engines, Particulate matter (PM), Performance, Exhaust emission
Abstract
The world is now facing a dual crisis of the rapid depletion of fossil fuels coupled with
environmental degradation. At present, the worlds’ energy supply largely depends on
petroleum based fuels, which have a finite supply and a restricted geographical availability.
Furthermore, the demand for energy is increasing rapidly due to the growing world
population and industrialisation. Therefore, development of a sustainable, long term
alternative fuel source has become critical. As a consequence, biodiesel made from various
crops, as well as animal fat, is receiving much attention of late and is emerging as an
alternative to conventional petroleum based fuels. The socio-economic advantages of using
biodiesel are many, including renewability, bio-degradability and low-toxicity compared with
petroleum fuels. The biodiesels which are commercially available today are mainly obtained
from edible vegetable oil feedstocks and are referred as first-generation biodiesels. Although
offering many advantages over conventional petroleum based fuels, the use of these types of
biodiesels are receiving serious condemnation due to their pressure on food sources and it is
this debate on "food versus fuel" which brings into question the sustainability of first-
generation biodiesels. Biodiesels produced from non-edible vegetable oils are considered a
possible alternative to overcome the socio-economic disadvantages of current biodiesel
technology. These types of biodiesel are called second-generation biodiesel, however they are
not commercially available today, mainly due to concerns in relation to secure feedstock
supply and a lack of technological assessment.
This study aimed to explore the potential of Australian native plants as a source of second-
generation biodiesel. The aim was achieved through several experimental measurements and
the numerical investigation of first-generation and second-generation biodiesel. While
conducting the research, necessary data were obtained from two sources: through
experimental investigation and data collected from literature published in peer reviewed
journals and conferences. A multivariate data analysis was conducted using principal
component analysis (PCA) to establish a correlation between important properties and the
chemical composition of biodiesel. The fuel properties investigated in this study were
kinematic viscosity, density, higher heating value, oxidation stability, cold filter plugging
point temperature, flash point temperature and iodine value. Using the data obtained, together
with results of the correlation study, a set of artificial neural network (ANN) models were
developed to estimate the above mentioned fuel properties of biodiesel. MatLab R2012a
software was used to train, validate and simulate the ANN model on a personal computer.
The network architecture was optimised using a trial and error methodology, in order to
obtain the best performance of the model. The ANN models were development in such a way
that they would be able to estimate the fuel properties for biodiesel obtained from Australian
native plants according to their respective chemical composition. For this purpose, biodiesels
were produced on a laboratory scale from eleven non-edible oil seed plants, as follows:
Beauty leaf (Calophyllum inophyllum), Candle nut (Aleurites Moluccana), Blue berry lily
(Dianella Caerula), Queen palm (Syagrus Romanzoffiana), Castor (Ricinus Communis),
Bidwilli (Brachychiton Bidwilli), Karanja (Pongamia Pinnata), Whitewood (Atalaya
Hemiglauca), Cordyline (Cordyline Manners – Suttoniae), Flame tree (Brachychiton
Acerifolius), Chinese rain (Koelreuteria Formosana). These plants are Australian natives and
naturally grow in Queensland. During the production of biodiesels from these native oil
seeds, the level of difficulty of seed processing, oil content in the seed kernels and free fatty
acid (FFA) content were also measured and reported. The chemical composition of the
obtained biodiesel samples were also measured in terms of fatty acid methyl ester (FAME)
profiles. The fuel properties of biodiesel were estimated by using chemical composition as the
input variable to developed ANN models. Based on physico-chemical properties, the native
species feedstock were then evaluated, and their quality was compared with each other using
a multi-criteria decision method (MCDM) software PORMETHEE-GAIA. In addition,
sensitivity analysis of native plant ranking was investigated by changing the weighting of
three important criteria: oil yield; oxidation stability and cold filter plugging point
temperature. This study found that among the native plant feedstock investigated, Beauty leaf
was the top ranked candidate for second-generation biodiesel production followed by Queen
palm, Castor and Karanja. Furthermore, Beauty leaf and Queen palm biodiesels were found
to be the better choice for tropical/sub-tropical regions, however the opposite was true for
cold weather conditions, where Castor, Cordyline and Flame trees were the best candidates
for biodiesel production.
In the next phase of the study, biodiesel from Beauty leaf oil seeds was produced on a pilot
scale and experimental evaluation of the fuel properties, emission and performance were
tested in a multi-cylinder diesel engine. While extracting bio-oil from the Beauty leaf oil
seeds, the effect of seed preparation, moisture content and oil extraction methods on bio-oil
yield were experimentally investigated. The oil extraction methods studied in this work were:
(1) mechanical oil extraction using a screw press, (2) static chemical oil extraction under
atmospheric conditions and (3) accelerated chemical oil extraction using high pressure and
temperature. In both chemical extraction methods, n-Hexane was used as the oil solvent. The
physico-chemical properties of Beauty leaf oil and biodiesel were determined experimentally
and compared with that of commercially available biodiesel. Due to the high concentration of
free fatty acid contained in Beauty leaf, a two-step biodiesel conversion method consisting of
acid catalysed pre-esterification and alkali catalysed transesterification was used. The
performance of biodiesel conversion and esterification reaction parameters were investigated
using response surface methodology (RSM) based on a Box-Behnken experimental design.
This study found that seed preparation and moisture content in the oil bearing seed kernels
had a significant impact on oil yields, especially in terms of the mechanical oil extraction
method. High temperature and pressure during the extraction process was also found to
increase the oil extraction performance. A clear difference was found in the physical
properties of Beauty leaf oils obtained using different oil extraction methods, which
ultimately affected the oil to biodiesel conversion process. However, Beauty leaf oil methyl
esters (biodiesel) were very consistent in terms of their physico-chemical properties and were
able to compete with commercially available biodiesel in terms of fuel quality. The reaction
conditions for the largest reduction in FFA concentration for acid catalysed pre-esterification
was 30:1 methanol to oil molar ratio, 10% (w/w) sulphuric acid catalyst loading and 75 °C
reaction temperature. In the alkali catalysed transesterification process a 7.5:1 methanol to oil
molar ratio, 1% (w/w) sodium methoxide catalyst loading and 55 °C reaction temperature
were found to result in the highest FAME conversion. The performance and emissions of a
four-cylinder common rail diesel engine were experimentally investigated using neat diesel
and biodiesel produced from Beauty leaf oil. Results indicated that 5% and 10% blends of
Beauty leaf oil biodiesel with diesel fuel can be used in conventional diesel engines without
engine modification. Beauty leaf biodiesel reduced the engine power, brake thermal
efficiency, cylinder peak pressure and specific nitrogen oxide (NOx) particle mass (PM)
emissions. At the same time, brake specific fuel consumption and particle number emissions
were found to be higher from Beauty leaf biodiesel compared with that of conventional diesel.
However, this variation is not unusual and is commonly found in conventional biodiesels,
mainly due to variations in physico-chemical properties between biodiesel and conventional
diesel.
This thesis advances knowledge in the field biofuel technology, by delivering an extensive
database of the properties of second-generation biodiesel and its application in a modern
diesel engine. The research methodology and numerical model developed in this study can be
used for a broad range of biodiesel feedstock and will facilitate further biodiesel research in
the future. The experimental study on modern automobile engines using Beauty leaf biodiesel
indicated the suitability of Australian native plants for use as fuel for modern automobile
diesel engines without engine modification. Therefore, the findings of this study are expected
to serve as the basis for further developments in the use of Beauty leaf as a feedstock for
industrial scale biodiesel production.
i
Table of Contents
Table of Contents ..................................................................................................................................... i
List of Figures ......................................................................................................................................... v
List of Tables ...................................................................................................................................... viii
List of Abbreviations ............................................................................................................................... x
List of Publication ................................................................................................................................ xii
Published journal papers: ..................................................................................................................... xii
Conference papers: ............................................................................................................................... xii
Statement of Original Authorship ........................................................................................................ xiv
Acknowledgements ............................................................................................................................... xv
CHAPTER 1: INTRODUCTION ....................................................................................................... 1
1.1 Background .................................................................................................................................. 1
1.2 Research questions ....................................................................................................................... 4
1.3 Aims and objectives ..................................................................................................................... 4 1.3.1 Aims ................................................................................................................................. 4 1.3.2 Specific study objectives .................................................................................................. 4
1.4 Significantce of this project ......................................................................................................... 5
1.5 Thesis outline ............................................................................................................................... 6
CHAPTER 2: LITERATURE REVIEW ......................................................................................... 13
1 Introduction ................................................................................................................................ 15
2.1 Biodiesel .................................................................................................................................... 19 2.1.1 Biodiesel feedstock ......................................................................................................... 20 2.1.2 First and second-generation biodiesel ............................................................................. 21 2.1.3 Potential second-generation biodiesel feedstock ............................................................ 23 2.1.4 Production of biodiesel ................................................................................................... 26 2.1.5 Chemical composition of biodiesel ................................................................................. 29 2.1.6 Biodiesel standards ......................................................................................................... 31 2.1.7 Fuel properties ................................................................................................................ 32 2.1.7.1 Kinematic viscosity ........................................................................................................ 33 2.1.7.2 Density ............................................................................................................................ 35 2.1.7.3 Cetane number (CN) ....................................................................................................... 36 2.1.7.4 Heating (calorific) Value ................................................................................................ 37 2.1.7.5 Flash point ...................................................................................................................... 37 2.1.7.6 Oxidation stability .......................................................................................................... 38 2.1.7.7 Cold temperature properties ........................................................................................... 38 2.1.7.8 Lubricity ......................................................................................................................... 40 2.1.7.9 Iodine value .................................................................................................................... 41
2.2 Biodiesel as a diesel Engine Fuel ............................................................................................... 41 2.2.1 Engine performance ........................................................................................................ 42 2.2.2 Exhaust emissions ........................................................................................................... 44
2.3 Artificial neural networks .......................................................................................................... 47 2.3.1 ANN in predicting engine emission and performance .................................................... 51 2.3.2 ANN in predicting fuel properties .................................................................................. 53
2.4 ANN modeling of second-generation biodiesel ......................................................................... 57
ii
2.5 Conclusions ............................................................................................................................... 59
CHAPTER 3: ARTIFICIAL NEURAL NETWORK (ANN) MODEL DEVELOPMENT ......... 61
3.1 Introduction ............................................................................................................................... 62
3.2 Data Collection .......................................................................................................................... 65
3.3 Results and discussion ............................................................................................................... 71 3.3.1 Chemical composition .................................................................................................... 71 3.3.2 Fuel properties ................................................................................................................ 74 3.3.3 Correlation of chemical composition and fuel properties ............................................... 78 3.3.4 Principle component analysis ......................................................................................... 82 3.3.5 ANN model development ............................................................................................... 84 3.3.6 Evaluation of ANN model performance ......................................................................... 88
3.4 Conclusion ................................................................................................................................. 91
CHAPTER 4: BIODIESEL FROM AUSTRALIAN NATIVE PLANTS ...................................... 93
4.1 Introduction ............................................................................................................................... 93
4.2 Potential native oil seed plants ................................................................................................... 94 4.2.1 Beauty leaf (Calophyllum inophyllum) ........................................................................... 94 4.2.2 Candle nut (Aleurites Moluccana) .................................................................................. 95 4.2.3 Blue berry lily (Dianella Caerula) ................................................................................. 96 4.2.4 Queen palm (Syagrus Romanzoffiana) ........................................................................... 97 4.2.5 Castor (Ricinus Communis) ............................................................................................ 98 4.2.6 Bidwilli (Brachychiton Bidwilli) .................................................................................... 99 4.2.7 Karanja (Pongamia Pinnata) ........................................................................................ 100 4.2.8 Whitewood (Atalaya Hemiglauca) ............................................................................... 101 4.2.9 Cordyline (Cordyline Manners – Suttoniae) ................................................................ 102 4.2.10 Flame tree (Brachychiton Acerifolius) .......................................................................... 102 4.2.11 Chinese rain (Koelreuteria Formosana) ....................................................................... 103
4.3 Seed Collection And Preparation ............................................................................................. 104 4.3.1 Kernel extraction .......................................................................................................... 104 4.3.2 Kernel grinding ............................................................................................................. 105 4.3.3 Kernel drying ................................................................................................................ 105
4.4 Oil extraction ........................................................................................................................... 106
4.5 Chemical composition ............................................................................................................. 109
4.6 fuel properties .......................................................................................................................... 113
4.7 Evaluation of native plant methyle ester .................................................................................. 116
4.8 Conclusion ............................................................................................................................... 120
CHAPTER 5: PILOT SCALE BEAUTY LEAF (CALOPHYLLUM INOPHYLLUM) BIODIESEL PRODUCTION .......................................................................................................... 123
5.1 Introduction ............................................................................................................................. 125
5.2 Seed preparation ...................................................................................................................... 127 5.2.1 Seeds collection ............................................................................................................ 128 5.2.2 Kernel extraction .......................................................................................................... 128 5.2.3 Kernel drying ................................................................................................................ 129 5.2.4 Kernel grinding ............................................................................................................. 129
5.3 Oil extraction ........................................................................................................................... 129 5.3.1 Mechanical oil extraction using oil press (OP) ............................................................. 130 5.3.2 Chemical oil extraction using n-Hexane (nHX) ........................................................... 131 5.3.3 Accelerated solvent extraction (ASE) .......................................................................... 131
5.4 Oil Yield .................................................................................................................................. 132
5.5 Comparison of oil extraction methods ..................................................................................... 133
iii
5.6 Oil Analysis ............................................................................................................................. 135
5.7 Biodiesel production ................................................................................................................ 136
5.8 Biodiesel analysis .................................................................................................................... 140
5.9 Fuel properties ......................................................................................................................... 143 5.9.1 Kinematic viscosity ...................................................................................................... 144 5.9.2 Density .......................................................................................................................... 144 5.9.3 Higher heating value ..................................................................................................... 145 5.9.4 Acid number ................................................................................................................. 145 5.9.5 Oxidation stability ........................................................................................................ 146 5.9.6 Iodine value .................................................................................................................. 146 5.9.7 Cetane Number ............................................................................................................. 147 5.9.8 Flash point temperature ................................................................................................ 147 5.9.9 Cold filter plug point (CFPP)........................................................................................ 148
5.10 Validation of beauty leaf biodiesel .......................................................................................... 150
5.11 Conclusion ............................................................................................................................... 153
CHAPTER 6: PRODUCTION PROCESS OPTIMISATION OF BIODIESEL ........................ 157
6.1 Introduction .............................................................................................................................. 159
6.2 Materials and method ............................................................................................................... 163 6.2.1 Beauty leaf oil extraction method ................................................................................. 163 6.2.2 Analysis methods .......................................................................................................... 163 6.2.3 Pre-esterification and transesterification methods ........................................................ 164
6.3 Results and discussion ............................................................................................................. 167 6.3.1 Beauty leaf oil characterisations ................................................................................... 167 6.3.2 Acid-catalysed pre-esterification .................................................................................. 168 6.3.3 Base-catalysed transesterification of pre-esterified Beauty leaf oil .............................. 172
6.4 Conclusions .............................................................................................................................. 176
CHAPTER 7: DIESEL ENGINE TESTING WITH BIODIESEL OF CONTROLLED CHEMICAL COMPOSITION ........................................................................................................ 177
7.1 Introduction: ............................................................................................................................. 179
7.2 Materials and methods ............................................................................................................. 181 7.2.1 Engine and fuel specification ........................................................................................ 181 7.2.2 Exhaust sampling and measurement system ................................................................. 183
7.3 Results and Discussion ............................................................................................................ 186 7.3.1 Specific PM emissions .................................................................................................. 186 7.3.2 Specific PN emissions .................................................................................................. 187 7.3.3 Particle number size distribution .................................................................................. 189 7.3.4 Particle median size ...................................................................................................... 189 7.3.5 NOx emissions .............................................................................................................. 190 7.3.6 Influence of fuel physical properties and chemical composition on particle
emissions ...................................................................................................................... 192 7.3.7 Comparison of engine performance and particle emissions among used
biodiesels ...................................................................................................................... 194
7.4 Conclusions .............................................................................................................................. 195
CHAPTER 8: AUTOMOBILE DIESEL ENGINE TESTING WITH BEAUTY LEAF BIODIESEL 205
8.1 Introduction .............................................................................................................................. 205
8.2 Instrumentation and methodology ........................................................................................... 206
8.3 ResultS and discussion ............................................................................................................. 208 8.3.1 Engine power ................................................................................................................ 208 8.3.2 BTE and BSFC ............................................................................................................. 209
iv
8.3.3 Cylinder pressure .......................................................................................................... 210 8.3.4 Nitrogen oxide (NOx) emission .................................................................................... 212 8.3.5 Particle mass (PM) and particle number (PN) .............................................................. 213
8.4 Conclusion ............................................................................................................................... 214
CHAPTER 9: CONCLUSIONS ...................................................................................................... 217
9.1 ConclusionS arising from this thesis ........................................................................................ 217
9.2 Limitations and Recommendations for future work ................................................................ 223
6 BIBLIOGRAPHY .......................................................................................................................... 227
APPENDICES ................................................................................................................................... 266 APPENDIX A: MatLab code for ANN models training......................................................... 266 APPENDIX B: The eigenvalue for each of the PCs ............................................................. 270
v
List of Figures
Figure 1-1: Outline on the thesis ................................................................................ 11
Figure 2-1. Biodiesel feedstocks around the world .................................................... 21
Figure 2-2. Transeterification reaction ....................................................................... 28
Figure 2-3: Soap formation ........................................................................................ 29
Figure 2-4: Acid pre-esterification ............................................................................. 29
Figure 2-5: Fatty acid profile of various biodiesel fuels ............................................ 31
Figure 2-6: Variation in fuel properties of various biodiesel ..................................... 33
Figure 2-7: Schematic diagram of a typical diesel engine fuel system ...................... 42
Figure 2-8: Biological neuron .................................................................................... 48
Figure 2-9: Multi-layer ANN model .......................................................................... 49
Figure 2-10: Working principle of ANN ................................................................... 49
Figure 2-11: Comparison of the performance of between ANN and various linear and non-linear prediction techniques ................................................ 50
Figure 2-12: Proposed structure of ANN model ........................................................ 58
Figure 3-1: Number and average weight in percentages of fatty acid methyl esters found in the samples .......................................................................... 74
Figure 3-2: Correlation of (a) C18:2 with oxidation stability; (b) H2 with CN ......... 77
Figure 3-3: Correlation of (a) ANDB with CN; (b) ANBD with IV ......................... 79
Figure 3-4: Effect of ACL on biodiesel (a) kinematic viscosity (KV) and (b) higher heating value (HHV) ......................................................................... 81
Figure 3-5: Principle component analysis and correlation of biodiesel properties with chemical composition: ........................................................ 84
Figure 3-6: Proposed flow chart of ANN prediction model development ................. 85
Figure 3-7: Structure of ANN .................................................................................... 86
Figure 3-8: Analysis of influence between chemical composition and fuel properties ...................................................................................................... 87
Figure 3-9: Biodiesel properties estimation accuracy of developed ANN models .......................................................................................................... 91
Figure 4-1: Beauty leaf tree growing along a beach front, and in a park and its distribution in Australia ............................................................................... 95
Figure 4-2: The tree and kernels of Candle nut ......................................................... 96
Figure 4-3: Blue berry lily plant and seeds. ............................................................... 97
Figure 4-4: Tree and kernels of Queen palm ............................................................. 98
Figure 4-5: Shrub and seeds of Castor ....................................................................... 99
vi
Figure 4-6: Bidwilli plant and seeds ......................................................................... 100
Figure 4-7: Karanja fruit and seeds ......................................................................... 101
Figure 4-8: Tree and fruit of Whitewood .................................................................. 101
Figure 4-9: The tree and fruit of Cordyline .............................................................. 102
Figure 4-10: Flame tree, fruit and seeds .................................................................. 103
Figure 4-11: Chinese rain tree and fruits ................................................................. 104
Figure 4-12: Kernel extraction ................................................................................. 105
Figure 4-13: Ground kernels .................................................................................... 106
Figure 4-14: ASE 350 cell loading process .............................................................. 107
Figure 4-15: n-hexan removing using DionexTM SETM 400 ..................................... 108
Figure 4-16: Oil yield of native plant seed kernels .................................................. 109
Figure 4-17: Extracted bio-oil sample from native plants ........................................ 109
Figure 4-18: (a) Graphical Analysis for Interactive Assistance (GAIA) plot for eight biodiesel showing 11 criteria and decision vector. (b) Corresponding complete ranking and Phi value of biodiesel from native plants ............................................................................................... 118
Figure 5-5-1: Flow chart of Beauty leaf seeds preparation ...................................... 128
Figure 5-5-2: Mechanical oil extraction through a screw press ............................... 130
Figure 5-5-3: Chemical oil extraction ...................................................................... 131
Figure 5-5-4: ASE oil extraction (a) Dionex™ ASE 350® (b) solvent removal with flow of nitrogen .................................................................................. 132
Figure 5-5-5: Beauty leaf oil yield from three different extraction methods. .......... 133
Figure 5-5-6: Soap formation in oils contains high FFA ......................................... 137
Figure 5-5-7: Acid pre-esterification ........................................................................ 137
Figure 5-5-8: Two step bio-diesel production process from Beauty leaf oil ............ 139
Figure 5-5-9: Beauty leaf oil esterification .............................................................. 139
Figure 5-5-10: (a) Graphical Analysis for Interactive Assistance (GAIA) plot for eight biodiesel showing 10 criteria and decision vector. (b) Corresponding ranking of biodiesel on their outranking flow. .................. 151
Figure 6-6-1: (a) Esterification and transesterification reactor; (b) Layer of Methanol-Water (top) and oil (bottom) after acid-catalysed pre-esterification; (c) Layer of Beauty leaf oil methyl ester (top) and glycerol (bottom) after base-catalysed Trans-esterification. ...................... 165
Figure 6-6-2: Scatter diagram of experimental FFA (%) and predicted FFA (%) of a linear model. ....................................................................................... 170
Figure 6-6-3: Response surface of FFA content against .......................................... 171
Figure 6-6-4: Scatter diagram of experimental and calculated FAME (%) of full quadratic model. .................................................................................. 174
vii
Figure 6-6-5: Response surface ester content against catalyst concentration vs. (a) methanol to oil molar ratio at 55 °C; (b) reaction temperature at 7.5:1 methanol to oil molar ratio. ............................................................... 175
Figure 7-7-1: Schematic diagram of used engine exhaust measurement system ..... 185
Figure 7-7-2: Brake specific PM emission at ........................................................... 187
Figure 7-7-3: Brake specific PN emissions at 1500 rpm 100% load (a) and 2000 rpm 100% load (b). ........................................................................... 188
Figure 7-7-4: Variations in particle median size among used fuels at 1500 rpm 100% load (a) and 2000 rpm 100% load (b), while (c) shows particle median size variation with total number concentration. ............................ 190
Figure 7-7-5: Brake specific NOx emission at (a) 1500 rpm 100% load and (b) 2000 rpm 100% load. ................................................................................. 192
Figure 7-7-6: Variation in specific PM and PN emissions with used fuel surface tension, viscosity and oxygen content ........................................... 193
Figure 7-7-7: Comparison of engine performance (power, BSFC) and particle emissions (PM, PN) among biodiesels and their blends where petroleum diesel was used as a reference fuel. .......................................... 195
Figure 8-8-1: Experimental setup............................................................................. 207
Figure 8-8-2: Variation of power output for neat diesel and biodiesel blends (a) brake power; (b) indicated power .............................................................. 209
Figure 8-8-3: (a) Brake thermal efficiency (BTE) and (b) Brake specific fuel consumption (BSFC) for neat diesel and biodiesel blends ........................ 210
Figure 8-8-4: Engine cylinder pressure for diesel and Beauty leaf biodiesel blends, (a) full load; (b) 75% load; (c) 50% load; and (d) 25% load ......... 211
Figure 8-8-5: NOX emission for diesel and BOME blend for different engine load conditions ........................................................................................... 213
Figure 8-8-6: Particle emission for diesel and BOME blend in different engine load condition (a) Brake specific particle mass (PM); (b) brake specific particle number ............................................................................. 214
viii
List of Tables
Table 2-1. Advances in biodiesel technology ............................................................ 22
Table 2-2: Second-generation biodiesel feedstock containing oil by dry weight ...... 26
Table 2-3: Reported optimum conditions for transesterification of oils for biodiesel production. .................................................................................... 28
Table 2-4: Chemical structure of common fatty acid in biodiesels ............................ 30
Table 2-5: International biodiesel standards .............................................................. 32
Table 2-6: Performance and emission of diesel engines with biodiesel .................... 43
Table 2-7: ANN used in automobile engine application ............................................ 55
Table 2-8: ANN in predicting fuel properties ............................................................ 56
Table 3-1 Biodiesel property test standard ................................................................. 66
Table 3-2: Biodiesel datasets investigated in this study ............................................. 68
Table 3-3: Structural formulae for fatty acids methyl ester found in biodiesel samples ......................................................................................................... 71
Table 3-4: Chemical composition of tested biodiesel ................................................ 72
Table 3-5: BREF experimental results of biodiesel properties .................................. 73
Table 3-6: Summary of the secondary data for biodiesel properties .......................... 76
Table 3-7: Number of input variables and optimised number of neuron in ANN model ............................................................................................................ 88
Table 4-1: Fatty acid profile and chemical composition of bio-oil produced from native plants ....................................................................................... 112
Table 4-2: Estimated biodiesel properties ................................................................ 115
Table 4-3: Variables and preference used in PROMETHEE-GAIA analysis .......... 117
Table 4-4: Comparative rank shift with different weighting for bio-oil yield ......... 119
Table 4-5: Comparative rank shift with different oxidation stability of biodiesel ... 119
Table 4-6: Comparative Rank shift with different cold filter plugging point temperature ................................................................................................. 120
Table 5-1: Advantages and disadvantages of the three extraction methods ............ 134
Table 5-2: Physical properties of Beauty leaf oil ..................................................... 136
Table 5-3: The fatty acid distributions of Beauty leaf and commercial biodiesels .................................................................................................... 142
Table 5-4: Fuel properties of Beauty leaf oil biodiesel and commercial biodiesels .................................................................................................... 149
Table 5-5: Variables and preference used in PROMETHEE-GAIA analysis .......... 150
Table 5-6: Comparative rank shift with different OS and CFPP weighting ............ 152
ix
Table 6-1: Experimental range and levels of independent variables ....................... 165
Table 6-2: Coded experimental design .................................................................... 165
Table 6-3: Fatty acid composition of Beauty leaf oil ............................................... 167
Table 6-4: Properties of Beauty leaf oil. .................................................................. 167
Table 6-5: Experimental conditions and results for acid-catalysed pre-esterification ............................................................................................... 169
Table 6-6: Regression coefficients for %FFA prediction ........................................ 170
Table 6-7: Experimental data for base-catalysed trans-esterification. ..................... 172
Table 6-8: Regression coefficients for FAME (%) prediction ................................. 173
Table 7-1: Test engine specification ........................................................................ 181
Table 7-2: Fatty acid profile of used biodiesels ....................................................... 183
Table 7-3: Important physicochemical properties of tested fuels ............................ 185
Table 8-1: Properties of Beauty leaf fatty acid methyl ester (BOME) and petroleum diesel ......................................................................................... 206
Table 8-2: Test engine specification ........................................................................ 208
x
List of Abbreviations
AFR air-fuel ratio ANN artificial neural network ANOVA analysis of variance BMEP brake mean effective pressure BSFC brake-specific fuel consumption BTE brake thermal efficiency C cylinder CB cylinder bore CFPP cold-filter plugging point CI compression ignition CME coconut methyl ester CN cetane number CNG compressed natural gas CO carbon monoxide CO2 carbon dioxide CP cloud point DU degree of unsaturation ECP Engine cylinder pressure CPO crude palm oil CR compression ratio EGR exhaust gas recirculation ES engine stock ET engine temperature FAME fatty acid methyl ester FAEE fatty acid ethyle ester FDR fuels blend ratio FFA free fatty acid FFR fuel flow rate GHG greenhouse gas H2 hydrogen HC hydrocarbon HV heating value HHV higher heating value ICE internal combustion engines IP injection pressure IT injection timing
IV iodine value KV kinematic viscosity L load Lb lubricity LHV lower heating value LPG liquid petroleum gas MLR multiple linear regression MRE mean relative error MSE mean square error N2 nitrogen NA naturally aspirated NOx nitrogen oxides O2 oxygen OS oxidation stability P power PCA principle component analysis PCR principle component regression PLS partial least square regression PM particulate matter PME palm oil methyl ester PP pour point R2 regression coefficient RME rapeseed methyl ester RPM rotation per minute S sulphur SI spark ignition SME soybean methyl ester SOx sulphur oxides T torque TC turbocharged Texh exhaust gas temperature TP throttle position UHC unburned hydrocarbons VT valve timing WCO waste cooking oil
xii
List of Publications
Published journal papers:
1. M. I. Jahirul, K. Wenyong, L Moghaddam, R. J. Brown, I. O'Hara, W. Senadeera, N. Ashwath. Biodiesel production from non-edible Beauty Leaf (Calophyllum inophyllum) oil: process optimization using response surface methodology (RSM), Energies 7(8), 5317-5331, 2014. I.F. 1.884.
2. M. M. Rahman, A. M. Pourkhesalian, M. I. Jahirul, S. Stevanovic, P. X. Pham, H. Wang, A.R. Masri, R. J. Brown and Z. D. Ristovski. Particle emissions from biodiesels with different physicochemical properties, Fuel 134, 201-208, 2014. IF. 3.357
3. M. I. Jahirul, R. J. Brown, W. Senadeera, Z Ristovski, I O'Hara. Artificial neural network approach in identifying sustainable future generation biofuel feedstock, Energies, Special issue: Alternative Fuels for the Internal Combustion Engines (ICE), 6, 3764-3806, 2013. I.F. 1.884.
4. M. I. Jahirul, J. R. Brown, W. Senadeera, N. Ashwath, C. Laing, J. Leski-Taylor, and M. G. Rasul. Optimisation of Bio-Oil Extraction Process from Beauty Leaf (Calophyllum Inophyllum) Oil Seed as a Second-Generation Biodiesel Source, Procedia Engineering, 56, 619-24, 2013.
Submitted Paper: 1. M. I. Jahirul, W. Senadeera, J. R. Brown, , N. Ashwath, M. G. Rasul, M. M.
Rahman, Muhammad Aminul Islam, and I. M. O’Hara. Physico-chemical Assessment of Beauty Leaf (Calophyllum Inophyllum) as Second-Generation Biodiesel Feedstock. Submitted to the journal of Energy Conversion and Management.
Conference papers:
1. M. I. Jahirul, W. Senadeera, R. J. Brown, L. Moghaddam. Estimation of Biodiesel Properties from Its Chemical Composition – An Artificial Neural Network (ANN) Approach. International Conference on Environment and Renewable Energy, Cité Internationale Universitaire de Paris, 17 Boulevard Jourdan, 75014 Paris – France, 7-8 May 2014.
2. M. I. Jahirul, W. Senadeera, P. Brooks, R.J. Browna, R. Situ, P.X. Pham and A.R. Masri. An Artificial Neutral Network (ANN) Model for Predicting Biodiesel Kinematic Viscosity as a Function of Temperature and Chemical Compositions. 20th International Congress on Modelling and Simulation, Adelaide, Australia, 1–6 December 2013, Paper no. 1021
xiii
3. Jahirul M. I., Brown R. J, Senadeera W, Z Ristovski. Influence of the physical properties of fractionated methyl ester on the ultra-fine particle emission of internal combustion engine. 8th Australia and New Zealand Aerosol Workshop, 26-27 November 2012. Canberra, Australia.
xiv
Statement of Original Authorship
The work contained in this thesis has not been previously submitted to meet
requirements for an award at this or any other higher education institution. To the
best of my knowledge and belief, the thesis contains no material previously
published or written by another person except where due reference is made.
Signature: QUT Verified Signature
Date: 17/04/2014
xv
Acknowledgements
First of all, I would like to express my praise to Almighty Allah Subhanawatala for
giving me the opportunity to finish the thesis successfully. I would like to express my
gratitude and profound respect to my supervisors Dr Wijitha Senadeera, A/Prof
Richard Brown, Prof Zoran Ristovski and A/Prof Ian O’Hara. Their supervision,
guidance and comments helped me finish the research work. I am grateful for their
generosity and amicable nature. I found them to be not only great supervisors and
scientists, but also caring and supportive men. Further, I should also express my
particular gratitude to A/Prof Richard Brown for his excellent support,
encouragement, guidance and care during my research journey.
My sincere appreciation goes to Queensland University of Technology for providing
me scholarships and access to various research instruments/facilities. Further, I
should thank the Science and Engineering Faculty and the School of Chemistry,
Physics and Mechanical Engineering at QUT for their continued administrative
support and financial assistance, particularly for conference and workshop
attendance. I would like to express thanks to the Biofuel Engine Research Facility
(BERF), QUT, for providing engine testing facilities. I would also like to extend my
thanks to all researchers and staff of this facility. Special thanks go to Mr Noel
Hartnett, Mr Scott Abbett, Mr. Tony Morris, Mr. Nathaniel Raup, Mr Shane Russel
and other technical staff of QUT, who also contributed to this work. I would also like
to extend my gratitude to Dr. Lalehvash Moghaddam, Centre for Tropical Crops and
Biocommodities (CTCB), QUT, for her significant contribution in this project.
I would like to thank my friends and colleagues at QUT for their continuous support
and encouragement, without which my PhD journey might be more difficult. Special
thanks to Chaminda Prasad Karunasena Helambage, M. Aminul Islam, Meisam
Babaie, Md Mostafizur Rahman, Prof. Nurun Nabi, Dr. Timothy Bodisco, Kabir
Suara, Farhad Hossain and many others, whom I always found beside me during this
project. I wish to commend the efforts of all undergraduate and postgraduate students
involved in this project for their assistance, especially, Wenyong Koh, Jakub L.
xvi
Taylor, Cameron Laing, Eerond Parez, Luke Asgill, Fadil Darmanto, Jagaddhita
Pandya Atmaja, Bodhimula Satyajati and Mohammad Hariz Bin Mokhtar.
I would like to thank A/Prof Nanajapa Ashwath and A/Prof Muhammad Rasul, as
well as all technical staff from the Centre for Plant and Water Sciences (CPWS),
CQU, for their sincere support during my Rockhampton bio-oil extraction visit. I
would like to acknowledge my gratitude to Professor Henning Bockhorn and
Michael Stroebele from the Karlsruhe Institute of Technology (KIT), Germany, for
their immeasurable support and guidance during my visit to Germany for biodiesel
property testing. Further, I am very thankful to A/Prof Bo Feng and George Thomas
from the University of Queensland (UQ), for their excellent support during the
engine testing campaign with Australian native plants. I would also like to express
my thank to Dr. Peter Brooks, University of the Sunshine Coast (USC) for extending
his support to my project and helping to determine the chemical composition of
biodiesel using his lab facilities.
Further, I should be very thankful to my wife Mst. Halima Akhter for her unique
love, passion, encouragement and commitment. She has been so close to me all the
time, along with our little daughter Arisha Islam, who has tried her best to cheer her
father with lovely smiles or cheerful activities. I also would like to thank my
relatives, friends and their families who helped me throughout the journey in many
ways.
Last but not least, I would like to acknowledge with gratitude, the support and love
of my parents, Abdul Mannan Mozumder and Mst. Nojumer Nesa. Although their
life is full of up-and-down situations and financial hardship, they always try to keep
me on the right track and consistently encourage me for moving forward with my
study. I strongly believe that this thesis would not be possible without their sincere
effort and wishes.
Chapter 1: Introduction 1
Chapter 1: Introduction
1.1 BACKGROUND
The current global energy supply is based on petroleum fuels (oil, natural gas, coal)
of which the reserves are finite. Given the growing world population, the increasing
energy consumption per capita and global warming due to greenhouse gas emissions,
the necessity of identifying long-term alternative energy sources is well recognised.
In order to counter greenhouse gas emissions, the European Union ratified the Kyoto
Protocol in 2002 and emphasised the potential for scientific innovation, which
unfortunately has not yet been fully achieved. Atmospheric CO2 concentration has
already exceeded the allowable level 10 years earlier than had previously been
predicted (Stern 2008).
Although the transport sector occupies third place (after industry and the building
sector) when considering total global energy consumption and greenhouse gas
emissions, it is the fastest growing sector. By 2030, the energy consumption and CO2
emissions of this sector are predicted to be 80% above the levels seen today (Miller,
Schmidt and Shindell 2006). Besides, it is also the sector that most heavily depends
on petroleum fuel (through the oil-derived liquid products gasoline and diesel) and
currently consumes 30% of global petroleum oil, which is predicted to increase to
60% by 2030 (Luque et al. 2008). Furthermore, the availability of petroleum oil is
geographically restricted and the era of cheap and secure oil is almost over. These
facts have forced automobile researchers to look for alternative carbon neutral
transport fuels which promise an harmonious amalgamation of sustainable
development, energy conversion, energy efficiency and environmental preservation
(Jahirul et al. 2010). As yet, no such option for fuel has been fully developed for the
transportation sector. Moreover, cars which emit no greenhouse gases (electric, solar,
hydrogen etc.) are a long way from becoming mainstream vehicles. Therefore, the
development of a sustainable, long-term alternative fuel has become essential, with
2 Chapter 1: Introduction
biodiesel receiving much attention and presenting as a promising alternative to
conventional fossil fuel (Luque et al. 2008).
In recent years, biodiesels have received much attention as a sustainable alternative
for petroleum diesel. It is liquid fuel made from various oil seeds crops, as well as
animal fat. The socio-economic advantages of using biodiesel are many, including
the fact that it is renewable, bio-degradable, non-toxic and eco-friendly compared
with petroleum fuels. Biodiesels are now produced on an industrial scale around the
world, using edible oil feedstocks such as soybean oil, palm oil, sunflower oil, corn
oil, olive oil, mustard oil etc (Bannikov 2011; Benjumea, Agudelo and Agudelo
2008; Jahirul, Brown, Senadeera, O'Hara, et al. 2013). These are called first-
generation biodiesels. In general, biodiesel derived from these sources can be defined
as mono-alkyl esters of long chain fatty acids (Rashid and Anwar 2008a). The mono-
alkyl esters that are the main chemical species of biodiesel, have properties similar to
diesel fuel (Fernando et al. 2007) and it can be used in modern diesel engines in its
pure form (B100) or may be blended with petroleum diesel (Lebedevas and
Vaicekauskas 2006). Although the range of biodiesels available reveals the flexibility
and potential of the biodiesel industry, this potential has not been fully realised by
first-generation biodiesel systems due to serious social and environmental concerns
that it will lead to food price increases and creates pressure on the agricultural lands
usually used for food production. Those concerns make first-generation biodiesel
unlikely to be sustainable. Therefore, first-generation biodiesels are limited in their
ability to contribute to climate change mitigation, economic growth, and as substitute
for petroleum production (Bioenergy 2008). Consequently, there was a need to
develop new technologies for producing alternative feedstocks to overcome the
major short-comings of the supply of first-generation biodiesel fuels. The biodiesels
obtained from these new technologies have been defined as second-generation
biodiesels, which are generally produced from non-edible feedstocks (Posten and
Schaub 2009).
On the other hand, the sustainability of any new biodiesel fuels depends on their
quality and suitability for use in the internal combustion engine, with the majority of
current vehicle engines not optimised for the utilisation of biodiesel. Therefore, these
Chapter 1: Introduction 3
engines are unlikely to operate efficiently when using biodiesel and will often
experience technical problems such as carbon deposition, corrosion, high lubricating
oil contamination, poor low temperature performance, heavy gum and wax formation
compared to petroleum diesel (Jayed et al. 2009). The distinctions between
petroleum diesel and biodiesel may be attributed to variations in their physical
properties and chemical composition. Petroleum diesel is composed of hundreds of
compounds boiling at different temperatures (determined by the petroleum refining
process and crude oil raw material), whereas biodiesel contains compounds,
primarily C8 to 24 carbon chain length alkyl esters (determined entirely by the
feedstock) (Graboski et al. 2003). Besides the major fatty ester components, minor
constituents of biodiesel include intermediary mono- and di-glycerides and residual
triglycerides resulting from the transesterification reaction, as well as methanol, free
fatty acids etc (Knothe 2009). As engines are manufactured for use with petroleum
diesel, OEM’s (Original Equipment Manufacturers) and industry associations have
been cautious in their acceptance of biodiesel as a fuel source, especially those from
new sources and biodiesel blends because these cannot be easily verified (Haseeb et
al. 2011b)
Australia has a large land area on which oil seed crops can be cultivated for biodiesel
production. Therefore, Australia has the potential to become a leading biodiesel
producer. However, contrary to this opportunity, Australian energy supply has
consistently relied upon imported liquid fuels, and the market for biodiesel
production in Australia has not been expanded over the last few years. The factors
contributing to demand for biodiesel in Australia are mainly an inability to supply
consistent and reliable quality fuel, limited acceptance by consumers and concerns
regarding engine warranties and performance ((APAC) 2009). In order to enhance
sustainability and to position Australian biodiesel as an attractive and superior future
transport fuel, it is necessary to evaluate its suitability as a diesel engine fuel.
However, little is known about the optimal use of second-generation biodiesel in
modern engines. There is very little knowledge on modelling the combustion of
biodiesel in engines and at present, there are no comprehensive models for
optimising the benefits of second-generation biodiesel, in terms of fuel quality and
suitability for diesel engines.
4 Chapter 1: Introduction
1.2 RESEARCH QUESTIONS
Based on the background study of these topics, the key research questions were as
follows:
1. What are the potential sources of second-generation biodiesel feedstock?
2. What are the key parameters that influence biodiesel quality as a diesel
engine fuel?
3. What are the Australian native plants that are able to produce substantial
amounts of non-edible bio-oil?
4. Are the non-edible bio-oils produced from Australian native plants able to
meet biodiesel quality standards?
5. What are the optimum conditions for the production of biodiesel from
Australian native plants?
6. Which Australian native species are the most suitable candidates for future
generation biodiesel feedstocks?
7. Are biodiesels made from Australian native plants suitable for use in
conventional diesel engines?
1.3 AIMS AND OBJECTIVES
1.3.1 Aims
The aim of this study was explore the potential of Australian native plants as a source
of second-generation biodiesel. The aim was achieved by undertaking several
experimental and numerical measurements using these biodiesels.
1.3.2 Specific study objectives
In an effort to achieve the aim of this research, the below objectives were formulated
and satisfied, according to the following order:
Chapter 1: Introduction 5
1. Determine the correlation between chemical composition of biodiesel and
important fuel properties.
2. Develop a numerical model for the estimation of properties of the biodiesel.
3. Produce/obtain second-generation biodiesel from different Australian native
plants.
4. Apply the numerical model to investigate the fuel quality of second-generation
biodiesel produced from Australian native plants.
5. Production process optimisation of second-generation biodiesel from Australian
native plants.
6. Experimentally investigate the effect of biodiesel physico-chemical properties on
diesel engine performance and emission.
7. Experimentally investigate diesel engine performance and emissions using
second-generation biodiesel produced from Australian native plants.
1.4 SIGNIFICANTCE OF THIS PROJECT
Recognising the importance of developing sustainable energy sources for the future,
oil companies, governments and non-government organisations are beginning to
invest more research funds into the development of second-generation biodiesel fuels
((APAC) 2009). However, these types of biodiesel, particularly from Australian
native plants, have not been adequately evaluated as a diesel engine fuel and are yet
to start commercial production. The fuel quality of Australian second-generation
biodiesel and the suitability of its use in conventional diesel engines is still unknown.
In order to overcome these barriers and improve the feasibility of producing second-
generation biodiesel from native species, this project investigated the use of
Australian native plants as a second-generation biodiesel feedstock. The
comprehensive experimental study and numerical model was designed to yield a
sensible and simplified characterisation of biodiesel in terms of its physico-chemical
properties and its performance in modern diesel engines. Therefore, the results of this
study will facilitate a more rapid uptake of renewable energy systems, by
investigating new second-generation biodiesels for the public and industries. In this
6 Chapter 1: Introduction
work, the Australian native crops considered as potential feedstocks for biodiesel
production were non-edible, high oil yield, commercially unavailable and not yet
experimentally investigated. The outcomes of this project are expected to trigger
great interest around Australia in relation to the use of second-generation biodiesels.
Establishing the wide-spread use of second-generation biodiesel will reduce the
dependence on crude oil imports and therefore, increase the stability of Australia’s
fuel market and improve its balance of trade. The range of ecological benefits
include massive reductions in greenhouse gas emissions, as well as reductions in
sulphur dioxide emissions (which are one of the main causes of acid rain) and other
cancer causing emission such as benzene. There will also be potential benefits for
agricultural and rural development, including new jobs and income generation.
Therefore, this project is expected to generate new knowledge in relation to how
Australian native plants can significantly contribute to overcoming the global energy
and environmental crisis, and at the same time add value to the Australian economy
and its development.
1.5 THESIS OUTLINE
This section gives a brief outline of the remaining chapters of this thesis. There are a
total of nine chapters which systematically address the key topics of this thesis,
according to five stages. Figure 1.1 presents the relationship between each of the
chapters and stages presented in this thesis. Further discussions on how the outputs
of each chapter form a unified research program are provided below:
In stage one (Chapter 2), a comprehensive literature review of current biodiesel
technology and artificial neural network (ANN) modelling tools for selecting future
generation biodiesel was conducted. This chapter begins with an overview of
biodiesel feedstocks, production processes, chemical compositions, standards,
physico-chemical properties and performance as a diesel engine fuel. The limitations
of commercially available fist-generation biodiesel feedstocks over second-
generation biodiesel feedstock were discussed. Then, the application of ANN in
modelling key biodiesel quality parameters and combustion performance in
Chapter 1: Introduction 7
automobile engines was reviewed. This review found that all biodiesels are fatty acid
methyl esters (FAME), produced from raw vegetable oil and animal fat. The
chemical compositions of biodiesel, in terms of fatty acid profile, are different from
one feedstock to the next; even for the same biodiesel feedstock from a different
origin. Moreover, the fatty acid profile of biodiesel determines the important fuel
properties of biodiesel and hence, the quality of biodiesel as an internal combustion
engine fuel. Although, ANN modelling is most accurate approach for prediction
however, some literatures found that this technique better than many of the linear and
non-linear statistical techniques. Therefore, in the next stage of this study,
correlations between the chemical composition of biodiesel and fuel quality were
investigated. The findings of this study have been published in the journal of
Energies.
In the second stage of this study (Chapter 3), experimental and statistical
investigation of the important chemical composition and fuel properties of biodiesel
were conducted. Section 3.3 describes the data collection methods, while in Section
3.4.4, a correlation study between the properties of biodiesel and its chemical
composition are analysed using principal component analysis (PCA). The PCA
analysis indicated that individual biodiesel properties have complex correlation with
the parameters of chemical composition. The average member of double bonds
(ANBD) and poly-unsaturated fatty fraction (PUFA) are the most most-influential
parameters that affect all biodiesel properties. Figure 3-9 shows the relationship
between eight important biodiesel properties and chemical composition. Based on
these results, several artificial intelligence-based models were developed to predict
specific biodiesel properties based on its chemical composition. The developed
models were tested via simulation studies using an unknown data set (the results of
which are presented in Section 3.4.6), which demonstrated that ANN was able to
predict the relationship between biodiesel chemical composition and fuel properties.
Therefore, the ANN model developed in this study could be a useful tool in
estimating biodiesel fuel properties, instead of undertaking costly and time
consuming experimental tests. Such an attempt was made in the next stage of this
study, when estimating fuel properties of biodiesels obtained from several Australian
native oil seed plants.
8 Chapter 1: Introduction
In stage three (Chapter 4), eleven Australian native plants were investigated for the
potential production of second-generation biodiesel. The native plants investigated
were: Beauty leaf (Calophyllum inophyllum), Candle nut (Aleurites Moluccana),
Blue berry lily (Dianella Caerula), Queen palm (Syagrus Romanzoffiana), Castor
(Ricinus Communis), Bidwilli (Brachychiton Bidwilli), Karanja (Pongamia Pinnata),
Whitewood (Atalaya Hemiglauca), Cordyline (Cordyline Manners – Suttoniae),
Flame tree (Brachychiton Acerifolius), Chinese rain (Koelreuteria Formosana). This
chapter gives a brief description of the investigated native plants for biodiesel
production, followed by the oil extraction techniques from the dry seeds. During oil
extraction, oil yields were measured and the results were presented in Figure 4-16.
Section 4.5 presents the chemical composition of second-generation biodiesel
produced from the investigated feedstocks. Important fuel properties were estimated
using the ANN model developed in the previous chapter and the results are discussed
in Section 4.6. The native species feedstocks were then evaluated and their quality
was compared with each other using the multi-criteria decision method (MCDM)
software PORMETHEE-GAIA. In addition, sensitivity analysis of native plant
ranking was carried out by changing the weighting three important criteria, OY, OS
and CFPP. Results of this analysis indicated that Beauty leaf was one of the best
potential candidates for second-generation biodiesel production among the native
plants investigated. Therefore, in the next stage of this study, biodiesel was produced
form Beauty leaf oil seeds on a pilot scale, for in-depth experimental investigation.
In stage four of this study, the physico-chemical properties of biodiesel obtained
from Beauty leaf oil seeds were assessed experimentally (Chapter 5) and the
biodiesel production process was optimised (Chapter 6). Bio-oil was extracted from
dry Beauty leaf seeds through three different oil extraction methods. The oil
extraction methods and oil yield results are discusse+d in Section 5.3 and Section 5.4,
respectively. Overall, the highest average oil yield was found to be 51.5% for dry
kernels of Beauty leaf seeds. The important physical and chemical properties of the
extracted Beauty leaf oils were experimentally analysed and compared with
conventional edible vegetable oils (presented in the Section 5.6). In Section 5.7, a
brief description of the two-step transesterification method is provided for the
Chapter 1: Introduction 9
conversion from Beauty leaf oil into biodiesel. In the last part of this chapter, a
PROMETHEE-GAIA analysis was conducted, in order to assess the quality of
Beauty leaf biodiesel and compare its physical and chemical properties with that of
commercially available first-generation biodiesel. The results of the PROMETHEE-
GAIA analysis are presented in the Section 5.10. This investigation indicated that
Beauty leaf oil biodiesel may be a better internal combustion engine fuel compared to
most commercially available biodiesel. Findings of this analysis have been submitted
to the journal of Energy Conversion and Management. In Chapter 6, the three main
factors that drive biodiesel (fatty acid methyl ester (FAME)) conversion from
vegetable oil (triglycerides) have been studied using a response surface methodology
(RSM), based on a Box-Behnken experimental design. The factors considered in this
study were catalyst concentration, methanol to oil molar ratio and reaction
temperature. Linear and full quadratic regression models were developed to predict
FFA and FAME concentration, and to optimise the reaction conditions which are
shown in Equations 6-3 to 6-6. The significance of these factors and their interaction
was determined using analysis of variance (ANOVA). The good agreement between
model outputs and experimental results found in this chapter demonstrated that this
methodology may be useful for the optimisation of industrial biodiesel production
from Beauty leaf oil and possibly other industrial processes as well. This chapter has
been published to the journal Energies. Finally, the Beauty leaf biodiesel obtained in
this stage was used for testing the automobile diesel engine in the next stage of this
study.
In the last stage of this study (stage five), experimental investigations were
conducted on multi-cylinder automobile diesel engines using artificially prepared
biodiesel (methyl ester) with controlled chemical composition and second-generation
biodiesel produced from Beauty leaf oil. In Chapter 7, effects of the physico-
chemical properties of biodiesel on diesel engine performance and emissions were
experimentally investigated. Alongside neat diesel, four biodiesels with various
carbon chain lengths and degrees of unsaturation were used at three blending ratios
(B100, B50, B20) in a common rail engine. Experimental results of the chemical and
physical properties for the tested biodiesel samples are presented in Section 7.2. The
variation in diesel engine performance and emissions with changes in the physico-
10 Chapter 1: Introduction
chemical properties of biodiesel is presented in Section 7.3. The experimental results
indicated that particle emissions consistently decreased with reductions in fuel
oxygen content, regardless of the proportion of biodiesel in the blends; whereas it
increased with fuel viscosity, carbon chain length and unsaturation percentages of the
biodiesel. Overall, it is evident from the results presented in this chapter that the
chemical composition and physical properties of biodiesel is important for
determining diesel engine performance and emissions. Therefore, the results of this
chapter are deemed to provide a basis on which the diesel emissions from second-
generation biodiesel can be interpreted. This chapter has been published in the
journal Fuel. In Chapter 8, an investigation was conducted to experimentally
evaluate the suitability of Beauty leaf biodiesel as an automobile engine fuel. A four
cylinder common rail diesel engine was run with 5% and 10% blends of Beauty leaf
biodiesel with conventional diesel, following a standard engine testing equipment
and procedure. The important engine performance and emission indicators were
measured for both Beauty leaf biodiesel blends and petroleum diesel. Section 8.3
summarises the engine testing results and presents a discussion on individual engine
emission and performance indicators. This study found that Beauty leaf oil biodiesel
can be utilised in conventional automobile diesel engines without any engine
modification. Although variations in engine performance and emissions were
observed, due to differences in the physicochemical properties of commercial diesel
and biodiesel produced from Beauty leaf biodiesel, such variations are not unusual
when testing an engine dedicated for petroleum diesel only and are commonly found
during engine testing using only commercially available biodiesel. In Chapter 9, the
conclusions, limitations and recommendations for future work of this thesis are
summarised.
Chapter 1: Introduction 11
Chapter 1: IntroductionThesisAim:InvestigatethepotentialofAustraliannativeplantsasasourceof
second-generationbiodiesel
Chapter 2: Literature review
TheUseofArtificialNeuralNetworksforIdentifyingSustainableBiodieselFeedstock
Publication:Journalof Energies.Specialissue" AlternativeFuelsfortheInternalCombustionEngines(ICE)2013"Vol.6,pp.3764-3806.
Chapter 3: Artificial neural network (ANN) model development
Correlationbetweenchemicalcompositionandpropertiesofbiodiesel– aprincipalcomponentanalysis(PCA)andartificialneuralnetwork(ANN)
approach
Chapter 4: Second-generation biodiesel production from Australian native plants in laboratory scale
PreparedandtestthebiodieselfromAustraliannativeplants
PropertiesestimationusingANNmodels
PROMETHEE-GAIAanalysistorankthebiodiesels
Pilot scale beauty leaf biodiesel production
Chapter5: Physio-chemicalassessmentofbeautyleafbiodieselassecond-generationbiodiesel
feedstockPublication:SubmittedtoEnergy
andManagement
Chapter6: biodieselproductionfromnon-ediblebeautyleafoil:processoptimisationusingresponsesurface
methodology(RSM)Publication: Energies.Vol.7(8),pp.
5317-5331,2014.
Diesel engine testing with biodiesel
Chapter7: Particleemissionofbiodieselswithdifferentphysicochemicalproperties
Publication: JournalofFuel,Vol.134,pp.201-208,2014
Chapter 9: ConclusionConclusionarisefromthisthesisLimitationandRecommendationsforfuturework
Stage 1
Stage 2
Stage 3
Stage 4
Chapter8: Automobiledieselenginetestingwithbeautyleaf
biodiesel
Stage 5
Figure 1-1: Outline on the thesis
Chapter 1: Literature Review 13
Chapter 2: Literature Review
The Use of Artificial Neural Networks for Identifying
Sustainable Biodiesel Feedstock
Md I. Jahirul1,*, Richard J. Brown1, Wijitha Senadeera1, Ian O’Hara2 and Zoran
Ristovski1
1 Biofuel Engine Research Facility (BERF), Science and Engineering Faculty,
Queensland University of Technology (QUT), Brisbane, Australia
2 Centre for Tropical Crops and Biocommodities (CTCB), Queensland University
of Technology (QUT), Brisbane, Australia.
Publication: Journal of Energies. Special issue " Alternative Fuels for the Internal
Combustion Engines (ICE) 2013" Vol. 6, pp. 3764-3806.
Author Contribution
Contributor Statement of Contribution
Md I. Jahirul Analysed the literature and drafted the manuscript Signature
Richard J. Brown Supervised the project and revised the manuscript
Wijitha Senadeera Supervised the project and revised the manuscript
Ian O’Hara Supervised the project and revised the manuscript
Zoran Ristovski Supervised the project and revised the manuscript
Principal Supervisor Confirmation
I have sighted email or other correspondence from all co-authors confirming
their certifying authorship.
Name
Dr Wijitha Senadeera
Signature
Date
14
Abstract
Over the past few decades, biodiesel produced from oilseed crops and animal fat
is receiving much attention as a renewable and sustainable alternative for
automobile engine fuels, and particularly petroleum diesel. However, current
biodiesel production is heavily dependent on edible oil feedstocks which are
unlikely to be sustainable in the longer term due to the rising food prices and the
concerns about automobile engine durability. Therefore, there is an urgent need
for researchers to identify and develop sustainable biodiesel feedstocks which
overcome the disadvantages of current biodiesel feedstocks. On the other hand,
artificial neural network (ANN) modelling has been successfully used in recent
years to gain new knowledge in various disciplines. In this study, recent literature
has been reviewed to assess the state of the art on the use of ANN as a modelling
tool for future generation biodiesel feedstocks. Biodiesel feedstocks, production
processes, chemical compositions, standards, physico-chemical properties and in-
use performance are discussed. Limitations of current biodiesel feedstocks over
future generation biodiesel feedstock have been identified. The application of
ANN in modelling key biodiesel quality parameters and combustion performance
in automobile engines are also discussed. This review has determined that ANN
modelling has a high potential to contribute to the development of renewable
energy systems by accelerating biodiesel research.
Keywords: Renewable Energy, Biodiesel, Artificial Neural Networks (ANN),
Second-Generation Feedstock
Chapter 2: Literature Review 15
1 INTRODUCTION
Since the beginning of the Industrial Revolution in the late 18th and early 19th
century, energy has become an indispensable tool for mankind, contributing to
economic growth and increased standards of living. World primary energy demand is
expected to grow by 1.6% per annum over the period 2010 to 2030, which will
require 39% additional energy by 2030 (BP 2012). There are many potential energy
sources around us from which energy can be converted for use and stored. These
sources can be classified as fossil, fissile and renewable. Fossil energy sources were
formed thousands of years ago, and are not renewable in a short time horizon. They
include liquid crude oil, coal, natural gas and tar sands. The major fissile energy
sources are uranium and thorium that are fissionable by neutrons with zero kinetic
energy. Renewable energy is generated from natural sources such as biomass, solar,
wind and geothermal resources.
Most of the primary energy used today comes from fossil-based resources,
predominantly crude oil (35%), coal (29%) and natural gas (24%), while nuclear and
renewable resources account for 7% and 5% of global energy consumption,
respectively (BP 2012). Fossil-based resources are therefore the single largest source
of energy, representing 88% of the total World consumption. However, fossil
resources are being consumed rapidly. Based on current production scenarios, it is
expected that a peak of global oil production will occur between 2015 and 2030
(Demirbas 2007). Therefore, fossil resources have practical limitations in their
capacity to supply future global energy requirements in which there are currently few
large scale alternatives available. Moreover, combustion of fossil fuels results in
greenhouse gas emissions and contributes to anthropogenic climate change. Despite
global measures such as the Kyoto Protocol and scientific innovation, atmospheric
CO2 concentration continues to increases and is exceeding benchmark levels much
earlier than had previously been predicted (Weitzman 2007).
16 Chapter 2: Literature Review
The transportation sector globally is the third largest energy consumer after the
industry and the building sectors which is the fastest growing sector. By 2030, the
energy consumption and CO2 emissions from this sector are predicted to be 80%
above the levels of today (Metz 2007). The transportation sector is also the sector
most reliant upon petroleum fuels (primarily through the crude oil-derived liquid
products gasoline and diesel). The sector currently consumes 30% of crude oil
globally and this is predicted to increase to 60% by 2030 (Luque et al. 2008).
Furthermore, the availability of conventional crude oil is geographically restricted
impacting on the security and cost of supply.
These issues have forced researchers to seek alternative carbon neutral transport fuels
which promote sustainable development, energy efficiency and environmental
preservation(Jahirul et al. 2010). As of yet, few large scale commercially viable
options exist for the transportation sector. Moreover, cars with no tailpipe
greenhouse gas emissions (e.g. electric, solar, hydrogen) are a long way from being
viable across the sector. Therefore, the development of sustainable long-term
solutions using alternative fuels is essential (Lapuerta et al. 2008; Jahirul et al. 2010)
As a consequence, biodiesel produced from oil crops and animal fats, is receiving
much attention as an alternative to conventional petroleum fuels. In particular, fuels
produced from biomass feedstocks have emerged as one of the more promising and
environmentally sustainable renewable energy options. Fuels produced from these
technologies are referred to as biofuels. Biofuels offer many benefits over
conventional petroleum fuels, including the wide regional distribution of biomass
feedstocks, high greenhouse gas reduction potential, biodegradability and their
contribution to sustainability (Reijnders 2006).
Biofuels produced by conventional technologies (ethanol and biodiesel) typically
contain oxygen levels of 10%–45% by mass, while petroleum fuels (gasoline and
diesel) contain very low oxygen levels. This makes the chemical properties of
biofuels more favourable for complete combustion, although it reduces energy
density. In addition, biofuels typically have very low sulphur contents
Chapter 2: Literature Review 17
and often low nitrogen contents reducing the production of potentially harmful
emissions. Biomass resources can be used to produce a variety of biofuels. This
includes liquid fuels such as biodiesel ethanol, methanol, and Fisher-Tropsch diesel;
and gaseous fuels such as hydrogen, syngas and methane. Liquid biofuels are
primarily used in vehicles, but can also be used in stationery engines or fuel cells for
electricity generation. Biodiesel is widely used as an alternative fuel for diesel
engines, whereas ethanol is used as a substitute for automotive gasoline (Demirbas
2007).
Biodiesel is currently produced in commercial quantities from edible oil feedstocks
such as soybean oil, palm oil, and canola oil. Biodiesels produced from these
feedstocks are generally referred to as first-generation biodiesels. Although
biodiesels from these feedstocks offer reductions in greenhouse gas emissions
(GHG) and improves domestic energy security, first-generation biodiesels are
unlikely to be sustainable in the longer term due to land use impacts and the price
and social impacts associated with using food-based feedstocks. Second-generation
biodiesels produced from non-edible feedstocks have the potential to overcome the
disadvantages associated with first-generation feedstocks while addressing many of
the challenges of climate change and energy availability.
However, the majority of current vehicle engines are not optimised for the use of
biodiesel. When using biodiesel as a fuel, optimised engines may show increased
problems with carbon deposition, corrosion, lubricating oil contamination, poor low
temperature performance, and heavy gum and wax formation compared to petroleum
diesel(Jayed et al. 2009). The differences in performance between petroleum diesel
and biodiesel may be attributed to the variation between their physical properties and
their chemical compositions. Petroleum diesel is composed of hundreds of
compounds boiling at different temperatures (determined by the petroleum refining
process and crude oil raw material), whereas biodiesel contains primarily 6–24
carbon chain length alkyl esters (determined entirely by the feedstock) (Graboski and
McCormick 1998). In addition to these alkyl esters, biodiesel may also contain minor
amounts of mono-, di- and triglycerides resulting from incomplete trans-
18 Chapter 2: Literature Review
esterification, methanol, free fatty acids, chlorophyll (in the case of algae) and sterols
(Knothe 2008).
As engines are currently manufactured to be optimised for petroleum fuels, Original
Equipment Manufacturers (OEMs) and industry associations have shown a cautious
response in their acceptance of biodiesel, especially from new feedstocks or
processes (Haseeb et al. 2011b). Engine manufacturers and users may be reluctant to
realise the potential of using second biodiesel in engines (especially from new
feedstocks) because its suitability as an automobile engine fuel may not be readily
verified. As a consequence, the potential of second-generation biodiesel is still
largely unexplored and is yet develop as a mainstream transportation fuel. This is in
spite of the fact that there are many such potential biodiesel feedstocks that have
already been identified including oilseed plants (Ahmad, Yasin, et al. 2011;
Banković-Ilić, Stamenković and Veljković 2012; Ashwath 2010b) and marine algae
(Mata, Martins and Caetano 2010; Chisti 2007). The slow uptake of biofuels can be
associated with the challenges of ensuring a consistent supply of feedstock, feedstock
cost, and the lack of experimental data to prove the quality of the fuel (resulting in
limited acceptance by consumers who are concerned about engine warranties and
performance) ((APAC) 2009). Undertaking experiments with automotive engines
and measuring multiple fuel quality parameters requires considerable quantities of
fuel, which can be a challenge from new sources. Fuel testing requires sophisticated
equipment and expert personnel which can be costly. These concerns potentially
restrict the progress of scientific research to establish widely acceptable second-
generation biodiesels.
In recent years, ANN modelling techniques have gained in popularity due to their
ability to accurately predict from small data sets (examples) rather than from larger
data sets requiring costly and time-consuming studies and experiments. ANN has
been successfully applied in various disciplines, including neuroscience (Alkım,
Gürbüz and Kılıç 2012), mathematical and computational analysis (Costa, Braga and
De Menezes 2012), learning systems (Carrillo et al. 2012), engineering design and
application (Samura and Hayashi 2012; Gao et al. 2012; Minnett et al. 2011b) and
Chapter 2: Literature Review 19
chemical and environmental engineering (Zendehboudi et al. 2012; Roosta, Setoodeh
and Jahanmiri 2011; Kumar 2009b; Ali Ahmadi et al. 2013).
In this paper, the potential of ANN modelling techniques in identifying sustainable
future generation biodiesel feedstock are identified based on the most recent
literature. Current biodiesel technology (feedstocks, properties, production processes,
chemical compositions and factors contributing to fuel quality and its applicability as
an alternative fuel) and the application of ANN in fuel technology and the potential
second-generation biodiesel feedstock is also discussed. Findings from this literature
review contain valuable information to assist biodiesel manufacturers and researchers
to make important decisions to accelerate the technological development of biofuels.
2.1 BIODIESEL
Fatty acid methyl or ethyl esters, commonly referred to as “biodiesel”, are a liquid
fuel alternative to diesel. They are made from agricultural products, forest organic
matter and animal fat feedstocks. Biodiesel is the only currently available alternative
transport fuel made from oilseed crops and animal fat which can be used directly in
conventional unmodified diesel engines. Biodiesel is safer to handle, store and
transport compared to petroleum diesel because it is biodegradable, non-toxic and
has a higher flash point than diesel. One of the major advantages of biodiesel is that
it has potential to reduce dependency on imported petroleum through the use of
domestic feedstocks for production (Demirbas 2008b; Fernando et al. 2007).
In fuel property terms, biodiesel has a higher cetane rating than petroleum diesel
which improves engine performance. In addition, it has better lubricant properties
than petroleum diesel which can extend engine life (Haseeb et al. 2011b). The use of
biodiesel reduces particulate emissions by up to 75% when compared with
conventional diesel fuel. Biodiesel also substantially reduces unburned
hydrocarbons, carbon monoxides and particulate matters, including elimination of
sulphur dioxide in exhaust emissions. The exhaust emissions of particulate matter
20 Chapter 2: Literature Review
from biodiesel have been found to be 30% lower than overall particulate matter
emissions from fossil diesel. The exhaust emissions of total hydrocarbons are up to
93% lower for biodiesel than diesel fuel (Hoekman et al. 2012).
As a fuel, there are currently several disadvantages to using biodiesel in diesel engine
applications. These differences mainly result from the difference in chemical
compositions between petroleum diesel and biodiesel. These major disadvantages
are: lower energy density, higher viscosity, higher copper strip corrosion and issues
with the degradation of fuel in storage for prolonged periods. Biodiesel also has a
higher cold-filter plugging point temperature than fossil diesel which means it will
crystallise into a gel at lower temperatures when used in its pure form. Biodiesel can
also cause dilution of engine lubricant oil, requiring more frequent oil changes than
when using petroleum diesel fuels in conventional diesel engines. This increase in
dilution and polymerisation of engine sump oil is due to the higher viscosity at lower
temperatures of biodiesel compared to petroleum diesel (Jayed et al. 2009).
2.1.1 Biodiesel feedstock
Feedstocks for biodiesel production can be classified into four groups. These are: (1)
virgin vegetable oil feedstocks such as rapeseed, soybean, sunflower, palm oil; (2)
waste vegetable oils; (3) animal fats including beef tallow, lard and yellow grease;
and (4) non-edible oils such as jatropha, neem oil, castor oil and the prevalence of
these feedstocks varies around the World (Figure 1). The regional availability of
feedstocks for biodiesel production depends greatly on climate, soil conditions and
options for alternate land use. Consequently, different regions are focusing their
efforts on different feedstocks (Lin et al. 2011). As an example, the widespread use
of soybeans in the USA as a food product has led to the emergence of soybean
biodiesel in that country. In Europe, rapeseed is the most common source of
biodiesel production. In India and Southeast Asia, the jatropha tree is used in
biodiesel production, and in Malaysia and Indonesia, palm oil is used as a significant
biodiesel source.
Chapter 2: Literature Review 21
Figure 2-1. Biodiesel feedstocks around the world
2.1.2 First and second-generation biodiesel
Many potential feedstocks for biodiesel production have been investigated including
soybean oil (Goodrum and Geller 2005), sunflower oil, corn oil, used fried oil, olive
oil (Anastopoulos et al. 2005), rapeseed oil (Terry, McCormick and Natarajan 2006),
lesquerella oil, milkweed (Asclepias) seed oil (Holser and Harry-O’Kuru 2006a),
Jatropha curcas, Pongamia glabra (karanja), Madhuca indica (mahua) and
Salvadora oleoides (Pilu) (Kaul et al. 2007b), palm oil (Raadnui and Meenak 2003)
and linseed oil (Agarwal 1999). Most of those are produced from edible oil feedstock
and known as first-generation biodiesels (Rashid and Anwar 2008b). Although the
range of biodiesels available reveals the flexibility and potential of the biodiesel
industry, this potential is challenged by social and economic concerns. The most
contentious issue affecting the production of first-generation biodiesel is the use of
agricultural land for biodiesel production. This issue is commonly referred to as the
“Food verses Fuel” debate, whereby the main two issues are the use of edible crops
for biodiesel production, and the land space devoted to the growing of inedible crops.
First-generation biodiesels are primarily made from edible vegetable oils, therefore
22 Chapter 2: Literature Review
crop space used for biodiesel production limits the crop space available for food
production. Farmers of these crops now have the choice to sell to the biodiesel
production market or the food market. Farmers must find ways to retain their
viability, and if a higher price is offered by the biodiesel production market, this will
more often than not be the option they will choose. This is of particular concern in
disadvantaged countries where crops used for biodiesel production displace the
production of food crops, causing shortages. Supply and demand dictates that a
shortage will cause a price rise, where countries such as Malaysia are currently
experiencing (Bradsher 2008). This issue became a global debate due to the 2007–
2008 world food price crises. Differing arguments exist for the cause of this crisis;
however, there has been widespread speculation that the increasing consumption of
biodiesel contributed to the food shortage and subsequent price increases (Kingsbury
2012). Although there is a global demand for biodiesel due to its proven benefits and
its potential to decrease dependence on fossil fuels, this should not lead to of people
suffering from hunger. As a result, first-generation biodiesels are unlikely to be
sustainable in the longer term, having limitations in their ability to contribute to
socio-economic growth (Bioenergy 2008). Therefore, an alternative must be
considered which eliminates the downfalls of first-generation biodiesels. Research is
currently underway in second and third-generation biodiesels which are targeted at
addressing the “Food verses Fuel” debate (Posten and Schaub 2009). A comparison
of first and second-generation biodiesel in terms of feedstock, advantages and
associated problems are shown in Table 2-1.
Table 2-1. Advances in biodiesel technology
Technology First-generation biodiesel Second-generation biodiesel
Feedstock Edible vegetable oil and animal fat Non-edible feedstock Cheap and abundant biomass
Advantage Commercially available Renewable Environmentally friendly Economic
Renewable Not competing with food Environmentally friendly More sustainable Social security
Problems Limited feedstock Competing with food
In development stage Unreliable sources of feedstock
Chapter 2: Literature Review 23
Engine durability Not likely to be sustainable
High production cost
2.1.3 Potential second-generation biodiesel feedstock
Non-edible oils, which are considered as second-generation biodiesel feedstock,
currently contribute less than 5% of total global biodiesel production (Banković-Ilić,
Stamenković and Veljković 2012). Therefore, second-generation biodiesels are yet to
make a significant impact on the mainstream alternative energy system. One of the
main reasons for this is the lack of reliable and commercially viable feedstock
sources. The feedstocks most often used in second-generation biodiesel production
are jatropha, cottonseed, mahua, and waste cooking oils. A considerable amount of
research has been carried out in an effort to determine alternative feedstocks for
biodiesel production over the last few years. A large number of non-edible oilseed
plants and algae species have been identified around the world which could be a
valuable source for future generation biodiesel. For example, Ashwath (Ashwath
2010b) investigated the biodiesel potential of more than 200 Australian native plants
species, based on their ability to produce abundant quantities of seeds in their natural
environment in Queensland. Among those, 20 species have been identified as
containing more than 20% of non-edible oil in their seed. This study has concluded
that Australia has potential to be a major source of next-generation biodiesel
feedstock, having vast areas of grazing (cleared) and degraded (mined) land on
which biodiesel crops could successfully be established.
In Thailand, 27 types of plants have been found to contain more than 25% (w/w) of
non-edible oil in the seed (Winayanuwattikun et al. 2008). Mohibbe Azam et al.
(Mohibbe Azam, Waris and Nahar 2005a) identified 75 different non-edible plant
seed oils produced from plant species growing naturally in India, and containing
more than 30% of oil in their seeds or kernels. Kumar and Sharma (Kumar and
Sharma 2011) presented a brief description of the biology, distribution and chemistry
of fifteen potential non-edible oilseed plants from India. Moreover, in a critical
review, Balat and Balat (Balat and Balat 2008) reported that there are more than 350
potential oilseed crops for biodiesel production which have been identified, with
24 Chapter 2: Literature Review
most of them being non-edible, yet useful as alternative fuel sources for diesel
engines.
Besides non-edible oilseed plants, researchers worldwide are giving more attention to
algae species as a promising source of second-generation biodiesel feedstocks. Algae
are usually found in seas, rivers, lakes, soils, walls, in plants and animals—almost
anywhere that there is light for photosynthesis. While growing, they convert sun
energy into chemical energy through photosynthesis and complete an entire growth
cycle every few days (Khan et al. 2009). Their growth and bio-oil production rates
can be accelerated by the addition of specific nutrients and sufficient aeration (Khan
et al. 2009). Therefore, it is possible to find species best suited to local environments
or which have specific growth characteristics, which is not possible with current
first-generation biodiesel feedstocks (e.g., soybean, rapeseed, sunflower and palm
oil) (Mata, Martins and Caetano 2010). Moreover, the oil productivity of many algae
greatly exceeds the oil productivity of even the best producing oil crops (Karmakar,
Karmakar and Mukherjee 2010). For example, some algae species are able to
produce bio-oil at a rate approximately 200 times higher than that of soybean plants
over an acre of land area (Hossain et al. 2008).
Algae also offers some significant advantages in the production of second-generation
biodiesel including: (1) providing a reliable and continuous supply for naturally
growing oil all year round; (2) can be harvested batch-wise nearly all year round; (3)
higher photon conversion efficiency (as evidenced by increased biomass yields per
hectare); (4) utilising salt and waste water streams, thereby greatly reducing
freshwater use; (5) possibility of combining CO2-neutral fuel production with CO2
sequestration; and (6) producing non-toxic and highly biodegradable biofuels
(Schenk et al. 2008). Furthermore, algae species are very diverse in nature with
diversity much greater than that of land plants. Hu et al. (Hu et al. 2008) reported
that over 40,000 species have already been identified, with potentially many more.
These species are commonly classified into multiple major groups which include:
cyanobacteria (Cyanophyceae), green algae (Chlorophyceae), diatoms
(Bacillariophyceae), yellow-green algae (Xanthophyceae), golden algae
(Chrysophyceae), red algae (Rhodophyceae), brown algae (Phaeophyceae),
Chapter 2: Literature Review 25
dinoflagellates (Dinophyceae) and “pico-plankton” (Prasinophyceae and
Eustigmatophyceae). Several additional divisions and classes of unicellular algae
have been described, and details of their structure and biology are available
(Demirbas and Fatih Demirbas 2011; Scott et al. 2010).
The concept of using diverse non-edible oils produced from oilseed plants and algae
species has been explored in depth over the past few decades. At the same time, a
large number of new feedstocks have been identified, based on oil content. Table 2-2
summarises the potential oilseed plants and algae species feedstock for future
generation biodiesel production, focusing on feedstocks containing at least 40% non-
edible oil by dry weight as reported in some recent studies. The chemical
compositions of most of the oil produced from the feedstock listed in the Table 2-2
have also been analysed (Gouveia and Oliveira 2009). However, the production of
those oils is primarily confined to high-value specialty oils with nutritional value,
rather than commodity oils for biofuel and scalable, commercially viable systems
have yet to emerge. In considering any new oil as a candidate for large-scale
production in order to use as a biodiesel fuel source, several other issues should be
considered which include the economic necessities revolving around the production
of biodiesel and its suitability for use as fuel in combustion engines.
However, major technical and economic hurdles are still to be overcome before they
can be widely deployed on a fully commercial scale, and unfortunately, the progress
of research to address these issues is not progressing strongly. The quality and
suitability of biodiesel produced from a large number of non-edible oil-bearing
feedstocks as the alternative to petroleum fuel for automobile engine application is
still uncertain. Therefore, many potential non-edible feedstocks are yet emerging to
mainstream biodiesel production, and it appears that its contribution to the global
energy system is a long way off (Sims et al. 2010). In order to accelerate the progress
of second-generation biodiesel as an alternative automobile fuel, more detailed
research is still needed to ensure that second-generation biofuels will provide
economic benefits whilst fulfilling quality standards and combustion performance in
automobile engines. Such research requires pilot scale biodiesel production plants,
specialised equipment and a skilled workforce - all of which involve high-level
26 Chapter 2: Literature Review
research investment and considerable economic risk. Therefore, it has become
challenging for many countries (especially developing or underdeveloped countries)
to take the initiative to establish their own biodiesel manufacture from locally grown
non-edible feedstock.
Table 2-2: Second-generation biodiesel feedstock containing oil by dry weight
(Banković-Ilić, Stamenković and Veljković 2012; Ahmad, Yasin, et al. 2011;
Ashwath 2010b; Mata, Martins and Caetano 2010; Gouveia and Oliveira 2009; No
2011; Amaro, Guedes and Malcata 2011) Non-edible oilseed plants Microalgae
Feedstock Oil content (% dry wet)
Feedstock Oil content (% dry wet)
Aleurites moluccana 42 Botryococcus braunii 25–75 Argemone mexicana 38–54 Chaetoceros calcitrans 40 Atalaya hemiglauca 18–43 Chlorella emersonii 25–63 Azadirachta indica 45–61 Chlorella minutissima 57 Brachychiton acerifolius 45–50 Chlorella protothecoides 14.6–57.8 Cerbera odollam 54 Chlorella vulgaris 5–58 Cocos nucifera 49–52 Crypthecodinium cohnii 20.0–51.1 Jatropha curcas 20–60 Cylindrotheca sp. 16–40 Euphorbia lathyris 48 Dunaliella tertiolecta 16.7–71.0 Garcinia indica 45.5 Isochrysis galbana 7–40 Hevea brasiliensis 40–60 Nannochloris sp. 20–56 Linum usitatissimum 35–45 Nannochloropsis sp. 31–68 Madhuca indica 35–50 Neochloris oleoabundans 29–65 Melia azedarach 19–45 Nitzschia sp. 16–47 Michelia chaampaca 45 Phaeodactylum tricornutum 18–57 Nicotiana tabacum 36–41 Porphyridium cruentum 9.0–60.7 Pongamia pinnata 21–50 Scenedesmus dimorphus 16–40 Putranjiva roxburghii 42 Scenedesmus obliquus 11–55 Raphanus sativus 40–54 Schizochytrium sp. 50–77 Ricinus communis 20–50 Skeletonema costatum 13.5–51.3 Salvadora oleoides 45 Simmondsia chinensis 45–55 Syagrus romanzoffiana 38–44
2.1.4 Production of biodiesel
More than 100 years ago, Rudolph Diesel demonstrated the operation of a diesel
engine using vegetable oil as a fuel, hence the potential of using these feedstocks has
been long recognised. However, vegetable oils are extremely viscous, with viscosity
ranging from 10 to 17 times higher than that of petroleum diesel. This makes
Chapter 2: Literature Review 27
vegetable oil unsuitable to use as a direct fuel in the modern diesel engine. As a
consequence, researchers and scientists have developed various methods to reduce
the viscosity of bio-oils to make them suitable for diesel engine use. Some of these
methods include dilution with other fuels, esterification, micro-emulsification,
pyrolysis and catalytic cracking. Of these techniques, esterification is the most
promising and widely used solution due to its high conversion efficiency, simplicity,
low conversion cost and the fuel qualities of the product. Transesterification of bio-
oils with alcohols to produce esters is a widely used technique for commercial
biodiesel production (Lin et al. 2011).
Transesterification is a chemical reaction in which oils (triglycerides) are converted
into esters as shown in Figure 2-2. Triglycerides react with alcohols (e.g., methanol,
ethanol) under acid or base catalysed conditions, producing fatty acid alkyl esters and
glycerol. A catalyst is used to improve the reaction rate and yield. Because the
transesterification reaction is reversible, excess alcohol is used to shift the
equilibrium to favour production of the ester. After the reaction is complete, glycerol
is removed as a by-product. The biodiesel produced may be denominated by the
feedstock used and the ester formed including Fatty Acid Methyl Ester (FAME),
Fatty Acid Ethyl Ester (FAEE), Soybean Methyl Ester (SME) and Rapeseed Methyl
Ester (RME). The total ester content in biodiesel is the measure of the completeness
of the transesterification reaction (Rajendra, Jena and Raheman 2009).
The yield of biodiesel in transesterification is affected by several process parameters.
These parameters include the reaction temperature, the molar ratio of alcohol to oil,
the type and concentration of catalyst and the reaction time (Balat and Balat 2010;
Demirbas 2008a). While conducting experiments to optimise the transesterification
reaction conditions of two feedstock oils (including waste cooking oil and canola
oil), Leung and Guo (Leung and Guo 2006b; Zhang and Jiang 2008a) determined
that the optimal values of these parameters for achieving maximum conversion of
triglycerides to esters depended on the chemical and physical properties of the
feedstock oils. Other research has also determined varying optimum process
conditions for different oil feedstocks as shown in Table 2-3.
28 Chapter 2: Literature Review
Figure 2-2. Transeterification reaction
Table 2-3: Reported optimum conditions for transesterification of oils for biodiesel production.
Feedstock
Reaction parameters Ester
(wt.%) References Temp
(°C)
Alcohol: Oil
(mol:mol)
Catalyst
(wt.%)
Time
(min)
Palm
38.44 6.44:1 1.25 26 98.02
Worapun et al. (Worapun,
Pianthong and Thaiyasuit
2012)
65 6:1 1.0 60 82
Darnoko and Cheryan
(Darnoko and Cheryan
2000)
Cottonseed
60 6:1 0.3 60 96 Hoda (Hoda 2010)
60 12:1 2 480 90 He et al. (Chen et al.
2007)
Rapeseed 65 6:1 1.0 120 96
Rashid and Anwar
(Rashid and Anwar
2008b)
Sunflower
60 6:1 1.0 120 97.1 Rashid et al. (Rashid et al.
2008)
70 3:1 0.28 60 96 Antolin et al. (Antolın et
al. 2002)
Jatropha 60 6:1 1.0 40 98.6
Nakpong and
Wootthikanokkhan
(Nakpong and
Wootthikanokkhan 2013)
Canola 45 6:1 1.0 60 98 Leung and Guo (Leung
and Guo 2006b)
Waste cooking 60 7:1 1.1 20 94.6 Leung and Guo (Leung
and Guo 2006b)
Soybean 70 9:1 0.5 180 99 De et al. (de Oliveira et al.
2005)
Chapter 2: Literature Review 29
Figure 2-3: Soap formation (Rajendra, Jena and Raheman 2009).
Figure 2-4: Acid pre-esterification (Rajendra, Jena and Raheman 2009)
Alkali-catalysed transesterification cannot be directly used to produce high quality
biodiesel from feedstocks containing high levels of free fatty acids (FFA). This is
because FFAs react with the catalyst to form soap (Figure 2-3), resulting in
emulsification and separation problems. To overcome this problem, a pre-
esterification process may be used to reduce the content of FFAs in the feedstock.
A typical pre-esterification processes uses homogeneous acid catalysts, such as
sulphuric acid, phosphorous acid combined with sulphonic acid, or heterogeneous
“solid-acid” catalysts, to esterify the free fatty acids (Zhang and Jiang 2008a) as
shown in Figure 2-4.
2.1.5 Chemical composition of biodiesel
Petroleum diesel fuels are saturated straight chain hydrocarbons with carbon chain
lengths of 12–18, whereas vegetable oils and animal fats consist of 90%–98%
triglycerides, small amounts of mono-glycerides and free fatty acids. The fatty acid
compositions of triglycerides differ in relation to the chain length, degree of
30 Chapter 2: Literature Review
unsaturation and the presence of other functional groups. The fatty acid compositions
are feedstock dependent and are affected by factors such as climatic conditions, soil
type, plant health, and plant maturity upon harvest. Using the carboxyl reference
system, fatty acids are designated by two numbers: the first number denotes the total
number of carbon atoms in the fatty acid and the second is the number of double
bonds indicating the degree of unsaturation. For example, 18:1 designates oleic acid
which has 18 carbon atoms and one C=C double bond. The most common fatty acids
found in biodiesels and their structures are listed in Table 2-4. However, biodiesels
from differing feedstock and origins have variations in the fatty acid in their
molecules, as shown in Figure 2-5.
Table 2-4: Chemical structure of common fatty acid in biodiesels
(Hoekman et al. 2012; Kapdan and Kargi 2006; Singh and Singh 2010; Ramos et al. 2009) Fatty acid Chemical structure
Caprilic (8:0) CH3(CH2)6COOH Capric (10:0) CH3(CH2)8COOH Lauric (12:0) CH3(CH2)10COOH Myristic (14:0) CH3(CH2)12COOH Palmitic (16:0) CH3(CH2)14COOH Palmitoilic (16:1) CH3(CH2)6 CH=CH (CH2)6 COOH Stearic (18:0) CH3(CH2)16COOH Oleic (18:1) CH3(CH2)7 CH=CH (CH2)7 COOH Linoleic (18:2) CH3(CH2)4 CH=CHCH2CH=CH (CH2)7 COOH Linolenic (18:3) CH3(CH2)2CH=CHCH2CH=CHCH2CH=CH(CH2)7 COOH Arachidic (20:0) CH3(CH2)18COOH Behenic (22:0) CH3(CH2)20COOH Erucic (22:1) CH3(CH2)9 CH=CH (CH2)9 COOH
Chapter 2: Literature Review 31
Figure 2-5: Fatty acid profile of various biodiesel fuels
(Hoekman et al. 2012; Kapdan and Kargi 2006; Singh and Singh 2010; Taravus,
Temur and Yartasi 2009; Chuck et al. 2009; Kinoshita et al. 2007; Koçak, Ileri and
Utlu 2007)
2.1.6 Biodiesel standards
Quality standards are crucial for the commercial use of any fuel product. They serve
as guidelines for production, assure customers that they are buying high-quality
fuels, and to provide authorities with approved tools for a common approach to
transport, storage and handling. Modern diesel engines using common rail fuel
injection systems are more sensitive to fuel quality. Therefore, engine and
automotive manufacturers rely on fuel standards in determining consumer
warranties. However, the chemical compositions of biodiesel and petroleum diesel
are very different, and these differences result in varying physico-chemical
properties. In order to improve the viability of biodiesel for as a commercial fuel for
direct replacement of petroleum diesel, the properties of biodiesel need to reflect
functional equivalence with diesel.
Biodiesel can be used as a pure fuel (B100) or blended with petroleum diesel in
varying concentrations. For B100, the most internationally recognised standards are
EN14214 (Europe) and ASTM D-6751 (USA). Both standards are similar in content,
32 Chapter 2: Literature Review
with only minor differences in some parameters (Hoekman et al. 2012). Many other
countries have defined their own standards, which are frequently derived from either
EN14214 or ASTM D-6751 (Hoekman et al. 2012). As a part of the Fuel Quality
Standards Act 2000, the Australian government released a biodiesel fuel standard,
“Fuel Standard (Biodiesel) Determination 2003”. A summary of the major fuel
quality parameters in these standards is detailed in Table 2-5.
Table 2-5: International biodiesel standards
(Singh and Singh 2010; Canakci and Sanli 2008)
Properties Units USA ASTM D-
6751 Europe EN
14214 Australia
Viscosity, 40 °C mm2/sec 1.9–6.0 3.5–5.0 3.5–5.0 Density gm/m3 n/a 0.860–0.900 0.860–0.900
Cetane number - 47 min 51 min 51 min Flash point °C 130 min 120 min 120 min Cloud point °C report report report
Acid number mg KOH/g 0.80 max 0.5 max 0.8 max Free glycerine wt.% 0.02 max 0.02 max 0.02 max Total glycerine wt.% 0.24 max 0.25 max 0.25 max Iodine number - - 120 max n/a
Oxidation stability h - 6 min n/a Monoglyceride Mass (%) - 0.8 max n/a
Diglyceride Mass (%) - 0.2max n/a Triglyceride Mass (%) - 0.2 max n/a
CFPP °C - - −4
2.1.7 Fuel properties
Biodiesel fuel properties vary significantly between feedstocks due to their differing
chemical compositions. Figure 2-6 summarises the key fuel properties of various
biodiesels reported in the more recent literature. The factors that influence biodiesel
fuel properties are discussed below.
Chapter 2: Literature Review 33
Figure 2-6: Variation in fuel properties of various biodiesel
(Ramos et al. 2009; Taravus, Temur and Yartasi 2009; Koçak, Ileri and Utlu 2007;
Benjumea, Agudelo and Agudelo 2010; Sanford et al. 2009; Canakci and Sanli 2008;
Canakci 2005b; Barnwal and Sharma 2005; Alptekin and Canakci 2008; Kinast 2003a)
2.1.7.1 Kinematic viscosity
Viscosity is defined as the resistance to shear or flow; it is highly dependent on
temperature and it describes the behaviour of a liquid in motion near a solid
boundary such as the walls of a pipe. The presence of strong or weak interactions at
the molecular level can greatly affect the way the molecules of an oil or fat interact,
therefore affecting their resistance to flow. Viscosity is one of the most critical
features of a fuel. It plays a dominant role in fuel spray, fuel-air mixture formation
and the combustion process. In a diesel engine, the liquid fuel is sprayed into
compressed air and atomised into small droplets near the nozzle exit. In the
combustion chamber, a fuel form a cone-shaped spray at the nozzle exit which
affects the viscosity affects the atomisation quality, penetration and size of the fuel
droplet (Alptekin and Canakci 2008). Higher viscosities result in higher drag in the
fuel line and injection pump, higher engine deposits, higher fuel pump duties and
increased wear in the fuel pump elements and injectors. Moreover, the mean
diameter of the fuel droplets from the injector and their penetration increases with an
34 Chapter 2: Literature Review
increase in fuel viscosity (Choi and Reitz 1999). Higher pressure in the fuel line can
cause early injection, moving the combustion of the fuel closer to top dead centre,
increasing the maximum pressure and temperature in the combustion chamber (Choi
and Reitz 1999; Lee et al. 2002; Tat and Van Gerpen 2003a). Therefore, fuel
viscosity significantly influences engine combustion, performance and emissions,
especially carbon monoxide (CO) and unburnt hydrocarbon (UHC) (Knothe and
Steidley 2005).
To estimate the influence of biodiesel viscosity on diesel engine exhaust emission,
Ng et al. (Ng, Ng and Gan 2012) conducted experiments on a light-duty diesel
engine using coconut methyl ester (CME), palm methyl ester (PME), soybean methyl
ester (SME) and blends with petroleum diesel. This study found that an increase in
kinematic viscosity by 1 mm2/s has the potential to raise the emitted CO
concentration by 0.02 vol%. UHC was increased by 1 ppm vol. for every 1 mm2/s
rise in kinematic viscosity. Problems with high viscosity in the fuel became more
severe in cold weather as viscosity of biodiesel increases with decreasing
temperature (Joshi and Pegg 2007a). However, very low fuel viscosity is not
desirable because the fuel then doesn’t provide sufficient lubrication for the precision
fit of fuel injection pumps, resulting in leakage or increased wear. Therefore, all
biodiesel standards define the upper and lower limits of biodiesel shown in Table 2-
4.
The viscosity of biodiesel is dependent on its fatty acid composition. A recent study
showed that viscosity increases with increasing length of both the fatty acid chain
and alcohol group (Geller and Goodrum 2004). As the lengths of the acid and alcohol
segments in the ester molecules increased, so did the degree of random
intermolecular interactions and consequently viscosity. The effect becomes more
evident at lower temperatures, where molecular movements are more restricted
(Knothe and Steidley 2007; Rodrigues Jr et al. 2006). However, Refaat (Refaat
2009b) reported that shorter fatty acid chains with longer alcohol moieties display
lower viscosity than ester with longer fatty acid chains and shorter alcohol moieties.
Other factors that influence biodiesel viscosity include: number and position of
double bonds (Cheenkachorn 2004), degree of saturation (Knothe 2005), molecular
Chapter 2: Literature Review 35
weight (Gunstone 2011), branching (Lee, Johnson and Hammond 1995) hydroxyl
groups and the amount of impurities, such as unreacted glycerides or glycerol
(Knothe 2005; Gunstone 2011).
2.1.7.2 Density
Density is an important fuel property that influences the amount of fuel injected into
the engine cylinder. This is because in a diesel engine fuel injection system, pumps
and injectors must deliver a precise amount of fuel to provide proper combustion
(Boudy and Seers 2009; Baroutian 2008). However, fuel injection pumps meter fuel
by volume and not by mass, leading to denser fuel which contains a greater mass for
the same volume. Thus, changes in the fuel density will influence engine output
power due to the different mass of the fuel injected, and this directly affects engine
performance characteristics (Alptekin and Canakci 2008). Moreover, density
increases the diameter of the fuel droplets in the combustion chamber. Since the
inertia of the bigger droplets is relatively large, their penetration in the combustion
chamber will be higher as well (Choi and Reitz 1999). When a fuel with lower
density and viscosity is injected, improved atomisation and mixture formation can be
attained which consequently affects exhaust emissions. Szybist et al. (Szybist et al.
2007) found that fuel density correlated with particulate matter (PM) and NOx
emissions, with higher densities generally causing an increase in PM and NOx
emission in diesel engines. However, while investigating the biodiesel fuel properties
on exhaust emission in a light-duty diesel engine, Ng et al. (Ng, Ng and Gan 2012)
found that fuel properties moderately affect CO emissions but have no significant
impact on NOx, UHC and smoke opacity levels. Density is also a key factor in the
design of reactors, separation processes and storage tanks in biodiesel production
(Veny et al. 2009).
Density of biodiesel is closely related to the fatty acid composition and the purity of
a biodiesel. Studies have shown that density increases with decreasing chain length
and an increasing degree of unsaturation (Blangino, Riveros and Romano 2008; Lang
et al. 2001).
36 Chapter 2: Literature Review
2.1.7.3 Cetane number (CN)
Cetane number (CN) is a widely used diesel fuel quality parameter, and is a
measurement of the combustion quality of diesel fuels during compression ignition.
It is related to the ignition delay (ID) time, that is, the time that passes between
injection of the fuel into the cylinder and the onset of ignition. A shorter ID time
results in a higher CN which provides better ignition properties (Meher, Vidya Sagar
and Naik 2006b). A high CN will help to ensure good cold start properties and will
minimise the formation of white smoke. On the other hand, lower CN may result in
diesel knocking and an increase in exhaust emissions.
Standards have been established worldwide for CN determination. It is measured by
comparing blends of two reference fuels. A long straight-chain hydrocarbon,
hexadecane (C16H34; trivial name “cetane”, giving the cetane scale its name) is the
high quality standard on the cetane scale with an assigned CN of 100. A highly
branched compound, 2, 2, 4, 4, 6, 8, 8,-heptamethylnonane (HMN, also C16H34), a
compound with poor ignition quality, is the low-quality standard and has an assigned
CN of 15. Both the Australian biodiesel standard and the European petroleum diesel
standard EN 590 limit the cetane number to a minimum value of 51, as shown in
Table 2-4.
CN is dependent on the fatty acid composition of a biodiesel. The longer the fatty
acid carbon chains and the more saturated the molecules, the higher the cetane
number (Bajpai and Tyagi 2006; Demirbas 2005). The most significant factor in
lowering CN is the degree of unsaturation. Geller and Goodrum (Geller and
Goodrum 2004) observed that a low CN was associated with highly unsaturated
compounds such esters of linoleic (C18:2) and linolenic (C18:3) acids, whereas high
CNs were observed for esters of saturated fatty acids such as palmitic (C16:0) and
stearic (C18:0) acids. Similar results have been reported by Knothe et al. (Knothe,
Matheaus and Ryan III 2003). Bangboye and Hansen (Bamgboye and Hansen 2008)
observed that a feedstock that is high in saturated fatty esters has a high CN, while a
feedstock with predominantly unsaturated fatty acids has lower CN values.
Chapter 2: Literature Review 37
2.1.7.4 Heating (calorific) Value
Heating value is a significant fuel property which influences the suitability of
biodiesel as an engine fuel, as it indicates the energy content in the fuel. Due to the
high oxygen content of biodiesel, it is generally accepted that biodiesels are about
10% less energy dense as compared with petroleum diesel. The heating value of
biodiesel is related to its fatty acid profile. Heating value increases with increasing
carbon number in fuel molecules due to mass fraction decreases (Demirbas 2003).
Studies have also found that unsaturated esters have lower mass energy content
(MJ/kg). Demirbas (Demirbas 2008b) studied the correlation between viscosity and
higher heating value (HHV) by performing a linear least square regression analysis.
This study found that there is a high correlation between the heating value and the
viscosity of vegetable oils and their methyl esters and that the heating value of
vegetables oils and biodiesels increases with viscosity.
2.1.7.5 Flash point
Flash point is often used as a descriptive characteristic of liquid fuel and is defined as
the lowest temperature at which the fuel will start to vaporise to form an ignitable
mixture when it comes to contact with air (Ali, Hanna and Cuppett 1995a). Hence,
flash point is an important parameter for assessing fire hazards during fuel transport
and storage. This is reflected by the respective limits within Australian standards
(≥120 °C) and the European fossil diesel standard, EN 590 (>55 °C). However, the
flash point of biodiesel is approximately double that of petroleum diesel, which
makes biodiesel a more acceptable engine fuel in relation to concerns about safety. It
is also an important parameter in engine combustion performance. Canakci and Sanli
[93] found that with a high flash point, NOx emission decreased due to low
combustion pressure and temperature. Moreover, high flash point is also important in
niche applications such as underground mining. On the other hand, studies show that
a high flash point can cause cold engine startup problems, misfiring and ignition
delay, which increases carbon deposition in the combustion chamber (Ali, Hanna and
Cuppett 1995a).
38 Chapter 2: Literature Review
Biodiesels from animal fat generally have higher flash points than those from
vegetable oils. This is the result of the highly saturated fatty acid compounds in
biodiesels from animal fats increasing the flash point temperature. Alcohol residue in
biodiesel significantly decreases flash point (Canakci and Sanli 2008).
2.1.7.6 Oxidation stability
Oxidation stability is an important fuel property which reflects the resistance to
oxidation during long-term storage. Usually biodiesels are more sensitive to
oxidative degradation than petroleum diesel due to their chemical composition. Fuel
quality declines due to gum formation during the oxidation process. This gum does
not combust completely resulting in poor combustion, carbon deposits in the
combustion chamber and lubrication oil thickening (Ma and Hanna 1999). Monyem
et al. (Monyem and H Van Gerpen 2001) observed that oxidised biodiesel starts to
burn earlier than unoxidised, increasing NOx emissions due to the associated increase
in viscosity and cetane number.
The chemical structure of biodiesel is an important factor in the oxidation reaction.
Oxidation is influenced by the presence of double bonds in the chains, that is,
feedstocks rich in polyunsaturated fatty acids are much more susceptible to oxidation
than the feedstocks rich in saturated or monounsaturated fatty acids (Graboski and
McCormick 1998). However, our understanding of oxidation is complicated by the
fact that fatty acids usually occur in complex mixtures, with minor components in
these mixtures catalysing or inhibiting oxidation. In addition, the rates of the
oxidation of different unsaturated fatty acids or esters can vary considerably. The
other factors that affect the oxidation stability of biodiesel include double bond
configuration, temperature, air, light and storage tank materials (Balat and Balat
2008).
2.1.7.7 Cold temperature properties
One of the major problems associated with the use of biodiesel in countries with a
cold climate include the poor cold flow properties when compared with petroleum
Chapter 2: Literature Review 39
diesel fuels. The parameters generally used to determine cold flow properties are
cloud point (CP), pour point (PP) and cold-filter plugging point (CFPP). CP is the
temperature at which a material becomes cloudy due to the formation of crystals and
solidification of saturates, with PP being the lowest temperature at which the fuel can
be pumped (Lee, Johnson and Hammond 1995). In general, the CP occurs at a higher
temperature than the PP, with the CFPP defined as the lowest temperature at which a
fuel portion will pass through a standardised filtering device in a specified time.
Studies have found that biodiesel from all feedstock has a relatively high CP, PP and
CFPP when compared with petroleum diesel (Durrett, Benning and Ohlrogge 2008).
Moreover, biodiesel contains relatively few components compared to petroleum
diesel, and each component has its own solidification temperature. Therefore, the
solidification of biodiesel is more rapid and difficult to control as one or two
components can tend to dominate. At low temperatures, solids and crystals rapidly
grow and agglomerate, clogging fuel lines and filters and causing major operability
problems.
The fatty acid composition of biodiesel greatly influences its cold-flow properties.
Biodiesels made from feedstocks containing higher concentrations of high-melting
point saturated long-chain fatty acids tends to have relatively poor cold flow
properties (Dunn and Moser 2010). Studies found that the freezing point of a
biodiesel fuel increases with increasing numbers of carbon atoms in the carbon chain,
and decreases with increasing double bonds (Graboski and McCormick 1998;
Knothe 2005; Demirbas 2003). Therefore, saturated components have poor cold flow
properties over unsaturated components. Moreover, the double bond position also
affects the cold flow properties. Rodrigues (Rodrigues Jr et al. 2006) reports that a
double bond position near the end of the carbon chain results in poor cold flow
properties as compared with a double bond found in the middle of the molecules.
While conducting studies on the effects of chemical structure on the crystallisation
temperature using a series of branched alcohol esters, Nascimento et al. (Nascimento
et al. 2005) observed in the studies that branching in the carbon chain reduces the
crystallisation of esters.
40 Chapter 2: Literature Review
2.1.7.8 Lubricity
Lubricity is defined as the ability of fuel to provide hydrodynamic and/or boundary
lubrication to prevent wear between engine moving parts. It is an important
parameter for diesel engine operation, because poor lubrication leads to the failure of
engine parts, such as fuel injectors and pumps as they are directly lubricated by the
fuel itself (Lacey and Lestz 1992). Increasingly strict regulations on the sulphur
content of commercial petroleum fuel have resulted in a decrease in fuel lubricity
over time. Therefore, the issue of lubricity in fuel is becoming increasingly important
with respect to diesel engine operation. Biodiesels typically have superior lubrication
properties when compared with petroleum diesel (Hu et al. 2005). Studies show that
biodiesel derived from vegetable oils can significantly increase diesel fuel lubricity at
blend concentrations of less than 1%. Therefore, biodiesel can be used as an additive
to improve the lubricity of petroleum fuel (Goodrum and Geller 2005; Anastopoulos
et al. 2005; Anastopoulos et al. 2001; Van Gerpen, Soylu and Tat 1999).
Knothe and Steidley (Knothe and Steidley 2005) examined the lubricity of biodiesel
and petroleum diesel components. They found better lubricity in fatty acid
compounds than hydrocarbons in petroleum diesel. This study also reported that pure
free fatty acids, mono-acylglycerols, and glycerol possess better lubricity than pure
esters. The main components responsible for biodiesel lubricity are FAMEs,
hydroxyl groups and mono-acylglycerols followed by free fatty acids (FFAs) and di-
acylglycerols, whereas triglycerols do not have any significant effect on the lubricity
of biodiesel (Geller and Goodrum 2004; Hu et al. 2005) On the other hand, the
structures of fatty acids have an impact on biodiesel lubricity. Increasing saturation
leads to a stronger lubrication layer as molecules can align themselves more easily in
straight chains and when they are packed closely on the surface (Wadumesthrige et
al. 2009). However different results have been reported by Knothe (Knothe 2005),
Anastopoulos et al. (Anastopoulos et al. 2001) and Bhuyan et al. (Bhuyan et al.
2006) who found that lubricity increases with the increase in length of fatty acid
chains, while Geller et al. (Geller and Goodrum 2004) showed that there is no
consistent trend relating chain length to lubricity enhancement.
Chapter 2: Literature Review 41
2.1.7.9 Iodine value
Iodine value (IV) is a common method used to determine the degree of unsaturation
in a mixture of fatty materials regardless of their relative share of mono-, di- and
polyunsaturated compounds (Knothe and Dunn 2003). When iodine is added to the
fat or oil, the amount of iodine in grams absorbed per 100 mL of oil is reported as the
IV. As it is a measure of total unsaturation, several studies found a linear correlation
between the degree of unsaturation (DU) and the IV: the more unsaturation in the oil,
the higher the iodine value (Lin, Lin and Hung 2006; Kyriakidis and Katsiloulis
2000). Methyl esters show an almost identical IV to that of corresponding vegetable
oils or animal fats (Knothe 2005).
Iodine value is an important parameter in regard to fuel quality because higher IV
biodiesel leads to a higher rate of polymerisation of glyceride, which increases
rapidly with temperature. This results in increasing fuel viscosity, adversely affecting
the fuel’s ease of flow, and causing the formation of engine deposits, thus adversely
affecting fuel injector spray patterns. This will ultimately lead to poor combustion,
high emissions and, consequently, engine failure (Knothe 2005).
2.2 BIODIESEL AS A DIESEL ENGINE FUEL
Diesel engines produce mechanical power from conversion of the chemical energy
contained in the fuel. Energy is released by the combustion and oxidisation of the
fuel inside the engine. The fuel-air mixture prior to combustion and the products
from combustion are the working fluids. The boundary work which provides the
desired power output occurs directly between these working fluids and the
mechanical components of the engine (Heywood 1988).
Since the advent of the diesel-powered engine, compression ignition engine
technology has been under continuous development. However, the basic components
of the engine (Figure 2-7) have been unchanged, with the main difference between a
42 Chapter 2: Literature Review
modern day engine and its predecessor being its combustion performance (Ferguson
and Kirkpatrick).
Biodiesel can be used in modern diesel engines in its pure form (B100), or blended
with petroleum diesel in any ratio (Lebedevas and Vaicekauskas 2006). There is an
increasing body of literature reporting on research into diesel engine performance
and engine emissions when fuelled with biodiesel. Some of these studies are
summarised in Table 2-6.
Figure 2-7: Schematic diagram of a typical diesel engine fuel system (Haseeb et al. 2011b)
2.2.1 Engine performance
Diesel engine performance parameters evaluated with biodiesel fuels in literature
typically includes engine power, torque, brake specific fuel consumption (BSFC),
thermal efficiency, and exhaust gas temperature (Table 2-5). While illustrating the
effect of biodiesel on engine power and/or torque, it is commonly argued that
biodiesel drops engine power and torque. This is mainly due to the lower heating
value of biodiesel compared with petroleum diesel. Utlu and Kocak (Utlu and Koçak
2008) ran a four-cylinder diesel engine with waste frying oil methyl ester (WFOME),
varying the engine speed from 1750 to 4400 rpm. They found on average a 4.5% and
Chapter 2: Literature Review 43
4.3% reduction in power and torque respectively. Similar results have been reported
in many other studies, with some fluctuations in the reduction percentage. Studies
found that the loss of power was 7.14% for biodiesel when compared to diesel on a
three-cylinder, naturally aspirated (NA) submarine diesel engine at full load, yet the
loss of heating value of biodiesel was about 13.5% when compared to diesel
(Karabektas 2009; Hansen, Gratton and Yuan 2006; Murillo et al. 2007). Hansen et
al. (Hansen, Gratton and Yuan 2006) observed that the brake torque loss was 9.1% in
biodiesel at 1900 rpm as the results of variation in heating value (13.3%), density and
viscosity. Findings from these studies confirm that the lower heating value of
biodiesel is not the only factor which influences engine power and torque. Other
biodiesel fuel properties including viscosity, density and lubricity have significant
effects on engine output power and torque.
Table 2-6: Performance and emission of diesel engines with biodiesel
Fuel type Engine Test condition Increase/decrease vs. diesel
References Power Torque BSFC BTE Texh CO CO2 NOx PM HC
Soybean 1C 1400–2000 rpm ↓ ↓ ↑ ↓ ↓ ↓ (Qi et al. 2010)
Cottonseed 1C 850 rpm ↑ ↓ ↓ ↑ ↓ (Nabi, Rahman and Akhter
2009)
Sunflower 1C 1000–3000 rpm ↓ ↓ ↑ ↓ ↓ ↓ ↓ (Ilkılıç et al. 2011)
Waste cooking 4C 1750–4400 rpm ↓ ↓ ↑ ↓ ↓ ↓ ↓ (Utlu and Koçak 2008)
Cottonseed 4C 1500 rpm ↓ ↓ ↑ ↓ ↓ ↑ ↓ (Kumar 2009a)
Sunflower oil 4C 1100–2800 rpm ↓ ↓ ↑ ↓ ↑ ↓ (Haşimoğlu et al. 2008)
Soybean 4C 1400 rpm ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↑ ↓ ↓ (Canakci 2007)
Waste cooking 4C 800–1400 rpm ↓ ↓ ↑ ↑ ↓ ↓ ↑ (Lin and Li 2009a)
Waste cooking 4C 1400–200 rpm ↓ ↓ ↑ ↓ ↓ ↓ ↑ (Lin and Li 2009a)
Mahua 1C 1600 rpm ↓ ↓ ↑ ↓ ↑ ↓ ↑ (Raheman and Ghadge 2007)
Tobacco seed 4C 1500–300 rpm ↑ ↑ ↓ ↑ ↓ ↓ ↑ (Usta 2005)
Rapeseed oil 4C 1200–2400 rpm ↑ ↑ ↓ ↓ ↑ (Karabektas 2009)
Cottonseed 1C 1200–2500 rpm ↑ ↑ ↓ ↑ ↓ ↓ ↑ (Aydin and Bayindir 2010)
For instance, higher viscosity of biodiesel improves air-fuel mixing by enhancing
spray penetration, and thus recovery in power and torque when compared to diesel
(Öner and Altun 2009; Monyem, Van Gerpen and Canakcl 2001). Higher viscosity
can also reduce engine power by decreasing combustion efficiency due to poor fuel
injection atomisation (Aydin and Bayindir 2010). On the other hand, the higher
density of biodiesel improves engine power and torque. Moreover the high lubricity
44 Chapter 2: Literature Review
in biodiesel may result in reduced friction power loss, and this will subsequently
recover engine output power and torque (Ramadhas, Muraleedharan and Jayaraj
2005). Therefore, it is not surprising that some studies have reported increased power
and torque from engines when running on biodiesel.
For an example, Song and Zhang (Song and Zhang 2008) showed power and torque
increased with an increase in biodiesel percentage in blends while running an engine
with soybean oil methyl ester. Usta (Usta 2005) also found similar results when
using tobacco seed oil in a four-cylinder turbo-charged diesel engine. Furthermore,
negligible variation in engine power and torque between biodiesel and petroleum
diesel has also been found (Luque et al. 2008; Pal et al. 2010). More interesting
results have been reported by Haşimoğin et al. (Haşimoğlu et al. 2008) while using
waste cooking oil biodiesel in a four-cylinder turbo-charged diesel engine operating
between 1100 and 2800 rpm. This study found lower engine torque and power at
lower engine speeds (1100 to 1600 rpm) while power and torque increased at
medium and high engine speeds. However, Carraretto et al. (Carraretto et al. 2004)
has overcome the power loss of biodiesel engine by optimising biodiesel combustion
through reducing the injection advance. It is therefore evident that power and torque
developed in biodiesel engines is not only dependent on feedstock and fuel
properties, but also on the engine type and operating conditions, such as engine
speed, load, injection timing and injection pressure. Similar correlations have been
found in the literature for other performance parameters such as brake specific fuel
consumption, thermal efficiency, exhaust gas temperature and combustion
characteristics (Lin et al. 2011; Raheman and Ghadge 2007; Aydin and Bayindir
2010; Murillo et al. 2007).
2.2.2 Exhaust emissions
Combustion chemistry in internal combustion engines (ICE) is very complex and
depends on fuel types and operating conditions. In the combustion chamber,
hydrocarbon reactions are generally grouped into three distinct steps. The first step is
the breakdown of hydrocarbons; the second step is the oxidation of hydrocarbons and
hydrogen; the third step is the oxidation of combustion reaction products. The
Chapter 2: Literature Review 45
exhaust gas from diesel engines contains many components including carbon dioxide
(CO2), carbon monoxide (CO), hydrogen (H2), oxygen (O2), sulphur oxides (SOx),
unburned hydrocarbons (HC), particular matter (PM), and nitrogen oxides (NOx).
These pollutants have various potential adverse health and environmental effects.
Numerous studies have been conducted to investigate the effect of biodiesel on
exhaust emissions in diesel engine applications. The emission parameters
investigated include carbon dioxide (CO2), carbon monoxide (CO), hydrocarbon
(HC), nitrogen oxides (NOx), sulphur oxides (SOx) smoke, particulate matter (PM).
Table 5 shows that most of the studies found a sharp reduction in all exhaust
emissions when biodiesel was used was compared with petroleum diesel fuel (except
NOx). However, a reduction in NOx in biodiesel use has also reported in some other
literature.
In general, biodiesel contains about 10% oxygen by mass, while diesel has little to no
oxygen. Biodiesel fuels result in more complete combustion and thereby reduces
exhaust emission, and various researchers have postulated reasons for this outcome.
The percentage change in emissions varies amongst these studies. The variety of
results reported can be attributed to variations in the fuel properties and chemical
structure of the biodiesels used, varying feedstocks and due to the variety of engines
used in tests. For example, Lin et al. (Lin and Li 2009a) conducted an experiment
with biodiesel from eight different feedstocks and found a significant reduction in
PM emissions (50%–73%). While conducting experiments with coconut, jatropha
and rapeseed oil biodiesel, Lance et al. (Lance et al. 2009) showed that rapeseed oil
biodiesel tended to give amongst the highest NOx emissions. Similarly, variations in
emissions from biodiesels using different feedstocks have been reported in many
other recent studies (Wu et al. 2009; Sahoo et al. 2009; Ozsezen et al. 2009;
Banapurmath, Tewari and Hosmath 2008). Monyem and Gerpen (Monyem and H
Van Gerpen 2001) found that oxidised biodiesel can significantly reduce emissions
while investigating the effect of biodiesel oxidation on diesel engine emissions. This
study found that oxidised biodiesel resulted in approximately 15% less CO emissions
and 21% less HC emission when compared with unoxidised biodiesel.
46 Chapter 2: Literature Review
The vast majority of the literature reports that NOx emissions are the only parameter
that increases while operating diesel engines on biodiesel as compared to petroleum
diesel. Therefore, NOx emission may be the single most critical parameter for
biodiesel application. Many studies have suggested that properties of biodiesel such
as cetane number, oxygen content, biodiesel feedstock, advance in fuel injection,
engine type and operating conditions have an important effect on the formation of
NOx (Xue, Grift and Hansen 2011; Fazal, Haseeb and Masjuki 2011). It is commonly
argued that high cetane numbers improve combustion, therefore the temperature in
the combustion chamber is expected to be higher which leads to the formation of
more NOx emissions in higher oxygen content fuel. Another common argument is
that a high cetane number reduces ignition delay which can cause higher NOx
emission. However, some authors oppose this argument. NOx usually forms in the
combustion phase and a high cetane number not only reduces ignition delay, but also
leads to lowering of the premixed combustion phase, which eventually reduces the
formation of NOx. Therefore, it is not surprising that some literature reports show a
reduction of NOx emission with biodiesel (Qi et al. 2010; Utlu and Koçak 2008;
Karabektas 2009). Moreover, the different chemical structure of biodiesels influences
the formation of NOx emissions and most of the authors agreed that shorter chain
lengths and more saturated biodiesels were preferable to reduce NOx emission.
Knothe et al. (Knothe and Steidley 2005) tested NOx in a six-cylinder diesel engine
with conventional diesel fuel and three different fatty acid methyl esters including
oleic (C18:1), palmitic (C16:0) and lauric (C12:0). They found a 4% and 5%
reduction in NOx emissions for the saturated palmitic and lauric esters, respectively,
and a 6% increase for the oleic ester. While blending short chain methyl esters such
as caprilic (C8:0) and capric (C10:0) with soybean oil biodiesel, Chapman and
Boehman, (Chapman 2006) also found a significant reduction in NOx emission.
Particulate matter is another important factor which needs to be considered while
using biodiesel as an engine fuel. Particulate matter emitted by petroleum diesel
engines consists of black carbon (soot), hydrocarbons, sulphates and metallic ashes
(Majewski 2002). Most studies found in the literature indicate that biodiesel reduces
diesel engine particulate emissions on a mass basis. However, adverse health effects
Chapter 2: Literature Review 47
from exposure to particulates may increase with a decreasing in particle size, even
though the particles are composed of toxicologically inert materials. For example,
fine particles with aerodynamic diameters lower than 2.5 µm (PM2.5) appear to have
considerably enhanced toxicity per unit mass compared to coarse particles with
aerodynamic diameters lower than 10 µm (PM10). Particles deposit in different parts
of the lung according to their aerodynamic diameter, and smaller particles are able to
penetrate deeper into the human lung. Furthermore, smaller particles tend to stay in
the atmosphere for longer periods of time, which means that there is a higher
probability that they will be inhaled and lead to respiratory diseases, inflammation
and damage to the lungs (Garshick et al. 2004). By considering these adverse effects
of ultrafine particles, an emission standard based on a solid particle number has
already been implemented by the European Union (EU) for its member states.
Particle number and size distribution is therefore a more appropriate parameter of
DPM when placing regulatory controls on the use of fuels, rather than relying on PM
mass alone (Surawski et al. 2009).
2.3 ARTIFICIAL NEURAL NETWORKS
The foundation of artificial neural networks (ANN) in a scientific sense begins with a
biological neuron as shown in Figure 2-8. In the brain, there is a flow of coded
information (using electrochemical media, the so-called neurotransmitters) from the
synapses towards the axon. The axon of each neuron transmits information to a
number of other neurons. Groups of neurons are organised into sub-systems and the
integration of these sub-systems forms the brain. On the other hand, an ANN is
composed of a large number of simple processing units called neurons which are
fully connected to each other through adoptable synaptic weight (Figure 2-9). This
resembles a brain in two aspects. Knowledge can be acquired through training and
knowledge can be stored. In the training process, weights are adjusted to minimise
the error between actual output and desired output.
The most important feature of artificial neural networks is their ability to solve
problems through learning by example, rather than by fully understanding the
48 Chapter 2: Literature Review
detailed characteristics of the systems. This feature makes it very useful because it
works like a “black box” model, and does not require detail or complete information
about the problem, and can be utilised when all that is available are sets of data
inputs and outputs of the system. It has a natural propensity to store experiential
knowledge and to make it available for use (Figure 2-10). Therefore, this nonlinear
computer algorithm can model large and complex systems with many interrelated
parameters.
Figure 2-8: Biological neuron (Kalogirou 2003)
Chapter 2: Literature Review 49
Figure 2-9: Multi-layer ANN model
Figure 2-10: Working principle of ANN
Since the development of high speed digital computers, the application of the ANN
approach has progressed at a very rapid rate. In recent years, this method has been
applied to various disciplines including automotive engineering, and in the
forecasting of fuel properties and engine thermal characteristics for various working
50 Chapter 2: Literature Review
conditions (Ramadhas et al. 2006b). Prediction accuracy of the ANN approach was
found to be superior when compared with other linear and non-linear statistical
techniques.
Balabin et al. (Balabin, Lomakina and Safieva 2011) compared the prediction
performance of artificial neural networks (ANN), multiple linear regression (MLR),
principal component regression (PCR), polynomial and Spline-PLS versions, and
partial least squares regression (PLS) for prediction of biodiesel properties from near
infrared (NIR) spectra. The model was created for four biodiesel properties density
(at 15 °C), kinematic viscosity (at 40 °C), water (H2O) content and methanol content.
This study reported the lowest root mean square errors of prediction (RMSEP) for
ANN when compared to other techniques as shown in Figure 2-11. Agarwal et al.
(Agarwal, Singh and Chaurasia 2010) compared the linear regression and ANN
techniques in predicting biodiesel properties. Results of this study indicated that
ANNs were able to predict the properties of biodiesel better than a linear regression
model. Similarly Cheenkachorn (Cheenkachorn 2004) predicted biodiesel properties
such as viscosity, high-heating value, and cetane number using the fatty acid
compositions of various vegetable oils by statistical methods and ANN. It was
observed that ANN was able to predict more accurately than statistical methods.
Figure 2-11: Comparison of the performance of between ANN and various linear and
non-linear prediction techniques (Balabin, Lomakina and Safieva 2011).
Chapter 2: Literature Review 51
On the other hand, ANN is not without having limitations. The main disadvantage of
an ANN model is due to the “black box” approach; it is difficult to gain insight in to
a problem without extra effort. Statistical techniques allow the user to determine the
most significant parameters among the important variables, hence, eliminating the
variables that do not fit the model which is not standby available with ANN.
However, this limitation can be overcome where necessary by combining the ANN
with multivariate data analysis, such as PCA or ANOVA. Moreover the requirements
of computational resources and standard software for ANN modeling may also be
considered as drawbacks over statistical techniques (Agarwal, Singh and Chaurasia
2010).
2.3.1 ANN in predicting engine emission and performance
One of the most important factors that affect the viability of an automobile fuel is its
suitability in terms of engine performance and exhaust emissions as discussed in the
previous section. However, conducting actual experiments with automobile engines
not only requires considerable amounts of fuel, heavy equipment and skilled
personnel, but it is also very time consuming and costly. Therefore, the use of ANN
modeling techniques in predicting performance parameters of internal combustion
engines has gained in popularity over the last few decades. Combustion related
performance using various types of fuels including diesel, gasoline, natural gas,
ethanol and biodiesel have successfully been modeled using ANN. Cay et al. (Cay et
al. 2012) developed an ANN model to predict the brake specific fuel consumption,
effective power and average effective pressure and exhaust gas temperature of the
methanol engine. A four-cylinder, four-stroke test engine was operated at different
engine speeds and torques to obtain model training and testing data. After training,
this study found an ANN prediction accuracy with R2 values close to 1 for both
training and testing data. RMS values less than 0.015 and mean errors less than 3.8%
for the testing data were reported. This study concluded that the ANN model is a
powerful technique for predicting performance parameters of internal combustion
engines.
52 Chapter 2: Literature Review
Similarly, while predicting engine performance emissions with ANN, Sayin et al.
(Sayin et al. 2007) found prediction performance with correlation coefficients in the
range of 0.983–0.996, mean relative errors in the range of 1.41%–6.66%, and very
small root mean square (RMS) values, and concluded that ANN was an alternative to
classical modelling techniques. Arcaklioğlu and Çelıkten (Arcaklioğlu and Çelıkten
2005) conducted experiments with petroleum diesel in a turbo-charged four-cylinder
diesel engine. Using experimental data, the ANN model was developed to predict
engine torque, power, brake mean effective pressure, specific fuel consumption, fuel
flow, and exhaust emissions such as SO2, CO2, NOx and smoke level based on
injection pressure, engine speed and throttle position. The study found the precise
prediction ability of the ANN in diesel engine performance and emission parameters
provided a good correlation between experiment and predicted values. The overall
mean squire errors in their study were less than 0.03% and R2 values where close to
1.
Yap and Karri (Yap and Karri 2012) conducted experiments on a single-cylinder,
spark ignition engine using various engine speeds and throttle positions. They have
developed an ANN model for the prediction of engine power, CO, CO2, HC and O2
emissions. Parlak et al. (Parlak et al. 2006) also successfully used ANN to predict
diesel engine brake specific fuel consumption and NOx emission for a Ricardo E6
type, single-cylinder diesel engine considering engine speed, break mean effective
pressure and fuel injection timing as input variables.
ANN modelling techniques have also been utilised for predicting engine
performance and emissions based on the physical properties of various fuels.
Canakci et al. (Canakci et al. 2009) used ANN model in predicting fuel flow rates,
maximum injection pressure, thermal efficiency, load, maximum cylinder pressure
and exhaust emission (CO, NOx, HC) of a diesel engine based on fuel density, kinetic
viscosity and lower heating value. Similarly, Karonis et al. (Karonis et al. 2003) were
successful in predicting exhaust emissions (CO, NOx, HC and PM) of a single-
cylinder diesel engine using fuel cetane number, density and kinetic viscosity.
Chapter 2: Literature Review 53
ANN has also been used to predict diesel engine emissions from in-cylinder pressure
along with engine operating parameters (Kesgin 2004; Manjunatha, Narayana and
Reddy 2010) investigated the effectiveness of various biodiesel fuel properties and
engine operating conditions on diesel engine combustion towards the formation of
exhaust emissions using ANN. They conducted experiments on a single-cylinder
direct injection (DI) diesel engine using blends of biodiesel from pongamia, jatropha
and neem oils. This study found good predictability of ANN modelling techniques in
predicting brake power, brake thermal efficiency, brake specific fuel consumption,
volumetric efficiency, and exhaust gas temperature regulated exhaust emissions (CO,
HC, NOx) based on diesel-biodiesel blend percentage, fuel properties and various
engine operating conditions. Similarly, different types of engines have been
successfully modeled using ANN techniques in a wide range of operating conditions,
performance parameters and fuels which are summarised in Table 2-7. All of these
studies demonstrated the suitability of ANN modeling techniques in internal
combustion engine application.
2.3.2 ANN in predicting fuel properties
The properties of fuel need to be estimated before their application to particular
combustion systems as these will significantly influence the end use performance of
such systems, as shown in previous sections. Modern official standards list more than
20 parameters that must be determined to certify any fuel’s quality before its use as
automobile engine fuel. However, testing of these properties requires considerable
amounts of fuel sample, standardised testing equipment, expert technicians and also
carries a significant cost (Balabin, Lomakina and Safieva 2011). Therefore it is a
worthwhile option to consider a prediction model to estimate the properties of any
new fuel prior to beginning large-scale production. As a consequence, a number of
models have been developed to predict the important fuel properties of diesel and
biodiesel using conventional linear regression and ANN techniques. Most of these
are based on fuel types, diesel-biodiesel blend ratios, chemical composition,
production methods etc., Agarwal et al. (Agarwal, Singh and Chaurasia 2010)
developed linear regression and an ANN model to predict several fuel properties of
biodiesel including heating value, density, viscosity, pour point, flash point, iodine
value and saponification value based on fatty acid composition. Experiments were
54 Chapter 2: Literature Review
conducted using biodiesel produced from various edible and non-edible vegetable
oils. This study found a good correlation between the properties of biodiesel and its
chemical composition, with the ANN demonstrating a higher prediction ability than
linear regression techniques in predicting all the fuel properties.
A similar study has been conducted by Cheenkachorn focused on determining
biodiesel fuel properties based on fatty acid profile only (Cheenkachorn 2004).
However, these studies did not consider other chemical compositions contained in
biodiesel through the model input, including the amount of free glycerol, free fatty
acid, methanol and impurities which may have an impact on the physical properties
of the biodiesel. Kumar and Bansal (Kumar and Bansal 2010) compared the
applicability of the traditional statistical technique of linear regression (principle of
least squares) and ANN techniques in estimating the flash point, fire point, density
and viscosity of diesel and biodiesel mixtures. They have optimised the network with
three training algorithms, along with ten different sets of weight and biases. Results
of this study show that neural network is the better choice over principle of least
squares to predict the fuel properties of various mixtures of diesel and biodiesel.
However, the performance of a neural network can further be improved by adjusting
the other training parameters like goal, epochs, learning rate, magnitude of the
gradient, etc.
Chapter 2: Literature Review 55
Table 2-7: ANN used in automobile engine application
References Model Input Model Target Engine Prediction accuracy Fuel used
(Arcaklioğlu and Çelıkten 2005) IP, RPM, TH T, P, BMEP, BSFC, FFR, SO2, CO2,
NOx, S
4S, 4C, TC R2: 0.9999; MSE: 8.5% Diesel
(Yap and Karri 2012) RPM, AFR, TH P, CO, CO2, HC, O2 2S, 1C, SI MSE: 2.27%–4.74% Gasoline
(Parlak et al. 2006) RPM, BMEP, IT BSFC, Texh 4S; 1C, NS, CI MSE: 1.93%–2.36% Diesel
(Choi and Chen 2005) CR, FFR, AFR, IT, EGR Start-of-combustion 4S, 1C, HCCI - Methane/n-heptane
(Karonis et al. 2003) CN, ρ, ν, Distillation curve CO, HC, NOx, PM 4S, 1C, CI R2: 0.937–0.99 Diesel
(Çelik and Arcaklioğlu 2005) RPM, P, Cooling water temperature BSFC, Texh, AFR 4S, 8C, CI R2: 0.99, MRE: 5.5% Diesel
(Deh Kiani et al. 2010) RPM, L, FBR P, T, CO, CO2, NOx, HC 4S, 4C, SI R2: 0.71–0.91 Gasoline, Ethanol
(Renald and Somasundaram 2012) L, FDR, Cylinder head geometry Texh, CO, CO2, O2, HC, NOx, ET 1C, 4S, SI - Gasolin, LPG
(Yusaf et al. 2010) RPM, FBR P, T, BTE, BSFC, Texh, NOx, CO, CO2 4S, 1C, CI R2: 0.95707–0.9934 Diesel, CNG
(Yusaf, Yousif and Elawad 2011) RPM, FBR P, T, BTE, BSFC, Texh, NOx, CO, CO2 4S, 1C, CI MSE: 0.0004 Diesel, CPO
(Obodeh and Ajuwa 2009) L, RPM NOx 4S, 4C, CI MRE: 0.68%–3.34%. Diesel
(Srinivasa Pai and Shrinivasa Rao 2011) L, IT, CR, FBR BTE, BSEC, Texh, NOx, HC 4S, 1C, CI MRE: 1.778%–5.889% Diesel, WCO biodiesel
(Tasdemir et al. 2011) RPM, Intake valve opening advance P, T, BSFC, HC 4S, 1C, SI - Gasoline
(Sayin et al. 2007) T, RPM, HV, Air inlet temperature BSFC, BTE, Texh, CO, HC 4S, 4C, SI R2: 0.983–0.996 Gasoline
(Kesgin 2004) ECP, CT, FAR, RPM, ES, CB, VT P, BTE, NOx, Heat Transfer TC, SI - Natural gas
(Canakci et al. 2009) RPM, HV, ρ, ν, Environmental conditions FFR, BTE, CO, NOx, HC, L, IP, ECP 4S, 4C, NA, CI R2: 0.99 WCO, Diesel
(Shanmugam et al. 2011) L, FBR BTE, CO, HC, CO2, NOx, S 4S, 1C, NA, CI R2: 0.975–0.999 Diesel, Bioethanol,
Cottonseed biodiesel
(Sharon et al. 2012) P, FBR BSFC, BTE, NOx, HC, CO, S 4S, 1C, NS R 2: 0.9989–0.999 Diesel, WCO biodiesel
(Manjunatha, Narayana and Reddy 2010) L, HV, CN, ρ, FBR P, BTE, BSFC, Texh, CO, HC, NOx 4S, 1C, CI R2: 0.95–0.99 Diesel, pongamia, jatropha,
neem oils biodiesel
Chapter 2: Literature Review 56
Table 2-8: ANN in predicting fuel properties
References Use of ANN model (Boudy and Seers 2009) Density prediction in different temperatures for palm oil biodiesel (Ramadhas et al. 2006b) Cetane number prediction based on the fatty acid profile of biodiesel (Agarwal, Singh and Chaurasia 2010)
Density, kinetic viscosity, water and methanol content prediction for various biodiesels
(Kumar, Bansal and Jha 2007)
Flash point, fire point, density and viscosity prediction based on diesel-biodiesel blend ratio
(Liu et al. 2007) Density, flash point, freezing point, aniline point and net heat of combustion prediction for various jet fuels based on their chemical composition
(Korres et al. 2002) Lubricity prediction from physical properties of diesel (Marinović et al. 2012) Prediction of diesel cold temperature properties based on density,
kinetic viscosity, conductivity, sulphur content and 90% distillation point
(Pasadakis, Gaganis and Foteinopoulos 2006)
Octane number prediction from chemical composition of gasoline
(Pasadakis, Sourligas and Foteinopoulos 2006)
Cold temperature properties and distillation curve prediction from the chemical composition of diesel
(Wu et al. 2006) Prediction of cold filter plugging point of diesel from physical properties
(Yang et al. 2002) Cetane number and density prediction using chemical composition of diesel
(Cheenkachorn 2004) Viscosity, cetane number and heat of combustion prediction from fatty acid composition of various biodiesel feedstocks
Satyanarayana and Muraleedharan (Satyanarayana and Muraleedharan 2010)
developed the ANN model to analyse the relation of esterification methods with fuel
properties for biodiesel produced from rubber seed. They used transesterification
reaction parameters such as the methanol-oil ratio, catalyst concentration, reaction
time, and reaction temperature in the input layer and acid value of biodiesel in output
layer while constricting the ANN model. Although a good prediction was obtained,
this study did not consider the initial acid value of the vegetable oil, which may have
an impact on the final acid value of the biodiesel. This issue has been addressed by
Rajendra et al. (Rajendra, Jena and Raheman 2009) while using ANN techniques to
predict the acid value of sunflower oil biodiesel, including its initial acid value as
well as transesterification reaction parameters. Liu et al. (Liu et al. 2007) compared
the reputability of standard fuel property testing methods with ANN while
developing a model to predict density, flash point, freezing point, aniline point and
net heat of combustion of 80 different jet fuels. This study found that the
Chapter 2: Literature Review 57
repeatability of neural network models for measuring density and flash point were
lower than the ASTM test methods. However, for the freezing point, aniline point
and net heat of combustion, the repeatability of the ANN methods are equal to the
ASTM methods. Therefore it can be said that not only the prediction accuracy but
also the ANN approaches are comparable to the repeatability values of the standard
ASTM methods, which are used for the experimental determination of the properties.
Similar conclusions have been made by Pasadakis et al. (Pasadakis, Gaganis and
Foteinopoulos 2006; Pasadakis, Sourligas and Foteinopoulos 2006) while predicting
pour point (CP), cloud point (CP) of diesel and octane number of gasoline based on
the chemical composition of the respective fuels. Several other studies which were
also successful in utilising ANN to estimate the properties of various fuels, which are
summarised in Table 2-8. The success of those studies proved that ANN has the
ability to accurately estimate the fuel properties instead of having costly and time
consuming experimental measurements.
2.4 ANN MODELING OF SECOND-GENERATION BIODIESEL
Although numerous feedstocks including oilseed crops and algae species have been
identified as being suitable for producing second-generation biodiesel, these types of
biodiesel have not yet been established, due to the unavailability of feedstock supply,
high production costs and a lack of knowledge about the fuel’s quality. Moreover, by
producing fuels from new feedstock, optimising production procesess, ensuring fuel
quality through measuring a number of physical and chemical properties, and
evaluating the end-use performance in automobile engines is costly, time-consuming
and requires a wide variety of specialised equipment and skilled workers. These
concerns have restricted the progress of second-generation biodiesel technology,
making it still unacceptable to both automobile engine manufacturers and customers,
which is yet to begin industrial-scale production. In order to address this issue, the
ANN modelling technique could be a very useful tool in predicting fuel quality and
engine combustion-related parameters when considering the chemical composition of
new biodiesel feedstocks. It would require laboratory scale biodiesel production and
basic chemical testing equipment, which will significantly reduce the research cost,
58 Chapter 2: Literature Review
and hence accelerate the investigation of future generation biodiesel. However,
researchers should move to develop a universal ANN prediction model for second-
generation biodiesel, and this will enable the instigation of a wide range of biodiesel
feedstock and automobile engine systems. While training the network, they need to
consider all possible parameters in feedstock that affect the production process, the
quality and the combustion performance of biodiesel in automobile engine
applications. A two-stage artificial neural network (ANN) prediction model can be
proposed for this purpose. At the 1st stage of the model, chemical composition of
biodiesel in terms of fatty acid profile can be used as input parameters and fuel
properties can be used as output or target variable at. In the 2nd stage of the model,
fuel properties alone with engine specification and operating condition can be used
as input layer, whereas, engine performance, emission and combustion parameters
can be used as the target vector of output layer. The structure of such a two-stage
ANN model as proposed is shown in Figure 12. It can be expected that such an
approach will generate new knowledge, based upon which, second-generation
biodiesel will be more sustainable, commercially available and a key contributor to
the mainstream global energy system.
Figure 2-12: Proposed structure of ANN model
Chapter 2: Literature Review 59
2.5 CONCLUSIONS
Biodiesel, produced from renewable feedstocks represents a more sustainable source
of energy and will therefore play a significant role in providing the energy
requirements for transportation in the near future. However, first-generation
biodiesels used around the World today are unlikely to be sustainable in the long
term as a result of being produced from edible oil feedstock. Second-generation
biodiesels produced from non-edible feedstocks have the potential to overcome this
challenge, and to serve as a more sustainable energy source in the near future.
Chemically, all biodiesels are fatty acid methyl esters (FAME), produced from raw
vegetable oil. Numerous fatty acids, ranging in chain length from 6 to 24, have been
found in various biodiesels, which are identical to their respective feedstock.
However, clear differences in chemical structure are apparent from one feedstock to
the next in terms of chain length, degree of unsaturation, number of double bonds
and double bond configuration-which all determine the important fuel properties of
biodiesel. This includes kinetic viscosity, density, cetane number, calorific value,
flash point, oxidation stability, cold temperature properties and iodine value.
Therefore, different levels of combustion performance and emission levels have been
observed in the literature when using different types of biodiesel as diesel engine
fuel. While considering production optimisation and engine durability issues, similar
trends have been observed. The literature reviewed in this study has assured that the
suitability of any biodiesel as automobile engine fuel can be explained largely
through the chemical composition of its respective feedstock.
ANN is a powerful computational modelling tool which has the ability to identify
complex relationships from input-output data. It can result in a higher level of
accuracy in its prediction ability when compared with other statistical methods.
Therefore, ANN has emerged and has found extensive acceptance in many
60 Chapter 2: Literature Review
disciplines for modelling complex real world situations. However, most of the
literatures compared the prediction accuracy of ANNs and statistical methods based
on MSE or RMS which may not much appropriate. It is recommended to consider
some other error measurement technique including residual plots, the maximum error
percentage, minimum error percentage etc.
Recent literature shows that the complex relationship between biodiesel chemical
composition, fuel properties and diesel engine combustion performance can be
established at different operating condition conation by using ANNs. Several ANN
models have been developed to estimate the combustion-related performance of
various fuels in automobile engine applications with a high prediction accuracy, as
shown in Table 6. However, applicability of these models is limited to a specific
engine and to fuel types that have been used to collect the experimental data upon
which the network has been trained. These models also have serious limitations,
considering the limited number of engine operating parameters used in the
experiments. The automobile engine is a complex system, with a large number of
parameters directly influencing its combustion performance. Moreover, no study has
considered the physical and chemical properties of fuel while developing the ANN
model for predicting combustion performance, in spite of there being a strong
correlation between these parameters. Therefore, it would be worthwhile for
researchers to develop a universal ANN model which will be able to predict the
combustion performance of versatile automobile engines and fuel types. To ensure
the most robust ANN model, data should be used which cover as much a range as
possible. This model will able to access the sustainability of the wide ranges of
biodiesel feedstock collecting from different origin.
Chapter 3: Artificial neural network (ANN) model development 61
Chapter 3: Artificial neural network (ANN) model development
Correlation between chemical composition and properties
of biodiesel – a principal component analysis (PCA) and
artificial neural network (ANN) approach
Abstract
Biodiesel, produced from renewable feedstocks, represents a more sustainable source
of energy and will, therefore, play a significant role in providing the energy
requirements for transportation in the near future. Chemically, all biodiesels are fatty
acid methyl esters (FAME), produced from raw vegetable oil and animal fat.
However, clear differences in chemical structure are apparent from one feedstock to
the next, in terms of chain length, degree of unsaturation, number of double bonds
and double bond configuration – all of which determine the fuel properties of
biodiesel. In the present study, the sensitivity of biodiesel fuel properties was
compared against its chemical composition using experimental data. The effective
fuel properties include kinematic viscosity, density, higher heating value, oxidation
stability, cold filter plugging point temperature, flash point temperature and iodine
value. Principal component analysis (PCA) was used to understand the relationship
between important properties of biodiesel and its chemical composition. Finally,
several artificial intelligence-based models were developed to predict specific
biodiesel properties based on its chemical composition. As the relationship between
biodiesel properties and its chemical composition is complex, and there is a lack of
available knowledge to develop traditional mathematical models, a data driven
modelling technique, namely an artificial neural network (ANN), was used in this
study. The experimental study was conducted in order to generate training data for
the ANN. Available (experimental) data from the literature was also employed for
this modelling strategy. The analytical part of this study found a complex multi-
dimensional correlation between chemical composition and biodiesel properties.
62 Chapter 3: Artificial neural network (ANN) model development
Average numbers of double bonds in the chemical structure (representing the
unsaturated component in biodiesel) and the poly-unsaturated component in biodiesel
had a great impact on biodiesel properties. The simulation study demonstrated that
ANN was able to predict the relationship between biodiesel chemical composition
and fuel properties. Therefore, the ANN model developed in this study could be a
useful tool in estimating biodiesel fuel properties, instead of undertaking costly and
time consuming experimental tests.
3.1 INTRODUCTION
Vegetable oil methyl or ethyl esters, commonly referred to as biodiesel, are a
renewable liquid fuel alternative to petroleum diesel. In technical terms, biodiesel is
diesel engine fuel comprised of mono-alkyl esters of long chain fatty acids derived
from vegetable oil or animal fats (Demirbas 2008b). These mono-alkyl esters are the
main chemical species that give biodiesel similar or better fuel properties compared
with petroleum diesel (Fernando et al. 2007). It is also safer to handle, store and
transport than petroleum diesel because it is biodegradable and non-toxic. It has a
higher flash point than diesel. One of the major advantages of biodiesel is that it has
the potential to reduce dependency on imported petroleum through the use of
domestic feedstocks for fuel production (Demirbas 2008b; Fernando et al. 2007).
Biodiesels are usually made from vegetable oils and animal fat feedstocks, through a
chemical reaction called trans-esterification. In this process, the pure oil and fat is
converted from natural oil (three long chain carbon molecules struck together by
glycerine) into three mono-alkyl esters (three separated long chain carbon
molecules). Triglycerides are allowed to react with alcohol (normally methanol)
under acidic or basic catalyst conditions, producing fatty acid esters of the respective
alcohol and free glycerol. After the complete reaction, glycerol is removed as a by-
product and esters remain, which are known as biodiesel (Jahirul 2013).
Chapter 3: Artificial neural network (ANN) model development 63
Quality standards are crucial for the commercial use of any fuel product, which serve
as guidelines for the production process, to assure customers are buying fuels at the
appropriate quality, and provide authorities with approved tools for the assessment of
safety risks and environmental pollution. The fuel quality, which eventually affects
fuel combustion performance, exhaust emissions and engine durability, is more
sensitive in modern diesel engines, as the use of high pressure (about 75,000 bar) in
common rail fuel injection systems has increased (Haseeb et al. 2011a). Cetane
number (CN), widely used as a diesel fuel quality parameter, is a measure of the
combustion quality of diesel fuels during compression ignition. It is related to the
ignition delay (ID) time (i.e. the time that passes between injection of the fuel into
the cylinder and onset of ignition). The higher the CN the lower the ignition delay.
An adequate CN is required for good engine performance. A high CN helps ensure
good cold start properties and minimises the formation of white smoke. On the other
hand, lower CNs may result in diesel knocking and increased exhaust emissions
(Meher, Vidya Sagar and Naik 2006a). Kinematic viscosity (KV) is one of the engine
fuel properties that play a dominant role in the fuel spray, fuel-air mixture formation
and combustion process in diesel engine applications. It effects engine combustion,
performance and emission, especially carbon monoxide (CO) and unburnt
hydrocarbon (UHC) (Knothe and Steidley 2007). In a light-duty diesel engine, the
CO and UHC could increase by 0.02% (by volume) and 1 ppm (by volume),
respectively, by increasing 1 cSt. fuel viscosity (Ng, Ng and Gan 2012). Moreover,
high viscosity is more of a problem in cold weather, as viscosity increases with
decreasing temperature (Joshi and Pegg 2007b). On the other hand, low fuel
viscosity is not desirable, because fuel with low viscosity does not provide sufficient
lubrication for the precision fit of the fuel injection pumps, resulting in leakage or
increased wear. Therefore, all biodiesel standards defined the upper and lower limit
of fuel viscosity. Heating value is another fuel property indicating the energy content
in the fuel. Depending on the amount of oxygen contained in the biodiesel, it is
generally accepted that biodiesel from all sources has about 10% less energy content
compared with petroleum diesel. Similarly, other important fuel properties like
density, oxidation stability (OS), cold filter plugging point temperature (CFPP), flash
point temperature (FP) and iodine value (IV) also effect the combustion performance
of diesel engines and have been discussed by Jahirul et al. (2013). However, those
properties are largely determined by the complex chemical structure of biodiesel. For
64 Chapter 3: Artificial neural network (ANN) model development
instance, the KV of biodiesel increases with increasing carbon chain length in FAME
(Knothe and Steidley 2007). As the lengths of the acid and alcohol segments in the
ester molecules increase, so does the degree of random intermolecular interactions
and consequently, the kinematic viscosity. As reported in the literature (Refaat
2009a), shorter fatty acid chains with longer alcohol content in biodiesel display
lower viscosity than esters with longer fatty acid chains and shorter alcohol. Other
factors that influence biodiesel properties include number and position of the double
bonds, degree of saturation, molecular weight, branching hydroxyl groups and the
level of impurities, such as free fatty acid and unreacted glycerides or glycerol etc.
However, understanding the relationship between chemical composition and the
properties of biodiesel is not a trivial task, since the chemical structure of biodiesel is
so complex. The available mathematical models are still limited in their ability to
describe biodiesel fluid properties, in terms of their corresponding chemical
structure.
In recent years, ANN modelling techniques have increased in popularity due to their
excellent capability to learn and model complex non-linear relationships. The most
important feature of artificial neural networks is their ability to solve problems
through learning by example, without having the process knowledge. More precisely,
ANNs work like a ‘black box’ model, and they can map any relationship based on
system input and output data without knowing the detailed or complete information
about the problem. Therefore, ANNs have been successfully applied in various
disciplines, including neuroscience (Alkım, Gürbüz and Kılıç 2012), mathematical
and computational analysis (Costa, Braga and De Menezes 2012), learning systems
(Carrillo et al. 2012) and engineering design and application (Samura and Hayashi
2012; Gao et al. 2012; Minnett et al. 2011a). The application of ANN has also been
used to predict the fuel properties of biodiesel. Ramadas et. al. (Ramadhas et al.
2006a) used ANN to predict the CN of biodiesel based on fatty acid profile. ANNs
have also been used to predict viscosity, flash point and fire point, based on diesel-
biodiesel blend ratio (Kumar et al. 2007). However, those prediction models were
limited to a specific biodiesel and/or experimental conditions. No investigation has
been reported in the literature to develop an ANN model to predict biodiesel
properties for a wide range of feedstocks. In addition, the available literature have
Chapter 3: Artificial neural network (ANN) model development 65
not considered the impurities generally contained in biodiesel. Therefore, this study
aimed to develop a robust ANN model to estimate the important fuel properties of
biodiesel from its chemical composition. During the model development process, this
study also aimed to make an in-depth investigation of chemical composition and
important fuel properties of biodiesel, and to analyse the relationship between them.
3.2 DATA COLLECTION
Two types of data were used in this study: BERF and literature data. BERF data was
obtained from the experimental study of nine biodiesel samples, using the biofuel
engine research facility (BERF) testing facility. Among the samples, biodiesel
derived from canola oil (COME), cotton seed oil (CSOME), tallow (TOME),
soybean oil (SOME) and waste cooking oil methyl ester (WCO) are commercially
available. The other biodiesel samples, named C810, C1214, C1618 and C1822,
were produced by the fractionating of palm oil biodiesel produced by Proctor &
Gambel. The chemical composition, fatty acid profile and glyceride content of nine
biodiesel samples were analysed using gas chromatography-mass spectrometry (GC-
MS). Biodiesel samples were diluted 1:100 with n-hexane and 1uL samples were
injected into a PerkinElmer Clarus 580 GC-MS fitted with an Elite - 5MS, 30m x
0.25mm x 0.25um column. The split ratio was 30:1, with a column flow of 1mL/min
He. The temperature program was as follows: 120 °C initial, holding 0.5min,
ramping 10 °C/min until 310 °C, and holding for 2 min. Masses were analysed over
the range 40-350m/z. The total amount of carbon (C), oxygen (O) and hydrogen (H)
content in biodiesel was obtained through elemental analysis. In addition, acid
number and six fuel properties of biodiesel, including cetane number (CN) kinematic
viscosity (KV), density, higher heating value (HHV), were obtained through
experimental study, following recognised international standards, as shown in Table
3-1.
66 Chapter 3: Artificial neural network (ANN) model development
Table 3-1 Biodiesel property test standard
Fuel properties Unit Test Method
Element analysis (C, O, H) wt.% DIN EN 15104 Cetane value (CN) - DIN 51773 Kinematic Viscosity (KV) cSt ASTM D445
Density Kg/l ASTM D4052
Higher heating value (HHV) Mj/kg ASTM D4868
Acid number (AN) - ASTM D664
Literature data were collected from papers published in recognised international
journals, conferences and reports of renowned research centres around the world.
Scientific and electronic databases, including Elsevier, Taylor and Francis,
DieselNet, Scopus, Springer, Wiley International, American Chemical Society,
IEEE, SAGA Publication, MDPI etc., were searched for relevant papers for this
study. More than 120 papers were collected, most of which were published in the last
decade, containing experimental results of the chemical composition of biodiesel
along with corresponding fuel properties. During secondary data collection, special
care was taken to ensure the quality of the data and eliminate duplication. Data was
only taken from the literature when the experiments were conducted by the authors
themselves, following recognised international standards. Some extreme data was
excluded from database, due to the unexpected nature of the results. Data was also
eliminated from the database if it was too dissimilar compared to the fuel properties
recorded in the primary data collection results. Furthermore, the experimental results
for density and kinematic viscosity of biodiesel are highly dependent on temperature
(Joshi and Pegg 2007b; Yuan, Hansen and Zhang 2009). Although 15 ºC and 40 °C
temperatures are recommended for density and kinematic viscosity respectively,
some researchers did not mention the test temperature. Therefore, those data were
excluded from the database. Since the properties of a particular biodiesel can be
varied depending on the type of alcohol (methyl, ethyl etc.) used in the production
process, this study only considered methyl esters for inclusion in the database. The
list of papers, including feedstock use and country, are tabulated in Table 3-2. A
large number of feedstocks were investigated worldwide for biodiesel production;
including edible and non-edible vegetable oils, waste cooking oils, beef tallow,
chicken fats, fish oils, algae etc. It is also interesting to note that many investigations
Chapter 3: Artificial neural network (ANN) model development 67
used pure methyl esters in order to represent actual biodiesel, which are mostly
produced by artificial chemical processes. The most popular edible feedstock for
biodiesel investigated worldwide was soybean, followed by palm, sunflower, canola
and rapeseed oil. Among non-edible oils, the most-investigated feedstock was
Jatropha, as shown in Table 3-2.
Chapter 3: Artificial neural network (ANN) model development 68
Table 3-2: Biodiesel datasets investigated in this study
Feedstock Authors’ affiliation References
Algae USA (Do et al. 2011) Almond Iran, Nigeria (Atapour and Kariminia 2011; Giwa and Ogunbona 2014) Apricot Turkey (Gumus and Kasifoglu 2010) Babassu USA, India, Brazil (Sanford et al. 2009; Rodrigues Jr et al. 2006; Barnwal and Sharma 2005; Nogueira Jr et al. 2010) Brassica Austria (Dorado et al. 2004) Camelina USA, China, Ireland (Sanford et al. 2009; Chung 2010; Wu and Leung 2011; Fröhlich and Rice 2005; Moser and
Vaughn 2010; Soriano Jr and Narani 2012) Canola USA, Turkey, China,
Canada (Sanford et al. 2009; Albuquerque et al. 2009; Koçak, Ileri and Utlu 2007; Davis et al. 2009; Hu et al. 2005; Haagenson et al. 2010; Chhetri and Watts 2012a; Moser 2008; Chhetri and Watts 2012b; Kinast 2003b; Do et al. 2011; Duncan et al. 2010; Cecrle et al. 2012)
Coconut Philippine, USA, India, Thailand, Brazil
(Sanford et al. 2009; Alleman and McCormick 2006; Rodrigues Jr et al. 2006; Tan, Culaba and Purvis 2004; Kumar et al. 2010; Nakpong and Wootthikanokkhan 2010; Duncan et al. 2010; Cecrle et al. 2012; Feitosa et al. 2010)
Coffee Greece, Japan (Deligiannis et al. 2011; Todaka et al. 2013) Corn Brazil, USA, Romania (Lin, Huang and Huang 2009; Rodrigues Jr et al. 2006; Dantas et al. 2011; Serdari et al. 1998;
Cursaru, Neagu and Bogatu 2013) Cottonseeds Brazil, Greek, USA (Albuquerque et al. 2009; Royon et al. 2007; Rashid, Anwar and Knothe 2009; Demirbaş 2002;
Tang, Salley and Simon Ng 2008; Nogueira Jr et al. 2010) Fish oil Taiwan, Chile, Turkey (Lin and Li 2009a; Reyes and Sepulveda 2006; Behçet 2011) Golden cress Egypt (Ali 2013) Grape Spain, Romania (Ramos et al. 2009; Cursaru, Neagu and Bogatu 2013) Hazelnut Turkey (Koçak, Ileri and Utlu 2007; Moser 2012; Demirbaş 2002) Hepar USA (Sanford et al. 2009) Jathropa USA, India, Canada,
South Africa, China, Japan
(Sanford et al. 2009; Sarin et al. 2007; Choudhury and Bose 2008; Jain and Sharma 2012; Chhetri et al. 2008; Singh and Padhi 2009; Aransiola et al. 2012; WANG et al. 2012; Chhetri and Watts 2012a, 2012b; Kumar Tiwari, Kumar and Raheman 2007; Todaka et al. 2013)
Chapter 3: Artificial neural network (ANN) model development 69
Lard Portugal, Korea (South), Spain, USA
(Lee, Foglia and Chang 2002; Dias, Alvim-Ferraz and Almeida 2009; Berrios et al. 2009; Kinast 2003b; Wyatt et al. 2005)
Linseed Brazil, Lithuania, India, Egypt
(Rodrigues Jr et al. 2006; Guzatto, De Martini and Samios 2011; Samios et al. 2009a; Lebedevas et al. 2006; Puhan et al. 2009; El Diwani and El Rafie 2008; Radha and Manikandan 2011)
Mahua India, Turkey (Ghadge and Raheman 2005; Godiganur, Suryanarayana Murthy and Reddy 2009; Kapilan and Reddy 2008; Demirbas 2009a)
Mustard Turkey, Brazil, USA (Bannikov 2011; Jham et al. 2009) Neem India, South Africa,
Pakistan (Ragit et al. 2011; Aransiola et al. 2012; Sivalakshmi and Balusamy 2012; Sardar et al. 2011; Radha and Manikandan 2011)
Olive Spain, Greece, USA, Romania
(Ramos et al. 2009; Dorado et al. 2003; Kalligeros et al. 2003; Cecrle et al. 2012; Cursaru, Neagu and Bogatu 2013)
Palm Malaysia, Indonesia, Greece, Colombia, Japan, Spain, India, USA, Canada, Romania
(Karavalakis, Stournas and Bakeas 2009; Sarin et al. 2007; Kousoulidou et al. 2010; Ramos et al. 2009; Benjumea, Agudelo and Agudelo 2008; Kalam and Masjuki 2002a; Ng and Gan 2010; Loh, Chew and Choo 2006; Crabbe et al. 2001; Kalam and Masjuki 2002b; Pérez et al. 2010; Barnwal and Sharma 2005; Moser 2008; Do et al. 2011; Vedaraman et al. 2011; Cecrle et al. 2012; Park et al. 2008; Cursaru, Neagu and Bogatu 2013)
Peanut USA, Spain, Turkey, India, China, Romania
(Lin, Huang and Huang 2009; Ramos et al. 2009; Moser 2012; Pérez et al. 2010; Davis et al. 2009; Kaya et al. 2009; Barnwal and Sharma 2005; SUN et al. 2008; Cursaru, Neagu and Bogatu 2013)
Popyseed Turkey (Demirbaş 2002) Rapeseed USA, Greece,
Lithuania, Turkey, Spain
(Karavalakis, Stournas and Bakeas 2009; Wu 2008; Lin, Huang and Huang 2009; Senatore et al. 2000; Rashid and Anwar 2008a; Sahoo et al. 2007; Ramos et al. 2009; Fröhlich and Rice 2005; Lebedevas et al. 2006; Demirbaş 2002; Pérez et al. 2010; Cecrle et al. 2012; Park et al. 2008; Todaka et al. 2013)
Rice barn India (Sinha, Agarwal and Garg 2008) Rubberseed India (Ramadhas, Jayaraj and Muraleedharan 2005; Ikwuagwu, Ononogbu and Njoku 2000) Safflower USA (Rashid and Anwar 2008a; Demirbaş 2002) Sesame Nigeria, Pakistan (Betiku and Adepoju 2013; Ahmad, Ullah, et al. 2011) Soybean USA, Brazil, Spain, (Ali, Hanna and Cuppett 1995b; Wu 2008; Armas, Yehliu and Boehman 2010; Albuquerque et al.
70 Chapter 3: Artificial neural network (ANN) model development
India, Turkey, China 2009; Sarin et al. 2007; Lin, Huang and Huang 2009; Ali, Hanna and Cuppett 1995a; Rodrigues Jr et al. 2006; Ramos et al. 2009; Guzatto, De Martini and Samios 2011; Canakci and Van Gerpen 2003; Alcantara et al. 2000; Schwab, Bagby and Freedman 1987; Pérez et al. 2010; Davis et al. 2009; Barnwal and Sharma 2005; Moser 2008; Kinast 2003b; Duncan et al. 2010; Candeia et al. 2009; Cecrle et al. 2012; Pereira et al. 2007; Tang, Salley and Simon Ng 2008; Wyatt et al. 2005; Canakci 2005a; Feitosa et al. 2010; Qi et al. 2009; Moraes et al. 2008; Nogueira Jr et al. 2010; Park et al. 2008; Shah et al. 2013)
Soapnut Canada (Chhetri and Watts 2012a, 2012b) Sunflower Greek, Turkey, Spain,
India, Greece, USA, Pakistan, Romania
(Sarin et al. 2007; Lin, Huang and Huang 2009; Royon et al. 2007; Serdari et al. 1998; Ramos et al. 2009; El Diwani and El Rafie 2008; Demirbaş 2002; Pérez et al. 2010; Barnwal and Sharma 2005; Kalligeros et al. 2003; Antolın et al. 2002; Moser 2008; Rashid et al. 2008; Cursaru, Neagu and Bogatu 2013)
Tall Turkey (Altıparmak et al. 2007) Tallow USA, Lithuania, Spain,
India, Turkey, Brazil, Japan
(Ali, Hanna and Cuppett 1995b; Sanford et al. 2009; Ali, Hanna and Cuppett 1995a; Lebedevas et al. 2006; Alcantara et al. 2000; Barnwal and Sharma 2005; Öner and Altun 2009; Ramalho et al. 2012; Kinast 2003b; Tang, Salley and Simon Ng 2008; Wyatt et al. 2005; Moraes et al. 2008)
Terebinth Turkey (Özcanlı, Keskin and Aydın 2011) Terminalia Brazil (Dos Santos et al. 2008) Turnip USA (Shah et al. 2013) Walnut USA (Moser 2012) Waste cooking Vietnam, Spain,
Taiwan, Turkey, Brazil, Spain, USA
(Lin, Huang and Huang 2009; Phan and Phan 2008; Encinar, Gonzalez and Rodríguez-Reinares 2005; Lin and Li 2009a; Koçak, Ileri and Utlu 2007; Guzatto, De Martini and Samios 2011; Lapuerta et al. 2008; Alcantara et al. 2000; Demirbas 2009b; Chhetri, Watts and Islam 2008; Cecrle et al. 2012)
Yellow grease USA (Canakci and Van Gerpen 2003; Kinast 2003b) Pure methyl ester
USA (Rodrigues Jr et al. 2006; Knothe 2005; Knothe and Steidley 2005; Knothe 2008; Moser 2011)
Chapter 3: Artificial neural network (ANN) model development 71
Table 3-3: Structural formulae for fatty acids methyl ester found in biodiesel samples
Acid chain C:N* Type* CL*** Chemical structure Caprilic C8:0 SFA 8 CH3(CH2)6COOH Capric C10:0 SFA 10 CH3(CH2)8COOH Lauric C12:0 SFA 12 CH3(CH2)10COOH Myristic C14:0 SFA 14 CH3(CH2)12COOH Palmitic C16:0 SFA 16 CH3(CH2)14COOH Palmitoilic C16:1 MUFA 16 CH3(CH2)6 CH=CH (CH2)6 COOH Stearic C18:0 SFA 18 CH3(CH2)16COOH Oleic C18:1 MUFA 18 CH3(CH2)7 CH=CH (CH2)7 COOH Linoleic C18:2 PUFA 18 CH3(CH2)4 CH= CHCH2CH =CH (CH2)7 COOH Linolenic C18:3 PUFA 18 CH3(CH2)2CH=CHCH2CH=CHCH2CH
=CH(CH2)7 COOH Gondonic C20:1 MUFA 20 CH3(CH2)7 CH=CH (CH2)9 COOH Erucic C22:1 MUFA 22 CH3(CH2)7 CH=CH (CH2)11 COOH * C: the number of carbon atoms and N: the number of double bonds of carbon atoms in the fatty acid chain. ** SAF: saturated fatty acids, MUFA: Mono unsaturated fatty acids and PUFA: Poly unsaturated fatty acid. *** CL: Chain length of hydrocarbon of respective methyl ester
3.3 RESULTS AND DISCUSSION
3.3.1 Chemical composition
The main chemical components of biodiesel are mono-alkyl esters of fatty acids,
which are feedstock dependent and vary significantly from feedstock to feedstock.
Therefore, a wide range of fatty acid profiles were found in the collected data.
Among those, most commonly found FAMEs are tabulated in Table 3-3, together
with their structure and name. These FAMEs are straight-chain compounds ranging
in size from 8−22 carbons, which are mainly one of three types: saturated, mono-
unsaturated and poly-unsaturated. In the saturated acid, no hydrogen can be added
chemically, and they contain only single bonds, whereas in mono-unsaturated fatty
acids, one hydrogen can be added and it contains one double bond. Similarly, in
poly-unsaturated fatty acids, more than one hydrogen can be added and it contains
multiple double bonds. In general, fatty acids are designated by two numbers: the
first number denotes the total number of carbon atoms in the fatty acid and the
72 Chapter 3: Artificial neural network (ANN) model development
second is the number of double bonds. For example, 14:1 designates Myristoleic acid
which has 14 carbon atoms and one double bond.
Table 3-4 shows the chemical composition of nine biodiesel tested in this study. The
C810, C1214 and C1416 samples were mostly saturated FAMEs of different chain
lengths. C1822 contained very high unsaturated fatty acid methyl, which was over
90% by weight. Although TOME contained about 50% unsaturated fatty acids, most
those were mono-unsaturated fatty acid methyl esters which accounted for 45.1% of
total fatty acids. The average chain length (ACL) of C810 was 8.93; which was the
lowest among the samples tested in this study. CSOME and C1875 had the highest
ACL among the samples tested, being 17.93. CSOME and SOME were rich in
PUFAs, containing 55.4% and 49.38 % by weight, respectively. The oxygen content
of the samples was found to be between 10-11% by weight, except for samples C810
and C1214, which contained 18.72% and 13.31% by weight, respectively. The
average number of double bonds (ANDB) in the samples ranged from 0 to 1.48. The
COME contained more than 1% mono-glycerides, whereas other samples were found
to have very low levels, ranging from 0% to 0.58% by weight. Acid number
(representing the free fatty acid content) was found to range from 0.22 to 2.69.
Overall, the results obtained in this study showed a wide range of chemical
compositions, which were useful for building a robust model.
Table 3-4: Chemical composition of tested biodiesel
Biodiesels COME CSOME TOME SOME WCO C810 C1214 C1618 C1822
FAME C8:0 0 0 0 0 0 52.16 0 0 0 C10:0 0 0 0.36 0 0 46.38 0.17 0 0 C12:0 0.31 0 0.36 0 0 1.38 47.8 0.1 0 C14:0 0.35 0.53 3.07 0 0 0 18.89 0.06 0.03 C16:0 11.93 2.51 26.62 11.08 11.23 0 10.19 21 4.45 C16:1 0.22 0.55 3.13 0.18 0 0 0 0 0.12 C18:0 2.43 4.13 19.89 5.23 3.5 0 2.55 9.47 2.53 C18:1 55.72 35.78 41.66 33.82 67.03 0 18.53 58.72 71.04 C18:2 25.87 55.04 2.57 48.38 18.3 0 1.76 9.98 18.69 C18:3 0.96 0.59 0 1.22 0 0 0 0 0 C20:0 0 0.31 0 0 0 0 0 0 0 C20:1 0.57 0.30 0.31 0 0 0 0 0.24 1.03 C22:1 0.21 0.01 0 0.24 0 0 0 0 0 Mono-glyceride 1.03 0.58 0.09 0 0 0.08 0 0 0 Acid number (AN) 0.91 0.22 2.31 0.34 1.4 0.91 2.69 1.04 1.48
Chapter 3: Artificial neural network (ANN) model development 73
Oxygen (O, wt.%) 10.81 10.80 10.97 10.95 10.93 18.72 13.31 10.97 10.63 Hydrogen (H,wt.%) 11.95 11.81 12.19 11.98 12.17 11.69 12.30 12.25 11.87 Carbon (C, wt.%)% 75.86 76.55 74.81 77.00 76.95 69.51 74.23 76.35 75.39 Ave. chain length (ACL)
17.74 17.93 17.22 17.78 17.78 8.98 14.15 17.57 17.93
Ave. number of double bond (ANDB)
1.12 1.48 0.51 1.33 1.04 0.00 0.22 0.79 1.12
Saturation (wt%) 16.41 8.32 52.33 16.38 14.67 100 79.71 31.06 9.12 MUFA* (wt.%) 56.72 36.64 45.10 34.24 67.03 0.00 18.53 58.96 72.19 PUFA* (wt.%) 26.87 55.04 2.57 49.38 18.30 0.00 1.76 9.98 18.69
Table 3-5: BREF experimental results of biodiesel properties
Fuel properties
Diesel COME CSOME TOME SOME WCO C810 C1214 C1618 C1822
Higher heating value (MJ/kg)
45.93 38.8 38.42 38.2 39.4 39.75 35.34 38.44 37.59 39.83
Kinematic Viscosity (cSt)
2.64 5.45 4.08 4.52 3.86 4.82 1.95 4.37 4.95 5.29
Density (kg/l) 0.838 0.898 0.893 0.888 0.893 0.890 0.867 0.875 0.879 0.889 Cetane value 50.6 - 87 - - 58.6 42 75.3 70 59.2
The most-common fatty acids found in biodiesel samples were: Oleic (C18:1)
followed by Stearic (18:0), Linoleic (C18:2), Palmitic (C16:0) and Linolenic (C18:3)
acid esters. Figure 3-1 shows that these fatty acid esters were found in almost every
biodiesel sample. It is interesting to note in Figure 4 that the Oleic (C18:1) and
Linoleic (C18:2) acid esters were not only represented in most of the biodiesel
samples, but they also showed highest average weight percentage in the biodiesel
samples, which were about 40% and 32%, respectively. On the contrary, an average
of 7.5% and 6.5% Linolenic (C18:3) and Stearic (18:0) acids methyl esters were
present in the samples. This also reflects the average values for chain length,
saturated fatty acid esters, mono-unsaturated esters and poly-unsaturated methyl
esters, as shown in Table 3-3. Apart from fatty acid methyl esters, other common
chemicals found in the biodiesel were unreacted mono-glyceride and free fatty acid,
represented as the acid value.
74 Chapter 3: Artificial neural network (ANN) model development
Figure 3-1: Number and average weight in percentages of fatty acid methyl esters
found in the samples
3.3.2 Fuel properties
Fuel properties of the tested biodiesels, in terms of higher heating value (HHV),
kinematic viscosity (KV), density and cetane number (CN), are summarised in the
Table 3-6. The properties of petroleum diesel are also shown in the Table 3-5 for
comparison purposes. Similar to chemical composition, clear differences in fuel
properties were found between biodiesels in the experimental results. Reflecting its
exceptional chemical composition, C810 showed the lowest values for all properties
compared with other biodiesels investigated. Other fractionated methyl esters
showed similar properties to commercial biodiesels. The HHV of biodiesel was
found to be just below 40 Mj/kg, which is about 10% lower than that of petroleum
diesel. This is mainly due to the oxygen content in the chemical structure of its
FAMEs (Demirbas 2008b; Refaat 2009a). The KV, density and CN of these
biodiesels were found to be much higher than petroleum diesel, except for C810. The
KV of commercial biodiesels was found to be double than that of petroleum diesel.
The highest KV was found for COME (5.45 cSt.) followed by C1822, C1618 and
CSOME. The density of biodiesels was found to be an insensitive parameter among
Chapter 3: Artificial neural network (ANN) model development 75
the biodiesels tested, however the commercial biodiesels showed a slightly higher
density compared with commercial biodiesel. The difference in biodiesel CN among
the biodiesel samples investigated in this study was very noticeable, ranging from 42
to 87. CSOME, C1214 and C1618 showed very high CN compared to petroleum
diesel. The CN of C810 was found to be 42, which was much lower than petroleum
diesels.
A good amount of quality data are required for the purpose of a correlation study and
robust model development, with the accuracy of the ANN model found to increase
with a greater number of data sets. Therefore, secondary data were collected from
published peer reviewed literature, as summarised in Table 4. Among the eight
biodiesel properties identified during secondary data collection, four of those
(including oxidation stability (OS), cold filter plugging point (CFPP), flash point
(FP) and iodine value) were not investigated experimentally, due to a lack of
laboratory facilities. A wide variety of biodiesel types were investigated in the
literature (as discussed in the previous section) and variation was also observed in
terms of the number of biodiesel properties examined, with no single work reporting
on all of the important fuel parameters. The most commonly studied biodiesel
properties were KV, Density and HHV. Overall, we were able to collect about 350
data sets for each of these properties, along with the corresponding chemical
composition. On the other hand, flash point and oxidation stability were the least
studied parameter, consisting of less than 200 data sets.
The average fuel properties reported in the collected secondary data were found to be
within the limits of European (EU), American (US) and Australian (AU) biodiesel
standards, except for OS. The average OS was found to be 4.73 hr, which is much
lower than the minimum OS requirement (6 hr minimum) of EU and AU biodiesel
standards. These results indicate that a vast number of the investigated biodiesels
were unlikely to fulfil EU biodiesel standards, and this is may be one of the major
issues that restricts the wide spread use of biodiesel in conventional diesel engines.
The biodiesels that showed poor OS and were rich in unsaturated FAME include
soybean, sunflower, safflower, corn, cottonseeds, linseeds, jatropha and camelina.
European and Australian biodiesel standards also impose tight restrictions on
76 Chapter 3: Artificial neural network (ANN) model development
biodiesels kinematic viscosity, limiting it to a minimum of 3.5 and a maximum of 5
cSt. However, the KV of biodiesels in the secondary data ranged from 0.99 to 7.21
cSt, which means that many of them would be unlikely to meet EU and AU biodiesel
standards in terms of KV. US biodiesel standards are more lax in terms of KV (1.9 –
6 cSt), however they place tighter restrictions on CN and FP, which means many
biodiesels identified in the secondary dataset would still be unlikely to meet US
standards. Overall, most of the maximum and minimum values for fuel properties
were outside the range of biodiesel standards, demonstrating the significant level of
variation the in secondary data collected. This was not unexpected, because the
secondary data was collected from a large number of different biodiesels with a wide
variety of chemical structures. Therefore, the collected secondary data were useful
for conducting an in-depth correlation study and developing the artificial neural
network model.
Table 3-6: Summary of the secondary data for biodiesel properties
Properties
Biodiesel standard
N* Max.* Min.* Ave.* ASTM D7651
EN 14214
Australian
Cetane number 47, min 51, min 51, min 227 86 37 54.59
Kinematic viscosity (cSt.) 1.9-6 3.5-5 3.5-5 368 6 1.97 4.42
Density (kg/l) n/a 0.86-0.90 0.86 -0.90 345 0.924 0.829 0.876
Higher heating value (Mj/kg) n/a n/a n/a 336 45.5 35.86 39.91
Oxidation stability (hrs) 3, min 6, min 6, min 198 11.4 0.2 4.73
Cold filter plugging point (°C) Report Report Report 254 17 -20 -1.11
Flash point (°C) 130, min 120, min 120, min 187 191 96 159.83
Iodine value (g iod/100 g) n/a 120, max n/a 214 184.5 0.3 96.1
*N =Number of data sets; Max = Maximum value; Min. = Minimum value; Ave. = Average value
Chapter 3: Artificial neural network (ANN) model development 77
(a)
(b)
Figure 3-2: Correlation of (a) C18:2 with oxidation stability; (b) H2 with CN
78 Chapter 3: Artificial neural network (ANN) model development
3.3.3 Correlation of chemical composition and fuel properties
Due to variations in chemical composition, the fuel properties of biodiesel
significantly differ from one another. Linoleic (C18:2) and linolenic (C18:3) fatty
acid methyl esters are not only present in most biodiesel, but also influence almost all
of the fuel properties. For instance, OS stability and CFPP decrease significantly
with an increase in both C18:2 and C18:3. Figure 3-2a shows that biodiesels with
C18:2 are more than 40% more likely to fail to meet the lower limit of Australian
and European biodiesel standards and over 60 also failed to meet the US biodiesel
standard. However, those fatty acid methyl esters do not have much effect on KV
and FP. The stearic acid methyl ester (C18:0), which is also found frequently in
biodiesel, has an influence on all fuel properties except HHV and OS. Other
individual fatty acid methyl esters investigated in this study may not have an effect
on all fuel properties, but they have some influence in certain numbers. For example,
short chain methyl esters such as caprilic (C8:0) and capric (C10:0) acid methyl
esters were found to only correlate with KV, HHV and FP, while the palmitoilic acid
methyl ester (C16:1) was unlikely to correlate with any of the biodiesel properties
investigated in this study.
The influence of the weight percentage of oxygen (O2), hydrogen (H2) and carbon
(C) content on fuel properties was also observed and it was found that CN was highly
influenced by biodiesel H2 content, as shown in Figure 3-2b, whereas O2 and C were
not correlated with biodiesel CN. On the hand, the correlation of O2 and C with KV
was much higher than that of H2. It is interesting to see that the carbon content of
biodiesel only affected the kinematic viscosity and density of the fuel.
Compared to individual chemical components, certain chemical characteristics
seemed to have a greater influence on biodiesel properties. For example, the average
number of double bonds (ANDB) in the biodiesel (which indicates the concentration
of unsaturated fatty acid methyl esters) was found to be very influential, affecting all
biodiesel properties investigated in this study. It had a strong negative correlation
with CN and a positive correlation with IV, as shown in Figure 3-3.
Chapter 3: Artificial neural network (ANN) model development 79
(a)
(b)
Figure 3-3: Correlation of (a) ANDB with CN; (b) ANBD with IV
80 Chapter 3: Artificial neural network (ANN) model development
This figure indicates that the biodiesel with an ANDB greater than two is likely to
fail to meet the lower limit of all biodiesel standards in terms of CN and IV. Further,
the OS of biodiesel decreased rapidly with the increase in ANDB, because a higher
number of double bonds in the fatty acid chain make it much more susceptible to
oxidation. ANDB also has a strong negative correlation with biodiesel CFPP and a
moderate positive correlation with density, HHV and FP.
As described earlier in this section, the fatty acid methyl esters found in biodiesel can
be divided into three categories. These include saturated, mono-unsaturated (MUFA)
and poly-unsaturated (PUFA) fatty acid methyl esters. This study found that PUFA’s
had the greatest influence on biodiesel properties, affecting almost all of the fuel
properties investigated in this study, including CN, density, OS, CFPP and IV. On
the other hand, MUFA’s only impacted greatly on biodiesel KV. Moreover, fuel
properties such as KV, OS and IV were found to have a high correlation with
saturated compounds.
The average chain length (ACL) was correlated with all of the fuel properties except
CFPP. It had a very strong correlation with KV, as shown in Figure 3-4a. This is
mainly due to the increase in carbon content, as well as random inter molecular
interaction in the FAME, which consequently increases the KV. For the same reason,
ACL were also found to have a strong correlation with calorific value (Figure 3-4b)
and flash point temperature. Biodiesels with an ACL less than 14 were unlikely to
meet the lower limit of both US and EU standards. On the other hand, biodiesel with
a very high ACL (over 19) are more likely to exceed the upper limit of biodiesel
standards in terms of KV.
Apart from fatty acid methyl ester compounds, mono-glyceride and acid number
(representing the amount of free fatty acid) also had some correlation with a number
of biodiesel properties. For example, biodiesel density had a strong inverse
correlation with acid number and a comparatively less strong correlation with mono-
glyceride. The acid number also had some influence the HHV and FP, whereas
mono-glyceride affected KV and IV. Overall the presence of mono-glyceride and
free fatty acid content in biodiesel was not found to correlate with any biodiesel
Chapter 3: Artificial neural network (ANN) model development 81
properties. This may be due to its relatively small concentration (less than 0.5% on
average) in the tested biodiesel.
(a)
(b)
Figure 3-4: Effect of ACL on biodiesel (a) kinematic viscosity (KV) and (b) higher
heating value (HHV)
82 Chapter 3: Artificial neural network (ANN) model development
3.3.4 Principle component analysis
The findings of the correlation study reported in the previous section indicate a
complex relationship between biodiesel quality and its chemical composition. A
particular fuel property does not depend on a single chemical parameter, rather it is
influenced by multiple parameters and factors. Therefore, multivariate data analysis
is required in order to gain a detailed understanding of this relationship. Principle
component analysis (PCA) is one of the popular multivariate data analysis techniques
used by almost all scientific disciplines. PCA is used to analyse data sets with highly
inter-correlated dependent variables. It reduces the complexity and dimensionality of
the problem, thereby extracting the most important information and analysing the
structure of the observations and variables. PCA changes the input variables into
principal components (PCs) that are an independent and linear combination of input
variables. PCA also represents patterns in the observations and variables by
displaying them as points on a diagram. In this study, PCA analysis was conducted in
order to observe the influence of chemical composition on individual fatty acid
compositions. The variables used for the principle components were individual fatty
acid methyl esters chain length ranging from 8 to 22, while the interaction terms
included average chain length (ACL), average number of double bonds (ANDB), and
weight percentages of oxygen (O), hydrogen (H), carbon (C), saturated fatty acids,
MUFA and PUFA. The variables also included the most commonly found impurities
in biodiesel, namely mono-glyceride and acid number (AN). In general, variables
which lie close to (±45°) an observation are correlated, those lying in opposite
directions (135–225°) are anti-correlated, and those lying in an orthogonal direction
have less or no influence. The direction and length of the variables is indicative of
their influence on the observation, with a short length indicative of little influence.
The results of eight fuel properties are graphically shows in Figure 3-5.
84 Chapter 3: Artificial neural network (ANN) model development
(g) (h)
Figure 3-5: Principle component analysis and correlation of biodiesel properties with
chemical composition:
(a) Cetane number (CN); (b) Kinematic viscosity (KV); (c) Density; (d) Higher
heating value (HHV); (e) Oxidation stability; (f) Cold filter plugging pint temp (CFPP); (g) Flash point temperature; (h) Iodine value (IV)
3.3.5 ANN model development
In the present work, several ANN models were developed to predict the biodiesel
properties from corresponding chemical composition. ANN represents a
mathematical relationship between input and output parameters of the system, like a
black box model. Figure 3-6 illustrates how several stages are involved in the ANN
model development process. The first step is pre-processing the training data,
whereby the raw data are rearranged according to the input-output structure of the
ANN. As the input ranges for ANN vary from one input to another, the rearranged
data are then rescaled into the range -1 to 1, and the data is spilt into training,
validation and testing data sets.
The second step involves setting the structure of the ANN, including the number of
hidden layers and hidden neurons. In this work a set of ANNs were developed where
each ANN predicted only one biodiesel parameter, which means that the ANN
predicts a single output based on several inputs. According to the literature, an ANN
(h)
Chapter 3: Artificial neural network (ANN) model development 85
with a single hidden layer is sufficient to generate single output-based model
properties (Tamura and Tateishi 1997; Hosen et al. 2014). Therefore, a single hidden
layer was used in this study and different hidden neuron sizes were used in the
training process, in order to diversify the structure of the ANNs. Finally, the best
ANN (whatever the neuron size), in terms of their performance index, was selected
for further study.
Figure 3-6: Proposed flow chart of ANN prediction model development
In order to initiate the training process, the initial training parameters and weights for
the ANN are assigned randomly, while after training, the weights are determined
85based on input and output data. When a satisfactory level of performance is
reached, the training stops and the network uses the optimal weights to make
decisions. Finally, the performance of the ANN was evaluated (in terms of
86 Chapter 3: Artificial neural network (ANN) model development
performance index) using the testing data set and the best ANN structure (in terms of
hidden neurons) was selected as final the ANN model.
The selection of input parameters (chemical composition of biodiesel) for ANN is
crucial in order to predict the parameters which reflect biodiesel properties. It is also
desirable to minimise the number of input parameter for an ANN system, in order to
reduce the computational time. In general, the best input parameters are selected
based on an understanding of the physics of the problem. In this study, the effective
parameters for chemical composition were used as input variables for the ANN,
whereas the single parameter of fuel properties was used as the output. In order to
avoid the repetitions of similar figures, a combine diagram representing eight ANN
models is shown in Figure 3-7.
Figure 3-7: Structure of ANN
Chapter 3: Artificial neural network (ANN) model development 87
Based on the correlation studies detailed in the previous section, eight parameters
were selected as biodiesel properties. The results of the correlation analysis are
presented in Figure 3-8. This figure shows that individual properties of biodiesels are
correlated with a certain number of chemical composition parameters, and, hence,
the effective inputs (from all parameters of chemical composition) for ANN were
selected as per Figure 3-10, in order to predict the corresponding parameter of fuel
properties. For an example, 14 parameters for chemical composition had a significant
influence on biodiesel oxidation stability, as shown in Figure 3-10. Therefore, these
14 parameters were selected as inputs during training of the ANN model for OS.
Variables CN KV Density HHV OS CFPP FP IV
C8:0
C10:0 C12:0 C14:0 C16:0 C16:1
C18:0 C18:1
C18:2 C18:3 C20:1 C22:1
O%
H%
C%
ACL
ANDB Saturation MUFA
PUFA Monogly AN
Figure 3-8: Analysis of influence between chemical composition and fuel properties
88 Chapter 3: Artificial neural network (ANN) model development
After selecting the input-output data for ANN training, the data were randomly split
into training, validation and testing data, as follows:
70% of total data were used for training data
10% of total data were used for validation data, and
20% of total data were used for testing data.
A single hidden layer was used for all ANNs. Different hidden neurons (6:2:30) were
used to train the ANNs, since ANN performance is sensitive to the size of hidden
neurons. Table 3-9 depicts the optimum number of hidden neurons for generating the
best ANN for estimating the corresponding parameters for biodiesel properties. Table
3-7 also shows the total number of inputs for corresponding biodiesel properties. In
this study, a logistic sigmoid (logsig) transfer function was used for the hidden layer,
while a purelinear (purelin) transfer function was used for the output layer. The
codes for the ANN training models are shown in Appendix A.
Table 3-7: Number of input variables and optimised number of neuron in ANN model
Properties Input variable
Neuron number
CN 11 20 KV 16 24 Density 14 23 HHV 13 18 OS 14 25 CFPP 11 18 FP 11 15 IV 11 20
3.3.6 Evaluation of ANN model performance
After the successful training of the network, the network was tested using the test
data set. Based on the results produced by the network, statistical methods were used
to investigate the prediction performance of the ANN results. To judge the prediction
Chapter 3: Artificial neural network (ANN) model development 89
performance of the ANN, several performance measures were used, including
statistical analysis in terms of absolute fraction of variance (R2), root mean squared
(RMS) and maximum average error percentage (MAEP). Formulas to calculate the
error parameters are shown in Equation 3-1 to 3-3.
2
1
2
12
)(1
NI
IMa
pa
Ni
i
EE
EER
Equation 3-1
N
EERMS
Ni
ipa
1
2
Equation 3-2
Ni
i a
pa
E
EE
1
100N
1 MEP
Equation 3-3
Where, Ea-Actual result; Ep-Predicted result; Em-Mean value of terget; N-Number of
sample size
The results of ANN model testing for biodiesel properties are shown in Figure 3-9.
Following statistical analysis, it was found that the absolute fraction of variance (R2)
was close to unity, ranging from 0.8932 to 0.9622, with Root-Mean-Squared (RMS)
errors ranging from 0.011 to 4.171 and maximum average error percentage (MAEP)
ranging from 1.86% to 5.53%. This variation in model performance may be due to
the quantity and quality of data, as well as the complexity of the correlation of input
and output variables. The best estimation accuracy was found for CN and IV, which
might be because these two parameters were correlated with fewer parameters, which
reduced the complexity of the system. The least accurate ANN model was found for
OS, due to highly complex correlation with input variables, and the smaller number
of data sets used during the training process. Overall, it can be seen that the ANN
was able to generate relationship between the measured and predicted fuel properties
of biodiesel. Therefore, the developed ANN models were trained well and can be
used to simulate fuel properties for a wide range of biodiesel fuels. However, the
prediction accuracy of the model should be further improved by increasing the
number and range of the training data sets.
Chapter 3: Artificial neural network (ANN) model development 91
(g) (h)
Figure 3-9: Biodiesel properties estimation accuracy of developed ANN models
(a) Cetane number (CN); (b) Kinematic viscosity (KV); (c) Density; (d) Higher
heating value (HHV); (e) Oxidation stability; (f) Cold filter plugging pint temp (CFPP); (g) Flash point temperature; (h) Iodine number (IN)
3.4 CONCLUSION
The aim of this paper was to investigate the correlation between important fuel
properties of biodiesel, as well as the ability of artificial neural networks (ANN) to
predict the important fuel properties of biodiesel based on its chemical composition.
Experiments were conducted on nine different biodiesel feedstocks, with a wide
range of chemical compositions. The data were further improved by collecting
experimental secondary data collected from over 120 published literatures. The fuel
properties investigated in this study were cetane number, kinematic viscosity,
density, higher heating value, oxidation stability, cold filter plugging pint temp, flash
point temperature and iodine value. Correlation of individual fuel properties and
chemical composition were investigated using principle component analysis and
based on the graphical representation of the results, a complex relationship was
found between chemical composition and biodiesel properties. The most influential
chemical composition parameters, which affected all biodiesel properties in this
study, were the poly-unsaturated fatty fraction and average number of double bonds
present in the biodiesels. Using the data obtained from experiments and the literature,
standard back propagation (BP) neural network models with LM algorithms were
92 Chapter 3: Artificial neural network (ANN) model development
developed for individual biodiesel properties. The parameters obtained from the
chemical compositions of biodiesel where were used as input variables and fuel
properties were used as output variables during training of the network. The input
variables were selected based on the correlation results obtained for individual
biodiesel properties. The performance of the developed ANN prediction models were
tested and it was found that the absolute fraction of variance (R2) was close to unity,
ranging from 0.8932 to 0.9622, with Root-Mean-Squared (RMS) errors ranging from
0.011 to 4.171 and maximum average error percentages (MAEP) ranging from 1.86%
to 5.53%. The results of this study also show that the ANN has the ability to learn
and generalise a wide range of experimental conditions. Therefore, the use of ANNs
may be recommended, in order to optimise the chemical composition of biodiesels,
as well as the fuel quality for internal combustion engine application. However, the
network should be further improved by including additional robust data sets during
the training process.
Chapter 4: Biodiesel from Australian native plants 93
Chapter 4: Biodiesel from Australian native plants
4.1 INTRODUCTION
Australia has vast areas of grazing (cleared) and degraded (mined) land on which
plants for biodiesel feedstock can be grown, including a large number of native
species containing non-edible oil in fruits, which have already been assessed for
growth on degraded land in Australia (Ashwath 2010b). However, for any plant to be
deemed a viable candidate for large scale production as a biodiesel fuel source,
several issues must be considered, including: cultivation requirements, such as the
extent of irrigation or drainage systems; having a tolerance to a range of
environments and soil types; and matching the ecology of available or desired areas;
as well as propagation, planting, weeding, fertilising, trimming/pruning and growth
rates. Aside from environmental and cultivation requirements, the potential oil yield
per year per hectare of land, along with difficulties in the extraction of oil from the
seeds or fruits also need to be considered, as does the quality of biodiesel produced
from these native species. This work investigated the suitability of 11 native
Australian plants for the production of second-generation biodiesel. The plants were
selected by consulting with plant scientist Assoc. Prof. Nanjappa Ashwath, Centre
for Plant and Water Sciences (CPWS), Central Queensland University, and a brief
description of the selected plant is presented below. The seeds of native plants were
collected from local seed suppliers in Queensland, Australia. These seeds were
ground-dried (i.e. had been on the ground for some time prior to collection) and
mostly collected from coastal locations in Northern Queensland.
94 Chapter 4: Biodiesel from Australian native plants
4.2 POTENTIAL NATIVE OIL SEED PLANTS
A large number non-edible vegetable oil crops grow naturally across the vast land
area of Australia, which can be a valuable source for future generation biodiesel
production. Important features of the species investigated in this study are described
in following sections.
4.2.1 Beauty leaf (Calophyllum inophyllum)
Calophyllum Inophyllum, more commonly known as Beauty leaf, is a moderate sized
tree that grows between 8–20 m tall and is most notable for its decorative leaves and
fragrant flowers, as can be seen in Figure 4-1. The tree grows in tropical and sub-
tropical climates (typical temperature range of 18–33oC) close to sea level. Beauty
leaf trees grow in free draining soils near shorelines, however, it has been observed
growing in various clay soils within Australia (Hyland et al. 2003). Beauty leaf can
be transplanted from nurseries very successfully, although for the purpose of
cultivation, it is more beneficial to grow the trees from direct seeding. It is a
moderately fast growing tree that can grow up to 1 m tall within a year. It has also
been seen to flourish in the presence of weeds and other species, so the plant can be
grown in mixed cultures and weeding is not necessary (Mohibbe Azam, Waris and
Nahar 2005a). The tree bears fruit twice a year, the timing of which depends on when
the tree was planted, as it is an evergreen. With two yields per year, a healthy tree
produces around 8,000 fruits which contain a kernel within a hard husk. At a tree
density of 400 trees per hectare, up to 16,000 kg of dry seeds can be produced per
hectare per year (Ashwath 2010b). The fruits can be harvested either directly from
the tree once they are fully matured or once they have fallen off. Given that a number
of animals eat the ripe fruit, it may also be preferable to harvest fruit from trees just
before they fully mature (Friday and Okano 2006; Mohibbe Azam, Waris and Nahar
2005a). The kernels themselves are contained in a hard shell, as seen in Figure 4-1.
Chapter 4: Biodiesel from Australian native plants 95
Figure 4-1: Beauty leaf tree growing along a beach front, and in a park and its
distribution in Australia
4.2.2 Candle nut (Aleurites Moluccana)
Aleurites Moluccana, also known as Candle nut due to the traditional use of its waxy
seeds and kernels as natural candles, grows naturally in both subtropical and tropical
dry or wet forest climates, reaching altitudes of up to 700 m (Okonko et al. 2009).
The tree needs 640-4300 mm of annual rainfall, however mature trees require little to
no maintenance. It begins bearing fruit after about 3-4 years of age. The tree
produces spherical fruits, typically 5-8 cm in diameter with a thick, rough and hard
nut shell, as shown in Figure 4-2 (Atabani et al. 2013). The fruits are usually fully
mature and begin dropping off the tree between March and May. The fruits can be
collected from the ground around the trees after they have fully matured and dropped
off. The kernels are extracted by cracking the hard shells of the seeds once fully
dried. The kernels have a dry weight of about 6 g, which, on a per hectare basis,
yields around 16,800 kg of dried kernels per year (Quintao et al. 2011).
96 Chapter 4: Biodiesel from Australian native plants
Figure 4-2: The tree and kernels of Candle nut
4.2.3 Blue berry lily (Dianella Caerula)
The Blue berry lily (Figure 4-3) is an herbaceous shrub growing about 0.5-1.3 m
height, commonly referred to as the Blue berry lily because of its characteristic
purple berries. The seeds range in size from about 2.9-3.7 mm, with a smooth, black,
shiny texture. The Blue berry lily can be found on the east coast of Australia from
Torres Straight Island to Tasmania and is most common throughout New South
Wales (Ashwath 2010b). The Blue berry lily is easily propagated via direct seeding
in a wide range of environments. This plant is known as one of the hardiest plants,
tolerating droughts, frost, high humidity and heat. For cultivation purposes, a spacing
of 1 x 0.5 m should be used, accommodating a total of 20,000 plants per hectare. In a
year, a typical plant will produce around 0.11 kg of seed. The seeds can be collected
by stripping the stalks on which they grow. The collected fruits should be either dried
or fermented before extracting the seeds. Considering a plant yields 0.11 kg of seed
per year, up to 400 kg of oil can be produced in a hectare (Genever, Grindrod and
Barker 2003).
Chapter 4: Biodiesel from Australian native plants 97
Figure 4-3: Blue berry lily plant and seeds.
4.2.4 Queen palm (Syagrus Romanzoffiana)
Queen palm (Figure 4-4) is widely cultivated as a decorative plant in Queensland and
the Northern Territory. It has the ability to spread rapidly and can grow in a wide
range ecosystems, from coastal sands to heavy creaking clay soil. The tree is a tall
slender palm reaching heights of around 20 m and it is tolerant of subtropical,
tropical and wetlands conditions (Paroissien 2012). This tree is best suited for acidic,
well-drained soils, thus drainage systems are necessary for large scale cultivation.
This tree also has substantial propagation abilities, partly due to its adaptability to a
large range of soil types, tolerance to a range of temperatures, and moderately low
watering requirements of 400-2000 mm per year. With proper conditions and
maintenance, especially watering, sunlight and soil conditions, around 2,000 kg of
dry seeds can be produced per hectare per year, assuming two harvests a year. The
fruits can be collected by cutting off the panicles on which they grow or by collecting
them from the ground once they have fully matured. The fruits are about 2.5-3 cm
long, have a diameter of 1-2 cm and contain a stringy or hairy kernel (Ashwath
2010b).
98 Chapter 4: Biodiesel from Australian native plants
Figure 4-4: Tree and kernels of Queen palm
4.2.5 Castor (Ricinus Communis)
Castor is a moderately sized shrub that grows to 1-4 m tall. Growing mainly in
riparian habitats, such as along water courses and flood plains, this shrub has been
declared a weed due to its rapid propagation and resilience to a number of
unfavourable conditions. This shrub is drought tolerant and can flourish in range of
temperatures, from 7 oC to 28 oC [37]. It has 15-45 long leaves, palmate and 5-12
deep lobes with tooth margins. The leaves come in various colours, such as dark
green, dark reddish purple or bronze, as shown in Figure 4-5 [38]. The shrub grows
best in well-drained, moisture retentive clays or sandy loam, under full sunlight. The
shrub also prefers high temperatures and humidity [37]. A typical healthy shrub can
produce a total of about 0.2 kg of seeds in a year. This results in approximately 1,000
kg of seed per hectare, per year [37]. The fruits are easily harvested by picking them
directly from the shrubs or simply collecting them from the surrounding area. The
seeds are then taken out of the fruits by cracking them open. This can be done almost
immediately after being picked, as they are already dry by the time they are able to
be picked [40].
Chapter 4: Biodiesel from Australian native plants 99
Figure 4-5: Shrub and seeds of Castor
4.2.6 Bidwilli (Brachychiton Bidwilli)
Bidwilli is usually classed as a shrub, but it can grow into a small tree. It grows in dry
rainforests and along various coastal regions, such as the central and southern
Queensland coasts. It produces fruits in a boat-shaped woody follicle containing a
number of hairy seeds, as shown in Figure 4-6 [41]. Bidwilli grows best in sand or
clay soils, requires minimal maintenance and approximately 2,700 plants can be
grown on a hectare of land area [13]. Due to its tolerance to dry conditions and
capability of growing in mixed cultures, Bidwilli holds a wide propagation potential
[42]. A healthy Bidwilli shrub or tree can produce up to 0.6 kg of seeds in a year,
totalling to about 1,600 kg of seeds per hectare [13].
100 Chapter 4: Biodiesel from Australian native plants
Figure 4-6: Bidwilli plant and seeds
4.2.7 Karanja (Pongamia Pinnata)
The Karanja tree prefers humid and sub-tropical environments, with an average
temperature of 27-38 oC and can be grown up to 14-25 m high. The fruits produced
by Karanja grow in clusters and have pod like features, with a though outer shell
which contains a number of small seeds inside, as shown in Figure 4-7 [43]. It can be
easily cultivated by sowing the seeds in most soil types. The seeds require no pre-
treatments and germinate anywhere between seven days and one month after sowing.
The fruits are best collected directly from the tree once it has matured. The bunches
of fruit are easily collected by cutting the main stems connecting then to the branches
of the tree. The fruits are then dried so that the pods can be cracked open easily,
revealing the seeds. Mature trees can yield 8-24 kg of seed per year [43]. Per hectare,
a total of 3,200-3,600 kg of seeds can be harvested per year.
Chapter 4: Biodiesel from Australian native plants 101
Figure 4-7: Karanja fruit and seeds
4.2.8 Whitewood (Atalaya Hemiglauca)
Whitewood, as shown in Figure 4-8, is a small tree that can grow up to 6 m tall. It is a
very drought tolerant tree growing throughout the open plains and alluvial flats of the
central regions of Australia [44]. Whitewood trees flower from spring until early
summer and produce a large quantity of fruits containing two or three seeds in each
pod [44]. Due to the ease with which it can be cultivated, coupled with its tolerance
to drought and light frost, this fairly small tree can also be transplanted from
nurseries, occupying 3 x 2 m2, allowing approximately 1,660 trees per hectare [13].
In a year, a healthy tree can produce about 2 kg of seed, which results in 3,320 kg of
seed per hectare, per year [13].
Figure 4-8: Tree and fruit of Whitewood
102 Chapter 4: Biodiesel from Australian native plants
4.2.9 Cordyline (Cordyline Manners – Suttoniae)
Cordyline (Figure 4-9) is a small tree that resembles the shape of a palm tree. It is an
evergreen tree that typically grows in rainforests, but has also been seen in various
other forest climates. The tree grows near swamps or other areas with poorly drained
soils. The tree usually grows to about 2-5 m in height [45], growing best in well
shaded areas with a plentiful supply of water. It is easy to identify when this tree
does not have sufficient shading, as the leaves quickly scorch under direct sunlight.
The tree can however, withstand dry conditions, so long as irrigation systems are
utilised appropriately. Cordylines continue to flourish in the presence of other
species and can be grown with mixed species as an understory plant [45]. The fruits
can be collected by cutting off the panicles on which they grow and the seeds can be
extracted with relative ease as the fruits are fairly soft. On a per hectare basis, around
3,000 kg of seeds can be produced by Cordyline trees [13].
Figure 4-9: The tree and fruit of Cordyline
4.2.10 Flame tree (Brachychiton Acerifolius)
Flame trees (Figure 4-10), a medium sized tree reaching a height of 30-35 m, are
widely spread through subtropical rainforests, from northern New South Wales to all
throughout Queensland. The tree produces spectacular floral displays after hot and
dry periods [46]. It is a hardy plant that can grow in a range of soils, it thrives in
temperate to tropical climates and it is able to tolerant both dry conditions and heavy
Chapter 4: Biodiesel from Australian native plants 103
rainfall. It can cope with droughts, however, it does not deal well with colder
temperatures, or harsh or salty winds [47]. The Flame tree flowers in late spring to
early summer and produces fruits with a capsular shape and leathery texture. These
fruits are small and contain 5-8 seeds. These seeds are extracted from the fruit by
allowing the fruit to dry and cracking open the outer shell. Care should be taken
when collecting the seeds as they can cause irritation. The optimal picking period for
the fruits is during the period when the fruits turn a coppery brown colour.
Figure 4-10: Flame tree, fruit and seeds
4.2.11 Chinese rain (Koelreuteria Formosana)
The Chinese rain tree, shown in Figure 4-11, is a moderately sized tree, able to
tolerate a range of conditions, allowing it to be grown in a range of ecosystems. Such
conditions range from frost to high heat, and from well-drained to wet soils [48]. It
can be grown from direct seeding or transplanting from nurseries. This tree can be
grown up to 18 m tall with a little maintenance. Chinese rain trees are able to survive
under a fairly wide range of conditions, however it thrives in warm climates under
full sunlight. Also, this tree can thrive on a number of different soil varieties, but
ideally, the soil should be free-draining [48]. The fruits of the Chinese rain tree grow
from February to March and mature by mid-year, yielding about 2,000 kg per year
per hectare. The fruit grows in drooping clusters and grows to about 50 mm in length
[48]. The seeds are spherical in shape and have a diameter of around 5 mm.
104 Chapter 4: Biodiesel from Australian native plants
Figure 4-11: Chinese rain tree and fruits
4.3 SEED COLLECTION AND PREPARATION
Seed preparation is critical in optimising the oil extraction process, because physical
conditions such as size, hardness and dryness of the seeds and kernels varies
significantly from one species to another. Several steps are involved including:
kernel extraction, kernel grinding and drying.
4.3.1 Kernel extraction
Dry seeds were cracked open to expose and obtain the oil bearing kernels. In order to
reduce kernel damage and oil loss, seed cracking was done with care using a mallet.
In this process, a handful of seeds were placed on a table surface and cracked
individually, as shown in Figure 4-12. During the seed cracking process, it was found
that rubber-headed mallets were the least preferred compared to wooden or steel-
headed mallets, as they tended to rebound excessively. In order to maximise yield of
oil through the extraction process, the individual kernels had to be separated from the
seeds and crushed. This was done to ensure that the solvent used during the chemical
extraction process would be as efficient as possible in extracting the oil. In this study,
dry seeds were cracked manually using a mallet. Depending on the species of seeds,
various levels of difficulty were found when opening the seeds. For example, Candle
nut, Blue berry lily, Beauty leaf, Karanja, Castor and Chinese rain were relatively
easy to process due to their fruit and seed structure. Other seeds, such as Queen
plam, Whitewood and Cordyline, contained a very hard shell around the kernel which
Chapter 4: Biodiesel from Australian native plants 105
made those seeds relatively difficult to process. Moreover, special care was needed
when handling the Bidwilli and Flame tree seeds. This is because those seeds
required some effort to crack open and the kernels inside were covered with tiny
hairs that can cause severe irritation to the skin. Latex gloves, safety goggles and
covered overalls were worn to prevent skin contact with these kernels.
Figure 4-12: Kernel extraction
4.3.2 Kernel grinding
The dried seed kernel samples were ground using a blender and coffee grinder to
obtain a fine consistency powder, in order to maximise particle surface area of the
kernels for exposure to the chemical solvent during the extraction process. After that,
the seeds were removed from the oven and placed in zip locked bags in a refrigerated
store room.
4.3.3 Kernel drying
The extracted kernels seeds naturally contained high levels of moisture, which
needed to be removed for effective oil extraction. After being ground into fine
particles, the kernels needed to be dried in order to reduce their moisture content.
The kernels are placed in an aluminium container (Figure 4-13) and left to dry in the
laboratory oven for 9 days at 70ᵒC. Each day, the kernels were weighed in order to
determine the amount of moisture which had evaporated. Each of the samples were
106 Chapter 4: Biodiesel from Australian native plants
then stirred before returned to the oven, in order to ensure that the kernels dried
evenly.
Figure 4-13: Ground kernels
Based on the weights recorded in the table below, there was a gradual decrease in the
weight of the kernels, with the greatest decreases observed during the first three days
of drying. Once the weight of the kernels displayed no significant changes, they were
considered to be dried. They were then stored in air tight zip lock bags, in order to
maintain their dryness.
4.4 OIL EXTRACTION
Oils from the processed and dried seed kernels were extracted chemically using n-
hexane as a solvent. For this purpose, an accelerated solvent extraction (Dionex™
ASE 350®) machine, shown in Figure 4-14, was used to extract oil at high pressure
and temperature using an accelerated solvent extraction method. The oil extractor
comes with an automated extraction control system that uses elevated temperatures
Chapter 4: Biodiesel from Australian native plants 107
and pressures to achieve extractions in a short period of time. Measured samples
were inserted into metal sample cells and the desired operating conditions were set
using the control interface. In order to observe the effect of temperature in oil yield,
the extraction process was conducted at the three different temperatures: 50 °C,
100°C and 150 °C. Due to the limitation of the equipment, oil extraction at over
150°C was not possible. N-hexane was pumped into the cell with pressurised
nitrogen gas to achieve a pressure of 1,600 psi. After the extraction process was
completed, all of the extracted oil samples were collected into the vessels in the
collection tray (Figure 4-14). The solvent was separated from the extracted sample
using the Dionex™ SE® 400 solvent evaporator system as shown in Figure 4-15.
Figure 4-14: ASE 350 cell loading process
108 Chapter 4: Biodiesel from Australian native plants
Figure 4-15: n-hexan removing using DionexTM SETM 400
The percentage of oil obtained from the seed kernels are shown graphically in the
Figure 4-16. Overall, a wide range of oil yields were found and of the eleven native
species investigated in this study, Beauty leaf had the highest oil yield, followed by
the Queen palm, Karanja and Candle nut. However, most of the tested oilseeds
contained more than 30% oil in their kernel. The Chinese rain oil seed showed
lowest oil yield, of about 5%, followed by Cordyline (about 18%). It is also
interesting to see from Figure 4-16 that extraction temperatures had a significant
influence on the oil yield of seed kernels. In general, the oil yield increased with
increasing temperature. This is because the thermal energy of the solvent increases
with increases in temperature, which helps to overcome cohesive and adhesive
interactions. Moreover, higher temperatures increase the molecular motion of
molecules and decrease hydrogen bond interactions. However, this effect varied from
one species to another and may be due to differences in the physical condition of the
kernels. There were also variations in the colour of oils extracted in this study, which
is evident in Figure 4-17.
Chapter 4: Biodiesel from Australian native plants 109
Figure 4-16: Oil yield of native plant seed kernels
Figure 4-17: Extracted bio-oil sample from native plants
A: Beauty leaf; B: Candle nut; C: Karanja; D: Queen palm; E: Blue berry lily; F:
Castor; G: Whitewood; H: Bidwilli; I: Chinese rain; J: Flame tree; K: Cordyline
4.5 CHEMICAL COMPOSITION
A systematic analysis of chemical composition and comparative fuel properties is
very important for selecting appropriate feedstock for biodiesel production. As
discussed in Chapter 3, chemical composition is a key factor in determining the
quality of biodiesel and the chemical composition of the extracted oil samples was
determined in terms of fatty acid profile and the percentage free fatty acid content
BAC D E F G H I J K
110 Chapter 4: Biodiesel from Australian native plants
(FFA). The fatty acid profiles were analysed by gas chromatography and flame
ionisation detection (GC-FID), in accordance with EN 14103 standards. The gas
chromatograph (GC) was a Hewlett-Packard 6890 System fitted with Varian Select TM 30 m × 0.32 mm × 0.25 µm column. The chemical composition of the tested bio-
oils are presented in Table 4-1, where it can be seen that the chemical compositions
of bio-oil obtained from native oil seeds were similar to those for conventional edible
oil and with the exception of Queen palm and Castor, they were mostly rich in
triglycerides of Oleic (C18:1), followed by Stearic (C18:0), Linoleic (C18:2),
Palmitic (C16:0) and Linolenic (C18:3) fatty acids. Queen palm bio-oil was mostly
comprised of shorts chain fatty acids, including 42.05 wt.% Luatic acid (C12:0) and
10.45 wt.% Myristic acid (C14:0), as well as small amounts of Caprilic (C8:0) and
Capric (C10:0) acids, which were not found in any other tested bio-oil. These fatty
acids not only consisted of a short carbon chain but they were also saturated fatty
acids, which explains why Queen palm showed the lowest average chain length
(ACL) account and highest percentages of saturated fatty acids among the tested bio-
oil samples. Likewise, Bidwilli bio-oil was rich in saturated fatty acids, containing
42.39% C16:0 and 14.37% C18:0 fatty acids. On the other hand, the highest ACL
and lowest saturated fatty acid content was found for Castor bio-oil, comprising 95
wt.% long chain length fatty acids and mono-unsaturated Gondonic (C20:1) acid,
with small amounts of C18:2 and C18:1 fatty acids. The only other oil to contain a
significant amount of Gondonic (C20:1) acid was Whitewood, which accounted for
25.04% by weight. Due to its high Linolenic (C18:3) fatty acid content, Candle nut
oil showed the highest level of poly-unsaturated fatty acids (PUFA), followed by
Cordilyne, Chinese rain and Blue berry lily oils. In contrast, very small amounts of
PUFA’s were found with Whitewood, Castor and Queen palm bio-oil. As reported in
Chapter 3, the higher the unsaturation of biodiesel, the greater the tendency for the
biodiesel to oxidise. On the other hand, unsaturated fatty acids had a positive
influence on other fuel properties, such as cold filter plugging point temperature.
Therefore, both saturated and unsaturated FAMEs have a role to play in finding the
optimal balance for high quality biodiesel.
FFA’s have a significant effect on biodiesel processing from vegetable oil, as
discussed in Chapter 2. In this study, FFA content of the native plant oil was
analysed using a D5555-95 (2011) standard test method and the results are shows in
Table 4-1. Although the literature (Dorado et al. 2002; Lam, Lee and Mohamed
2010; Kumar Tiwari, Kumar and Raheman 2007; Ramadhas, Jayaraj and
Muraleedharan 2005) suggested that the FFA content of bio-oil should be below 5%
for alkali-trans-esterification, most of the bio-oil tested in this study contained much
Chapter 4: Biodiesel from Australian native plants 111
higher FFA’s. These results indicate that FFA can be one of the issues impeding the
success of biodiesel production from native species. In particular, Flame tree oil
contained 36.7% FFA’s, which was the highest among the native bio-oils, followed
by Beauty leaf (22%), Queen palm (15%) and Blue berry lily (13.1%). The lowest
FFA content was found in Chinese rain oil, which consisted of 1.8%. In addition,
Queen palm contained an exceptionally higher amount of oxygen, which accounted
for 14.19% on a per weight basis. The oxygen content of the other methyl esters
range d from 10.25 to 11.82%, as shown in Table 4-1.
Chapter 4: Biodiesel from Australian native plants 112
Table 4-1: Fatty acid profile and chemical composition of bio-oil produced from native plants
Fatty acid profile
Beauty leaf
Candle nut
Karanja Queen palm
Blue berry lily
Castor oil
Whitewood Bidwilli Chinese rain
Flame Cordyline
C8:0 (wt. %) 0.00 0.00 0.00 2.50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 C10:0 (wt. %) 0.00 0.00 0.00 3.47 0.00 0.00 0.00 0.00 0.00 0.00 0.00 C12:0 (wt. %) 0.00 0.00 0.00 42.05 0.00 0.00 0.00 0.00 0.00 0.00 0.00 C14:0 (wt. %) 0.00 0.00 0.00 10.45 0.00 0.00 0.00 0.00 0.00 0.00 0.00 C16:0 (wt. %) 13.66 5.50 10.14 8.23 12.90 0.00 3.99 42.39 18.15 17.48 7.07 C16:1 (wt. %) 0.24 00 00 0.12 00 00 00 00 00 00 00 C18:0 (wt. %) 16.55 6.70 8.88 1.72 4.83 0.83 1.49 14.37 5.84 3.50 3.09 C18:1 (wt. %) 42.48 10.50 66.18 29.01 29.12 2.05 65.73 24.68 20.08 52.40 21.22 C18:2 (wt. %) 25.56 48.50 12.48 2.57 53.16 2.11 1.13 18.56 55.07 23.95 68.61 C18:3 (wt. %) 0.00 28.50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.68 0.00 C20:1 (wt. %) 0.00 0.00 0.39 0.00 0.00 95.00 25.04 0.00 0.00 0.00 0.00 C22:1 (wt. %) 0.00 0.00 0.00 0.00 0.00 0.00 1.28 0.00 0.00 0.00 0.00 FFA (wt. %) 22 12.3 6.3 15.2 13.1 8.4 4.5 8.2 1.8 36.7 4.6 O2 (wt. %) 11.32 11.44 11.22 14.19 11.51 10.35 10.94 11.82 11.48 11.55 11.46 H2 (wt. %) 11.95 11.49 11.92 12.14 11.89 12.32 12.01 12.29 11.80 12.05 11.75 C (wt. %) 75.22 76.77 74.93 73.79 76.61 77.32 75.71 75.89 75.87 76.41 76.78 Sat. (wt. %) 27.83 12.20 20.95 65.92 17.72 0.83 12.82 56.76 24.85 20.98 10.17 MUFA(wt. %) 47.49 10.50 66.57 29.01 29.12 97.05 86.05 24.68 20.08 52.40 21.22 PUFA(wt.%) 24.68 77.00 12.48 2.57 53.16 2.11 1.13 18.56 55.07 26.63 68.61 ACL 17.76 17.84 17.87 14.37 17.74 19.90 18.52 17.15 17.65 17.65 17.86 ANDB 0.97 1.64 0.91 0.34 1.35 1.01 0.88 0.61 1.30 1.05 1.58
Chapter 4: Biodiesel from Australian native plants 113
4.6 FUEL PROPERTIES
Based on the chemical composition of the bio-oil, the corresponding biodiesel
properties were estimated using the ANN model developed in Chapter 3. The
estimated biodiesel properties were: cetane number (CN), kinematic viscosity (KV),
density, higher heating value (HHV), oxidation stability (OS), cold-filter plugging
point temperature (CFPP), flash point temperature (FP) and iodine value (IV). Table
4-2 shows the results of estimated biodiesel properties and corresponding values of
AU, US and EU biodiesel standards. Estimation results indicated that the biodiesel
from most of the native plants met the biodiesel standards for all property indicators,
except oxidation stability.
Cetane number (CN), which is a measurement of the combustion quality of diesel
fuels during compression ignition, is a significant parameter for indicating fuel
quality. CN is associated with the ignition delay time of a fuel, which is the time that
passes between injection of the fuel into the cylinder and the onset of ignition. A
shorter ID time results in a higher CN and vice versa (Gopinath, Puhan and
Nagarajan 2009). According to AU and EU standards for biodiesel, the minimum CN
should be 51, whereas the US standards set a minimum value of 47. Most of the
eleven native species investigated in this study met the minimum requirements for
CN, except for Candle nut and Cordyline seed oils, which is likely to be a result of
their high PUFA concentration. For the same reason, the IV of Candle nut and
Cordyline were found to be very high, exceeding the recommended maximum value
in the EU biodiesel standard. Moreover, biodiesel must have an appropriate KV,
which plays a dominant role in fuel spray, fuel-air mixture formation and the
combustion process (Ramírez-Verduzco, Rodríguez-Rodríguez and Jaramillo-Jacob
2012). The KV of all native plant species listed in Table 4-2 where within the range
recommended by US standards, which was 1.9 – 6.0 cSt. However, the KV values
for Queen palm biodiesel were slightly below the minimum limit of AU and EU
standards. Another important indicator of biodiesel properties is density, which
determines the amount of fuel injected in the cylinder. Changes in fuel density will
influence the stoicheometric ratio of air and fuel, and consequently the engine output
114 Chapter 4: Biodiesel from Australian native plants
power and exhaust emissions (Knothe 2009). Therefore, AU and EU biodiesel
standards are set at 0.86 – 0.90 kg/l for biodiesel density, and all biodiesel from the
native species investigated in this study were within the standard range. The HHV,
which indicates the energy content of the fuel, was also within the standard range
(39.87 - 40.12 MJ/kg) for regular biodiesel, which is normally 10% to 12% less than
that obtained for petroleum-derived diesel (46MJ/kg) (Ramírez-Verduzco,
Rodríguez-Rodríguez and Jaramillo-Jacob 2012). Oxidation stability is another
important parameter, which indicates the resistance to oxidation during long-term
storage. Usually biodiesel quality declines due to gum formation during the oxidation
process. However, most of the biodiesel investigated in this study performed poorly
in terms of OS. Queen palm biodiesel was the only sample that fulfilled all biodiesel
standards in terms of OS, due to its very high saturated fatty acid content. On the
other hand, the oxygen stability of Candle nut, Blue berry lily, Chinese rain and
Cordyline oil biodiesel were below the specified limits for all biodiesel standards.
However, the US biodiesel standard imposed less restriction in OS, which is set at a
minimum of 3 hours, and therefore, the biodiesel from Beauty leaf, Karenja, Castor,
Whitewood, Bidwilli and Flame tree oils comfortably met this standard. CFPP is a
critical factor in determining biodiesel quality, especially for cold climate conditions.
High CFPP values were found for the biodiesel from Queen palm, Bidwilli, Beauty
leaf and Whitewood, which is acceptable because the chemical composition of that
feedstock contained a high portion of saturated fatty acids. Flash point is another
important parameter for assessing biodiesel quality, indicating the fire hazards during
fuel transport and storage. It is the lowest temperature at which the fuel will start to
vaporise to form an ignitable mixture when it comes into contact with air (Ali, Hanna
and Cuppett 1995a). This is reflected by the respective limits within the AU, EU and
US biodiesel standards shown in Table 4-2. All of the investigated biodiesel listed in
Table 4-2 show good quality in terms of FP. The lowest FP was found for Beauty
leaf biodiesel, which was well above the recommended minimum value in all
biodiesel standards.
Chapter 4: Biodiesel from Australian native plants 115
Table 4-2: Estimated biodiesel properties
Properties Unit
Biodiesel standard
Bea
uty
leaf
Can
dle
nu
t
Kar
anja
Qu
een
pal
m
Blu
e be
rry
li
ly
Cas
tor
oil
Wh
ite-
woo
d
Bid
wil
li
Ch
ines
e ra
in
Fla
me
Cor
dyli
ne
AUS US EU
CN - 51, min 47, min 51, min 64.12 38.42 59.38 62.28 51.79 59.84 59.18 64.55 50.43 56.64 45.78
KV cSt. 3.5-5 1.9-6 3.5-5 4.39 3.79 4.66 3.38 4.24 5.81 5.15 4.61 4.26 4.48 4.15
Density Kg/l 0.86-0.90 n/a 0.86-0.90 0.878 0.883 0.874 0.871 0.879 0.873 0.873 0.871 0.879 0.876 0.882
HHV Mj/kg n/a n/a n/a 40.12 39.87 40.01 39.39 39.97 40.10 40.04 39.95 39.97 39.98 39.97 OS hours 6, min 3, min 6, min 4.34 1.45 4.26 6.46 2.57 3.88 4.38 5.46 2.68 3.67 1.61 CFPP °C Report Report Report 12.54 -4.22 8.52 7.19 -4.85 -15.17 12.06 14.91 1.16 -5.09 -9.46 FP °C 120, min 93, min 120, min 142.78 157.57 148.23 185.44 162.43 201.87 164.74 163.28 165.61 160.92 163.95
IV g iod/ 100g
n/a n/a 120, max 79.25 164.77 77.51 29.86 114.32 79.75 74.04 53.16 112.19 93.17 131.25
Chapter 4: Biodiesel from Australian native plants 116
4.7 EVALUATION OF NATIVE PLANT METHYLE ESTER
In order to constitute an ideal source of sustainable biodiesel, the feedstock should
contain a sufficient amount of bio-oil, with a suitable chemical composition, in order
to elicit good fuel quality properties. However, selection of the most suitable
feedstock for industrial production is a multi-criteria problem, as it involves multiple
quality indicators. In this study, a multi-criteria decision method (MCDM) software
PORMETHEE-GAIA was used for the selection of biodiesel for large scale
production. The most suitable native plant species were selected from the eleven
native species investigated in this study, based on the parameters listed in Table 4-3.
The GIGA plan displays how the alternatives perform in terms of the different
criteria, as shown in Figure 4-18. The length of the criteria vectors and their
directions indicate the influence these criteria have on the decision vector (red line in
Figure 4-18) and preference of the species. The preference functions of criteria were
modelled as Min (i.e. lower values are preferred for good biodiesel) or Max (higher
values are preferred for good biodiesel), as per Table 4-3. When the criteria are
oriented in an opposite direction they are in conflict and when they are oriented in a
similar direction they express the same preference (Espinasse, Picolet and Chouraqui
1997). For example, maximum values for oil yield (OY), IV and OS were preferable
for good biodiesel and therefore, those criteria lines are in same direction of decision
vector. The preference functions were obtained by principle component analysis
(PCA) techniques (Brans 2002), however preference function selection also
influenced the orientation of criteria (which was also suggested in (Islam et al.
2013)). For example, IV and OS were inversely related but still showing in the same
direction within ±45○. This is because OS was preferred to maximum, but iodine
number was preferred to minimum, as shown in Table 4-4. Therefore, criteria which
are in the same preference (min/max) and lie close to ±45° are correlated. The
decision vector, which is marked as red line in Figure 4-18, is the direction in which
the decision maker is invited to decide. The direction and length of criteria are
indicative of their influence on the decision vector (Islam et al. 2013), such that the
very short length of some criteria (i.e. difficulty level (DL) and HHV) indicate the
little effect they had on the decision vector. However, the freedom of decision vector
Chapter 4: Biodiesel from Australian native plants 117
is modelled by the preference weight of individual criteria and therefore, if the
weights are modified, the decision maker is invited to decide in another direction
(when the position of the criteria and alternative remain unchanged) (Brans 2002).
The Phi value is the net flow score, which could be negative or positive depending
upon the angular distance from the decision vector and the distance from the centre.
Figure 4-18 shows the raking results of eleven native species biodiesels with its
corresponding phi value for the equally weighted criteria and preferred function
listed in Table 4-3. Results showed that Beauty leaf was most aligned with the
decision vector and in the farthest relative position from the centre, giving it the
highest of ranking followed, by Queen plam, Castor and Karenja. On the other hand,
biodiesel from the Flame tree was at the bottom of the ranking.
Table 4-3: Variables and preference used in PROMETHEE-GAIA analysis
No Variables Preference for PROMETHEE-GAIA
1 Seeds production per year per hector Max 2 Difficulty level of seed processing Min 3 Oil yield Max 4 Free fatty acid content in oil Min 5 Kinematic viscosity Min 6 Density Min 7 Higher heating value Max 8 Oxidation stability Max 9 Iodine value Min 10 Cetane number Max 11 Flash point temperature Max 12 Cold filter plug point Max
118 Chapter 4: Biodiesel from Australian native plants
Figure 4-18: (a) Graphical Analysis for Interactive Assistance (GAIA) plot for eight
biodiesel showing 11 criteria and decision vector. (b) Corresponding complete
ranking and Phi value of biodiesel from native plants
The quality ranking analyses of biodiesel shown in the previous section was
conducted with an equal weighting of all parameters. However, for industrial and
economical biodiesel production, the oil content of the feedstock may be a more
important criterion than others. Moreover, the importance of some fuel properties
depends on the country and place where it will be used and stored. For an example,
in tropical/sub-tropical regions, CFPP was not considered to be of importance,
however elevated temperatures in these regions are likely to affect the OS of
biodiesel. On the other hand, in colder climate conditions, CFPP are more important
than OS. In this study, ranking sensitivity analysis was conducted for the criteria OY,
CFPP and OS, by increasing the weighting from 1 (equal to other parameters) to 10,
and the results are shown in the Table 4-4 to 4-6. As shown in Table 4-4, Beauty leaf
was always ranked number 1 with the increasing OY weighting, whereas the rank of
Candle nut and Flame tree gradually improved from 7 to 2 and 11 to 6, respectively.
Rank Biodiesel Phi
1 Beauty leaf 0.1917
2 Queen palm 0.1500
3 Castor 0.1333
4 Karanja 0.0917
5 Whitewood 0.0583
6 Chinese rain
0.0083
7 Candle nut -0.0083
8 Cordyline -0.0500
9 Bidwilli -0.1500
10 Blue berry lily
-0.2000
11 Flame tree -0.2250
(a) (b)
Chapter 4: Biodiesel from Australian native plants 119
With a weighting of 10 for OY, Queen palm biodiesel dropped down from a ranking
of 2 to 3, however it increased to first position when the weighting of OS was
increased to 2 and it remained there as the weighting increased (see Table 4-5). The
largest improvement in ranking was observed for Bidwilli biodiesel, which increased
in rank from 9 to 4 when the weighting of OS was increased to 6. Whitewood also
showed a significant improvement in rank (from 5 to 2) for a heavier weighting of
OS. The rank of Flame tree also improved from 11 to 8 following the same increase
in weighting.
Table 4-4: Comparative rank shift with different weighting for bio-oil yield
feedstock Weighting
2 4 6 8 10
Beauty leaf 1 ‐ 1 ‐ 1 ‐ 1 ‐ 1 ‐ Queen palm 2 ‐ 2 ‐ 2 ‐ 2 ‐ 3
Castor 3 ‐ 4 5 5 ‐ 5 ‐ Karanja 4 ‐ 5 4 4 ‐ 4 ‐ Whitewood 6 6 ‐ 6 ‐ 6 ‐ 7
Chinese rain 7 10 10 ‐ 10 ‐ 10 ‐ Candle nut 5 3 3 ‐ 3 ‐ 2
Cordyline 8 ‐ 7 9 9 ‐ 9 ‐
Bidwilli 9 ‐ 8 8 ‐ 8 ‐ 8 ‐
Blue berry lily 11 11 ‐ 11 ‐ 11 ‐ 11 ‐
Flame tree 10 ‐ 9 7 7 ‐ 6
Table 4-5: Comparative rank shift with different oxidation stability of biodiesel
Feedstock Weighting
2 4 6 8 10
Beauty leaf 2 2 ‐ 2 ‐ 2 ‐ 3
Queen palm 1 1 ‐ 1 ‐ 1 ‐ 1 ‐ Castor 3 ‐ 5 6 6 ‐ 6 ‐ Karanja 4 ‐ 4 ‐ 5 5 ‐ 5 ‐ Whitewood 5 ‐ 3 3 ‐ 3 ‐ 2
Chinese rain 6 ‐ 7 7 ‐ 7 ‐ 7 ‐ Candle nut 7 ‐ 8 9 9 ‐ 9 ‐ Cordyline 9 10 11 ‐ 11 ‐ 11 ‐
Bidwill 8 6 4 4 ‐ 4 ‐
Blue berry lily 11 11 ‐ 10 ‐ 10 ‐ 10 ‐
Flame tree 10 9 8 8 ‐ 8 ‐
120 Chapter 4: Biodiesel from Australian native plants
As shown in Table 4-6, a massive change in raking was observed with the heavier
weighting of CFPP. Due to their high saturated and monounsaturated fatty acid
content, the ranking of Beauty leaf and Queen palm biodiesel dropped dramatically
from 1 to 10 and 2 to 7, respectively. In addition, the ranking of Karanja, Whitewood
and Bidwilli also decreased with the heavier weighting of CFPP. In contrast, the
ranking of Castor, Cordyline, Bidwilli, Flame tree, Candle nut and Blue berry lily
improved significantly with a weighting of 10 for CFPP, being 1-5, respectively (see
Table 4-6).
Table 4-6: Comparative Rank shift with different cold filter plugging point temperature
Feedstock Weighting
2 4 6 8 10
Beauty leaf 2 6 9 9 ‐ 10
Queen palm 3 3 4 5 7
Castor 1 1 ‐ 1 ‐ 1 ‐ 1 ‐ Karanja 4 ‐ 7 8 8 ‐ 8 ‐ Whitewood 5 ‐ 9 10 10 ‐ 9
Chinese rain 7 5 6 7 6
Candle nut 8 4 3 4 4
Cordyline 6 2 2 ‐ 2 ‐ 2 ‐
Bidwilli 11 11 ‐ 11 ‐ 11 ‐ 11 ‐
Blue berry lily 9 10 7 6 5
Flame tree 10 8 5 3 3 ‐
4.8 CONCLUSION
The need for a new source of renewable energy is growing ever more critical and
together with the issue of food vs. fuel, it is essential that second-generation biodiesel
feedstocks do not compete with food crops or lead to land-clearing, whilst still
providing the same benefits as first-generation biodiesel (i.e. greenhouse-gas
reductions). This chapter reported on the investigation of eleven species of native
Australian plants, in order to discover their viability for serving as an alternative
feedstock for biodiesel. Dry seeds of selected plants were collected from local
Chapter 4: Biodiesel from Australian native plants 121
sources in Queensland and processed for bio-oil extraction. DionexTM ASETM 350
Accelerated Solvent Extractor was used for this purpose and the respective oil yields
were determined in terms of oil percentage. The free fatty acid content of the
collected bio-oils were determined, before they were converted into biodiesel using
methyl alcohol. The chemical composition of biodiesel from native plants were
determined and important fuel properties were estimated using the ANN model
described in the previous chapter. Based on their dry seed production capability,
level of seed processing difficulties, bio-oil content in the seed kernel, amount of free
fatty acids and estimated fuel properties of biodiesel, the native species feedstock
was the evaluated and compared using the multi-criteria decision method (MCDM)
software PORMETHEE-GAIA. In addition, sensitivity analysis of native plant
ranking was investigated by changing the weighting of three important criteria - OY,
OS and CFPP. Overall, this study found that Beauty leaf, Queen palm, Castor and
Karanja were the top ranked candidates for biodiesel production. Based on the
variation of weighting for certain criteria, Beauty leaf and Queen palm biodiesel were
found to be a good choice for second-generation biodiesel production in tropical/sub-
tropical regions, however the opposite was true for cold weather conditions. For cold
climate conditions, Castor, Cordyline or Flame tree might be a better choice than
Beauty leaf and Queen palm.
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 123
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
Physico-chemical Assessment of Beauty leaf (Calophyllum
Inophyllum) as Second-Generation Biodiesel Feedstock
Jahirul M. I*, Brown R. J, Senadeera W, Ashwath N, Rasul M. G, Rahman M. M,
Muhammad A. I and O’Hara I. M
Publication: Submitted to the journal of Energy Conversion and Management
Author Contribution
Contributor Statement of Contribution
Jahirul M. I Conducted the experiments, performed the data analysis and drafted the manuscript Signature
Brown R. J Supervised the project, aided with the data analysis, development of the paper and extensively revised the manuscript
Senadeera W Supervised the project, aided with the development of the paper
Ashwath N Assisted with conducting the experiment
Rasul M. G Assisted with conducting the experiment and revised manuscript
Rahman M. M Assisted with conducting the experiment
Muhammad A. I Assisted data analysis
Ian O’Hara Supervised the project and revised the manuscript
Principal Supervisor Confirmation
I have sighted email or other correspondence from all co-authors confirming
their certifying authorship.
Name
Dr Wijitha Senadeera
Signature
Date
124 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
Abstract
Recently, second-generation (non-vegetable oil) feedstocks for biodiesel production
are receiving significant attention due to the cost and social impacts associated with
using food products for the production of energy products. The Beauty leaf tree
(Calophyllum Inophyllum) is a potential source of non-edible oil for producing
biodiesel because of its suitability for production in a wide range of climate
conditions, easy cultivation, high fruit production rate, and the high oil content in the
seed. In this study, oil was extracted from Beauty leaf tree seeds through three
different oil extraction methods. The important physical and chemical properties of
these extracted Beauty leaf oils were experimentally analysed and compared with
other commercially available vegetable oils. Biodiesel was produced using a two-
stage esterification process consisting of an acid catalysed pre-esterification process
and an alkali catalysed transesterification process. Fatty acid methyl ester (FAME)
profiles and physico-chemical properties including kinematic viscosity, density,
higher heating value and acid value were measured. Other fuel properties including
oxidation stability, iodine value, cetane number, flash point and cold filter plugging
point temperature were estimated using ANN models based on the FAME analysis.
Physico-chemical properties of Beauty leaf oil biodiesel are described and compared
with biodiesel standards and commercially available biodiesels produced from other
feedstocks. The quality of Beauty leaf biodiesel has been assessed based on 10
important chemical and physical properties through a Preference Ranking
Organisation Method for Enrichment Evaluation (PROMETHEE) and Graphical
Analysis for Interactive Assistance (GAIA) analysis. The results show that
mechanical extraction using a screw press produces oil at a low cost, however,
results in low oil yields. The study found that seed preparation has a significant
impact on oil yields, especially in the mechanical oil extraction method. High
temperature and pressure in the extraction process increases the oil extraction
performance. On the contrary, this process increases the free fatty acid content in the
oil. A clear difference was found in the physical properties of Beauty leaf oils, which
eventually affected the oil to biodiesel conversion process. However, Beauty leaf oils
methyl esters (biodiesel) were very consistent physico-chemical properties and able
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 125
to meet almost all indicators of biodiesel standards. Results of this study indicated
that Beauty leaf is a suitable feedstock for commercial production of second-
generation biodiesel. Therefore, the findings of this study are expected to serve as the
basis for further development of Beauty leaf as a feedstock for industrial scale
biodiesel production.
Key words: Beauty leaf, second-generation biodiesel, oil extraction, physico-
chemical properties, PROMETHEE-GAIA
5.1 INTRODUCTION
Rapid growth in population, urbanisation and energy demand, together with the
depletion of conventional fossil fuel reserves and degradation in air quality are
continuously motivating researchers to find more sustainable and cleaner energy
sources. As a consequence, biodiesels produced from vegetable oil and animal fat
feedstocks are receiving significant attention as an alternative to fossil-based diesel.
The first recorded production of biodiesel occurred in 1937 when fatty acid methyl
esters from palm oil were produced via trans-esterification. Biodiesel research
continued from this time but its potential was not fully realised until the 1970s
energy crisis when interest in alternative fuels was renewed (Lim and Teong 2010).
Since this time, biodiesel has been produced on an industrial scale and a multitude of
feedstocks have been assessed. It is generally held that biofuels offer many benefits
over fossil-based fuels, including ability to produce from regionally available
biomass sources, lower greenhouse gas emissions, enhanced biodegradability and
enhanced sustainability characteristics (Reijnders 2006; Ellabban, Abu-Rub and
Blaabjerg 2014). Biodiesel typically contains oxygen levels of 10–45% by mass
while fossil-based diesel has virtually no oxygen. This higher oxygen content makes
the chemical properties of biodiesel more favourable for complete combustion. In
addition, biodiesels typically have very low sulphur contents and many have low
nitrogen levels, which improves air quality from fuel combustion (Hoekman et al.
2012). At the same time, the rise in production and consumption of biodiesels has
focused attention on biodiesel quality standards (Behçet 2011). The most
126 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
internationally recognised biodiesel standards are: EN14214 (in Europe) and ASTM
D-6751 (in USA). Other countries have defined their own standards which typically
derive from either EN14214 or ASTM D-6751(Hoekman et al. 2012).
A large number of potential biodiesel feedstocks have been investigated including
soybean oil, sunflower oil, corn oil, used cooking oil, olive oil, rapeseed oil, castor
oil, lesquerella oil, milkweed seed oil, Jatropha curcas, Pongamia glabra (karanja),
Madhuca indica (mahua) and Salvadora oleoides (Pilu), palm oil and linseed oil.
(Goodrum and Geller 2005; Holser and Harry-O’Kuru 2006b; Kaul et al. 2007a;
Raadnui and Meenak 2003; Lin and Li 2009b; Marchetti, Miguel and Errazu 2008;
Leung and Guo 2006a). However, only a few feedstocks, including rapeseed,
soybean, sunflower, waste cooking oil and tallow are being used to commercially
produced biodiesel 6on industrial scale (Jahirul, Brown, Senadeera, #039, et al.
2013). These commercial biodiesels are produced from edible oil feedstocks and are
typically referred to as first-generation biodiesels (Rashid and Anwar 2008b). The
most contentious issue affecting the production of first-generation biodiesels is the
use of high quality agricultural land for biodiesel production. Farmers of these crops
now have the choice to sell feedstock to the biodiesel production market or food
market. If the biodiesel production market is offering a higher price, farmers will
choose this option more often than not to make a living. This is of particular concern
in poorer countries where crops used for biodiesel production displace the production
of food crops, thus causing a shortage. Supply and demand dictates that a shortage
will cause a price rise, which countries such as Malaysia are already experiencing.
This issue caused global debate due to the 2007-2008 world food price crises.
Different arguments exist regarding the cause of this crisis, however there has been
speculation that the increased consumption of biodiesel caused a food shortage and
subsequent price increases (Kingsbury 2007). Therefore, an alternative must be
considered which eliminates the disadvantages of first-generation biodiesels.
Research is currently taking place on second-generation biodiesels which do not
compete with food production.
In a recent study (Ashwath 2010a), a large number of non-edible oil seed plants were
been identified which have the potential to be used as feedstocks for second-
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 127
generation biodiesel production and have the ability to grow on previously cleared or
degraded land. Of the potential feedstocks assessed, Beauty leaf was identified as the
most suitable feedstock for future biodiesel production as a result of the high oil
productivity of the seeds. Beauty leaf is a moderately sized (8-20 m high) quick
growing tree that can grow up to 1 meter tall within a year. It has been seen to
flourish in the presence of weeds and other species, so can be grown in mixed
cultures with minimal cultivation (Mohibbe Azam, Waris and Nahar 2005b). The
tree naturally grows in tropical and sub-tropical climates (typical temperature range
of 18–33 oC) and in free draining soils near shorelines. Beauty leaf has been seen to
grow in clay soils within Australia and throughout southern and central Asia
including Indonesia, Sri Lanka and India (CSIRO 2010). The Beauty leaf tree bears
fruit twice a year and a healthy plant is able to produce around 8,000 fruits per year.
The fruits contain an oil bearing kernel within a hard husk. The fruits can be
harvested either directly from the tree once they are fully matured or from the ground
once they have fallen off the tree. With around 4,000 fruits per harvest (or 8,000
fruits per year), seed productivity can be as high as 40 kg of seeds per tree per year.
At a tree density of 400 trees per hectare, up to 16,000 kg of dry seeds can be
produced per hectare per year (Mohibbe Azam, Waris and Nahar 2005b; Okano
2006). However, its potential as a source of second-generation biodiesel is yet to be
realised, due to a lack of knowledge of the process for oil extraction from the seed,
oil quality and biodiesel quality. Therefore, this study aims to access different oil
extraction methods for Beauty leaf oil seed and to evaluate the quality of the oil and
biodiesel produced.
5.2 SEED PREPARATION
Seed preparation is critical in optimising the oil extraction process from plant to oil
seed. This is because the physical conditions such as size, hardness and dryness of
seeds and kernels varies significantly from one species to another. Several steps are
involved including: seed collection, kernel extraction and drying. Figure 5-1 shows
Beauty leaf seed preparation steps and brief descriptions of these steps are given in
following sections.
128 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
Figure 5-5-1: Flow chart of Beauty leaf seeds preparation
5.2.1 Seeds collection
About 140 kg of dry Beauty leaf seeds were collected from local seed suppliers in
Queensland, Australia. These seeds were in a ground-dried state (i.e. had been on the
ground for some time prior to collection) and collected mostly from coastal locations
of northern Queensland. In this state the flesh was not present with the endocarp
being the outermost layer. Furthermore, the kernels had shrunk from their fresh state
and a rattle could be heard when seeds were shaken.
5.2.2 Kernel extraction
Dry Beauty leaf seeds were cracked open to expose and obtain the oil bearing
kernels. In order to reduce kernel damage and oil loss, seed cracking was done with
care using two tools which are stompers and mallets. With the stomper, a large
number of seeds were placed on the ground and worked until a number had been
cracked after which the kernels and the waste husk were removed. For the mallet,
operators placed a handful of seeds on a table surface and cracked them individually,
before removing the kernels and the waste husk. During the seed cracking process, it
was found that rubber-headed mallets were preferred less than wooden or steel-
headed mallets, as they tended to rebound excessively. Using mallets meant that
seeds were cracked either individually or several at a time, whereas the stomper was
capable of cracking numerous seeds at a time. However, due to the variability in size
(roughly 1.5–4.5 cm in diameter) of the seeds, the efficacy of the stomper was
reduced as it only struck the largest seeds with each blow. About 51 kg of usable wet
kernels were obtained from cracking 140 kg of Beauty leaf seeds resulting in a kernel
Green seeds and plant
Dry seeds
Oil extraction
Kernel drying
Ground karnel
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 129
yield of 36%. Assuming a seed productivity of 16,000 kg of dry seeds per year per
hectare, it is likely that the Beauty leaf plant is able to produce ~5800 kg of wet
kernel per year per hectare.
5.2.3 Kernel drying
The kernels of Beauty leaf seeds naturally contain high moisture which needs to be
removed for effective oil extraction. Drying was conducted using a Clayson Electric
oven with temperature controller. Kernels were placed in the foil trays; generally 2
kg per tray to ensure the product was spread adequately for uniform drying. The trays
were weighed before placed in the drying oven for 3 days at 40 °C. After that the
temperature of the oven was increased to 70 °C and the drying progress was
monitored by weighing the trays several times daily. Because a fan-forced oven was
used, the tray positions in the oven seemed to impact on its drying, especially those
trays nearest to the oven walls. In order to reduce this effect, the trays were rotated in
the oven to ensure uniform drying rates. The seed was dried until it was observed
that the weight was remaining constant for one day. The moisture content of the
kernels was approximately 32%. Therefore it is expected that about 3960 kg of dry
kernel can be produced from Beauty leaf plant per hectare per year. However, this
may vary depending on seasonal variation, location and maturity of the seeds.
5.2.4 Kernel grinding
The dried seed kernel samples were ground using a blender and coffee grinder to
obtain a fine consistency to maximise particle surface area of the kernels to exposure
in the chemical solvent during the extraction process. After that the seeds were
removed from oven and placed in zip locked bags and placed in a refrigerated store
room.
5.3 OIL EXTRACTION
Oil was extracted from the kernel by three different methods which are: mechanical
oil extraction using an electric powered screw press, chemical oil extraction using n-
130 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
hexane as the solvent at room temperature and pressure and chemical oil extraction
under high pressure and temperature condition. Each of the extraction methods has
its advantages and limitations. A brief description of the oil extraction methods
conducted in this study is given in the following sections.
5.3.1 Mechanical oil extraction using oil press (OP)
A Mini 40 electric motor powered screw press shown in Figure 5-2 was used for the
mechanical oil extraction. As the screw press used in this study was not designed for
Beauty leaf seeds a degree of experimentation was undertaken to optimise pressure
and speed. Beauty leaf kernel was found to be very difficult to process using the
screw press due its physical properties and several cycles were required to extract the
oil. It was also difficult to control the soft kernel paste after one pass and to keep the
process clean. Two operators were required to constantly attend to the machine and
the rate of oil production was very low, typically taking over an hour to process just
about 200 g of sample. Mixing of rice husks with kernels significantly accelerated
the rate of oil production. It was also observed that temperature (both ambient and
barrel/product) have a significant impact on the oil yield. This was evident when
attempting to expel oil at low ambient temperatures (e.g. cold mornings) which took
longer. However, improvements in Beauty leaf oil extraction using the screw press
may be possible by optimising key design parameters of the machine including
pressure, compression ratio, speed, and temperature.
Figure 5-5-2: Mechanical oil extraction through a screw press
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 131
5.3.2 Chemical oil extraction using n-Hexane (nHX)
In this process, oil was extracted using n-hexane as an oil solvent at ambient
conditions. The ground kernels were put into conical flasks in which hexane were
added at a ratio of 2:1 (mL hexane: g kernel). The mixture was given an initial stir to
ensure that all kernels were wetted with hexane. Conical flask openings were covered
with aluminium foil and placed on an orbital mixer under the fume hood and the
samples were left to mix for at least eight hours. After this, the hexane/oil mixtures
were collected, filtered and decanted into aluminium foil containers for solvent
evaporation, and placed under the fume hood (Figure 5-3) for 8 - 10 hours. Hexane
was again added to the conical flask of kernels, but at a ratio of 1:1 for the second
extraction, and a similar procedure was followed for recovery of the oil. When it was
determined that the hexane had been fully evaporated, the oil was transferred into
containers for analysis. It was observed that the n-hexane oil extraction method
resulted in a much greater oil yield than the mechanical oil extraction process.
Figure 5-5-3: Chemical oil extraction
5.3.3 Accelerated solvent extraction (ASE)
The accelerated solvent extraction (Dionex™ ASE 350®) machine shown in Figure
5-4a was used to extract Beauty leaf oil at high pressure and temperature using
accelerated solvent extraction method. The oil extractor comes with an automated
extraction control system that uses elevated temperatures and pressures to achieve
extractions in a short period of time. Measured samples were inserted into metal
sample cells and the desired operating conditions were set using the control interface.
Although the machine allows for the use of up to 3 different types of solvents only n-
132 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
hexane was used as the solvent for lipid extraction. The electric oven maintains the
cell contents at the selected operating temperature throughout the extraction process
and was set to 150 °C. n-hexane was pumped into the cell with pressurised nitrogen
gas to achieve a pressure of 1600 psi. After the extraction process was completed, all
the extracted oil samples were collected into the vessels in the collection tray (Figure
5-4a). The solvent was separated from the extracted sample using the Dionex™ SE®
400 solvent evaporator system as shown in Figure 5-4b.
(a) (b)
Figure 5-5-4: ASE oil extraction (a) Dionex™ ASE 350® (b) solvent removal with
flow of nitrogen
5.4 OIL YIELD
The Beauty leaf oil yield from the three extraction methods used in this study are
summarised in Figure 5-5. All the results are averages of three replicates for each
extraction method. Overall the highest oil yield was obtained using the ASE oil
extraction method which produced on average 51.5 g of oil per 100 g of dry kernels.
The static n-hexane extraction methods produced on average 48 g of oil from 100 g
of dry kernels. These results indicated a 3-4% oil yield increase for the higher
pressure and temperature conditions. This result is likely to be due to the
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 133
improvement in solvation power of n-hexane under higher temperatures. With
increases in temperature, the thermal energy of the solvent increases, which help to
overcome cohesive and adhesive interactions. Moreover, higher temperatures
increase the molecular motion of molecules and decrease hydrogen bond
interactions. Higher pressure facilitates more interactions between the solvent and oil
especially oil that is trapped in pores and would normally not be contacted by
solvents under ambient conditions. These results indicate that about 1.56 tons of oil
per hectare per year can be produced from Beauty leaf plant using chemical
extraction methods with high pressure and temperature. The results also indicate that
the chemical method is more repeatable, and, given the relative ease of preparation
and no requirement for extensive training, it is considered to be more reproducible.
Seed preparation has a significant impact on oil yields especially for the screw press
extraction method. Mechanical extraction using the screw press can produce oil from
appropriately prepared product, but overall this method is ineffective, with relatively
low yields for a great deal of effort.
Figure 5-5-5: Beauty leaf oil yield from three different extraction methods.
5.5 COMPARISON OF OIL EXTRACTION METHODS
134 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
During Beauty leaf oil extraction using the three different oil extraction methods, it
was observed that all the methods have techno-economic advantages and limitations
compared with each other. For example, although the ASE method had higher oil
yields over the other two techniques, it requires high investment, sophisticated
equipment and skilled operators. The advantages and disadvantages of the three oil
extraction methods are summarised in Table 5-1.
Table 5-1: Advantages and disadvantages of the three extraction methods
Methods Advantages Disadvantages
Oil press Virgin oil is more sought after No potential for solvent contamination
Relatively inexpensive after initial capital costs
Minor consumables cost Low preparation is required Whole seeds or kernels can be processed
Time and labour intensive Low oil yields Operators require experience to achieve best results
High dependence on kernel moisture content
Relatively dirty process Filtration or degumming process of oil is required
Low and inconsistent oil production High oil loss
n‐Haxane Repeatable and reproducible
results and process High oil yields Relatively simple process Suitable for bulk oil extraction Low capital investment No especial equipment required
Less sought after than virgin oil High potential for solvent contamination Safety issues and environmental concerns regarding the use of hexane
Relatively costly High hexane requirement Only kernel can be processed
ASE Automatic technique
Condition can be optimised
More efficient
Clean process
Relatively less solvent
consumption
Less time and labor incentives
High oil yield
Very high initial cost
High preparation required
Special equipment and skill required
Potential for solvent contamination
Only kernel can be processed
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 135
5.6 OIL ANALYSIS
Experiments were conducted to determine the quality of the oil extracted in terms of
acid value, density, kinematic viscosity, surface tension and higher heating value.
Viscosity was measured using a Brookfield DV-III Rheometer according to the
ASTM D445 standard test method. Oil density and surface tension were analysed in
accordance with ASTM D1298 and ASTM D971-12 standard test methods using a
KSV Sigma 702 Tensiometer. Higher heating value of biodiesel was measured in
accordance with ASTM D240-09 standard method using a Parr 6200 oxygenated
bomb calorimeter. Acid value of oils was measured using D5555-95 (2011) standard
test method. All experiments were undertaken in triplicate and average results were
used. The experimental results are shown in Table 5-2 along with similar parameters
of other vegetable oil results obtained from literature. Beauty leaf oil obtained from
the screw press showed higher density, kinematic viscosity and lower heating value
compared to oil obtained from ASE and ambient n-hexane methods. This might be
due to the presence of suspended small particles remaining in the oil from
mechanical extraction although large particles were removed via centrifugation. The
acid values of oil from the screw press (36.26 mgKOH/g) and ASE oil (39.22
mgKOH/g) were much higher than the acid value of oil obtained from ambient n-
hexane extraction (24 mgKOH/g). The high pressure and temperature involved in
ASE and press oil might be responsible for creating high free fatty acid in the oil.
When compared with conventional vegetable oils, all of the Beauty leaf oil samples
showed much higher acid values in Table 5-2. These results confirm that raw Beauty
leaf oil is not suitable directly as a fuel for diesel engine application because of
having high acid value and kinematic viscosity and conversion to fatty acid methyl
esters is required prior to use as a fuel.
136 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
Table 5-2: Physical properties of Beauty leaf oil
Vegetable Oil Acid value (mgKOH/g)
Density (kg/lit.)
Higher heating value (Mj/kg)
Kinematic viscosity
(40 °C, cSt)
Beauty leaf*
Oil press 36.26 0.964 38.10 56.74
n-Haxane 24.00 0.936 39.52 42.24
ASE 39.22 0.945 39.34 44.05
Rapeseed** 0.39 0.907 40.05 38.25 Canola** 0.16 0.912 39.74 33.34
Soybean** 0.82 0.914 39.62 32.85
Sunflower** 0.20 0.916 39.49 31.63
Cottonseed** 0.30 0.914 39.40 33.70
Palm** 0.90 0.916 40.14 39.65 * Experiment; **Literature (Singh and Singh 2010; Ramos et al. 2009;
Hoekman et al. 2012)
5.7 BIODIESEL PRODUCTION
Like other conventional vegetable oils shown in Table 5-2, the kinematic viscosity of
Beauty leaf oils (42.24 – 56.74 cSt) are much higher than that of petroleum diesel (2-
3 cSt). Higher viscosities result in higher drag in the fuel line and injection pump,
higher engine deposits, higher fuel pump duties and increased wear in the fuel pump
elements and injectors. This can adversely influence fuel spray, fuel-air mixture
formation and the combustion process which eventually affects engine performance
and emissions (Jahirul, Brown, Senadeera, et al. 2013). Therefore, raw Beauty leaf
oils, as well as other vegetable oils, are not suitable to use as a direct fuel in the
diesel engine. In order to reduce the viscosity of bio-oils to make them normally
suitable for diesel engine use, transesterification is widely used due to its high
conversion efficiency, simplicity, low conversion cost and the fuel qualities of the
product (Lin et al. 2011; Gerpen 2005; Issariyakul et al. 2007). Transesterification is
a chemical reaction in which oils (triglycerides) react with alcohols (eg. methanol,
ethanol) under acid or alkali catalysed conditions, producing fatty acid alkyl esters
and glycerol. The catalyst is used to improve the reaction rate and yield of esters.
After the reaction is complete, glycerol is removed as a by-product and the esters are
purified to produce biodiesel (Fernando et al. 2007). However, alkali catalysed
transesterification cannot be directly used to produce biodiesel from feedstocks
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 137
containing high levels of free fatty acids (FFA). This is because FFAs react with the
catalyst to form soap (Figure 5-6), resulting in emulsification, separation problems
and reduction in biodiesel yield. To overcome this problem, a pre-esterification
process may be used to reduce the content of FFAs in the feedstock. A typical pre-
esterification process uses homogeneous acid catalysts, such as sulphuric acid, or
heterogenous ‘solid-acid’ catalysts, to pre-esterify the free fatty acids (Zhang and
Jiang 2008b; Haas 2005; Samios et al. 2009b) as shown in Figure 5-7.
Figure 5-5-6: Soap formation in oils contains high FFA (Jahirul, et al. 2013)
Figure 5-5-7: Acid pre-esterification (Jahirul, et al. 2013)
A schematic of a two-step process of biodiesel production from high free fatty acid
Beauty leaf oil is shown in Figure 5-8. Both acid-catalysed pre-esterification and
base-catalysed transesterification were conducted in a 500 ml triple neck bottom
flask reactor shown in Figure 5-9a. An oil quantity of 40 g was used for the acid-
catalysed pre-esterification experiments and 30 g was used for each base-catalysed
transesterification trial. For each experiment, oil was carefully transferred into the
reaction flask and preheated in an oil bath to the reaction temperature. For acid-
catalysed esterification trials, sulphuric acid (H2SO4) was used as catalyst. The
sulphuric acid and methanol solution was freshly prepared and added to the
138 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
preheated oil and the mixture was agitated for 2 hours. At the completion of 2 hours,
the mixture was centrifuged in a self-standing tube for 7 minutes to separate the
methanol-water and esterified oil phases as shown in Figure 5-9b. The majority of
the excess methanol, sulphuric acid and impurities were separated into the top phase.
The bottom phase containing the oil was collected for base-catalysed trans-
esterification. The procedures were undertaken in triplicate and average values were
taken. It was found that after acid-catalysed pre-esterification, the acid value of
Beauty leaf biodiesel reduced to 5.14, 3.66, and 6.30 respectively for screw press,
ambient n-Hexane and ASE extracted oils.
In the base-catalysed transesterification trials, sodium methoxide (NaOCH3) was
used as catalyst with a reaction time of 1.5 hours. Similarly to the acid-catalysed pre-
esterification trials, the phases of the transesterification product were separated using
a centrifuge and the bottom layer drained using a separation funnel as shown in
Figure 5-9(c). The top layer containing crude Beauty leaf biodiesel was collected and
washed to remove the soap, unreacted methanol and other contaminant. All
experiments were undertaken in triplicate. The average methyl ester conversion for
the screw press, n-Hexane and ASE extracted Beauty leaf oils were 75.47%, 93.05%
and 83.76%, respectively. The results clearly indicated the dependency of the
biodiesel conversion process the on the presence of free fatty acid in the base oil.
Therefore after analysing these results, it is clear that methyl ester production
efficiency not only depends on feedstock, but also the oil extraction methods.
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 139
Figure 5-5-8: Two step bio-diesel production process from Beauty leaf oil
Figure 5-5-9: Beauty leaf oil esterification
(a) Esterification reactor (b) Layer of Methanol-Water (top) and oil (bottom) after
acid-catalysed pre-esterification; (c) Layer of Beauty leaf oil methyl ester (top) and glycerol (bottom) after base-catalysed Trans-esterification
140 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
5.8 BIODIESEL ANALYSIS
The chemical composition of biodiesels is very important for determining their
suitability for automobile engine application. Chemically, all biodiesels are mono
alkyl esters of fatty acids, commonly referred as fatty acid methyl or ethyl esters.
Depending on the feedstock and production process, the fatty acids are different in
relation to the chain length, degree of unsaturation or presence of other chemical
functions. Fatty acids are commonly designated by two numbers: the first number
denotes the total number of carbon atoms in the fatty acid and the second is the
number of double bonds. For example, 18:1 designates oleic acid which has 18
carbon atoms and one double bond. Table 5-3 shows the fatty acid methyl ester
composition of Beauty leaf oil (BLOME) produced through three different oil
extraction methods along with traditional biodiesel obtained from soybean (SOME),
canola (COME), palm (POME), rapeseed (ROME) and sunflower (SOME) oil
feedstocks. The FAME composition of Beauty leaf, soybean and canola biodiesels
were analysed by gas chromatography and flame ionisation detection (GC-FID) in
accordance with EN 14103 standards. The gas chromatograph (GC) was a Hewlett-
Packard 6890 System fitted with Varian Select TM 30 m × 0.32 mm × 0.25 µm
column. FAME compositions of other biodiesel were collected from literature [28].
The prominent fatty acids found in chemical composition of biodiesels were Palmitic
(Hexadecanoic, C16:0), Stearic (Octadecanoic, C18:0), Oleic (9-Octadecenoic,
C18:1) and Linolenic (9, 12-Octadecadienoic, C18:2) acids.
Three main types of fatty acids were found in the biodiesel samples: saturated
(Cn:0), monounsaturated (Cn:1) and polyunsaturated with two or three double bonds
(Cn:2,3). The percentage of these compounds for each vegetable oil is given in Table
6-3. Based on this composition, average chain length (ACL) and Average number of
double bond (ANDB) were estimated using equations 5-1 and 5-2.
∑ ∙ : 0, 1, 2, 3, .% ……………………………..Equation 5-1
1 ∙ : 1, . % 2 ∙ : 2, .% 3. : 3, %
100
……Equation 5-2
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 141
where n is the number of carbon atom in fatty acid chain.
Similarly to other biodiesels shown in Table 5-3, Beauty leaf oil biodiesels were also
high in Palmitic (C16:0), Stearic (C18:0), Oleic (C18:1) and Linolenic (C18:2) acids
esters. Mono-unsaturated stearic (C18:1) acid methyl ester is the most prominent
consisting of 38.6-40.29% by weight followed by poly-unsaturated Linolenic (22.81
- 27%), saturated Stearic (16.59 – 18.64%) and Palmitic (14.48-14.73%). Beauty leaf
oil biodiesels have higher saturated esters consisting of 32.7 – 34% which after palm
oil biodiesel which is 44.6%. However, Beauty leaf biodiesel showed higher long
chain saturation factor (10.72 – 11.81) over palm oil biodiesel (7.37). This is because
palm oil biodiesel is richer in short chain saturated Palmitic acid esters compared
with Beauty leaf biodiesel. Overall, the chemical compositions of Beauty leaf
biodiesels were closer to palm oil biodiesel than any other biodiesels shown in Table
5-3.
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 142
Table 5-3: The fatty acid distributions of Beauty leaf and commercial biodiesels
FAME Formula BLOME* SOME* COME* POME** ROME** SFOME**
OP ASE nHX Lauric C12:0 0 0 0 0 0.4 0.1 0 0 Myristic C14:0 0 0 0 0 0.53 0.7 0 0 Palmitic C16:0 14.73 14.48 14.68 13.04 13.19 36.7 4.9 6.2 Palmitoilic C16:1 0.19 0.22 0.24 0.28 5.6 0.01 0 0.1 Stearic C16:1 0 0.04 0 0 0 0 0 0 Oleic C18:0 16.59 18.64 18.25 6.32 3.04 6.6 1.6 3.7 Linoleic C18:1 39.3 40.29 40.18 26.59 47.1 46.1 33 25.2 Linolenic C18:2 27 22.81 23.23 45.34 27.2 8.6 20.4 63.1 Gondonic C18:3 0.28 0.17 0.19 6.9 5.23 0.3 7.9 0.2 Erucic C20:0 0.95 1.04 1.02 0.44 0.55 0.4 0 0.3 Lauric C20:1 0.29 0.24 0.23 0.3 0.95 0.2 9.3 0.2 Myristic C22:1 0 0 0 0 0 0 23.0 0.1 Saturated (wt%) 32.7 34.7 34.4 20.4 18.8 44.6 6.5 10.4 Mono-unsaturation (wt%) 39.86 40.88 40.51 27.28 53.97 46.31 65.30 25.30 Poli- unsaturation (wt%) 27.28 22.98 23.42 52.24 32.43 8.90 28.30 64.00 Average chain length (ACL) 17.74 17.74 17.74 17.76 17.67 17.28 19.01 17.94 Average number of double bond (ANBD)
0.95 0.87 0.87 1.32 1.19 0.64 1.22 1.53
* Experiment; **Literature (Singh and Singh 2010; Ramos et al. 2009; Hoekman et al. 2012)
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 143
5.9 FUEL PROPERTIES
Biodiesel properties from any types of feedstock need to meet the relevant quality
standard before being accepted as an acceptable automobile fuel. However, biodiesel
properties can vary substantially from one feedstock to the next due to differences in
the compositional profiles describe above. In some cases properties also vary in
similar feedstocks from different origins and production processes. However, the
quality standards are crucial for the commercial use of any fuel product, which
serves as guidelines for the production process, to assure customers are buying high-
quality fuels, and provide authorities with approved tools for the assessment of safety
risks and environmental pollution. The most internationally recognised biodiesel
standards are: EN14214 (in Europe) and ASTM D-6751 (in USA). Numerous other
countries have defined their own standard, which in many cases derive from either
EN14214 or ASTM D-6751. With the increasing production of biodiesel within
Australia and as a part of the Fuel Quality Standards Act 2000, the Australian
government has released a biodiesel fuel standard, “Fuel Standard (Biodiesel)
Determination 2003”. This standard is an adaptation of the above US and EU
standards and fuel standards differ only slightly to conform to Australian climate
related requirements. A summary of the important fuel quality parameters of Beauty
leaf oil biodiesels and conventional biodiesels across all three standards are shown in
Table 5-4. Among the fuel properties listed in the Table 5-4, kinematic viscosity,
density, higher heating value and acid value were obtained from experiment for
Beauty leaf biodiesels, soybean and canola biodiesel. A similar experimental
procedure was followed for these four parameters were described in the previous
section. For comparison purposes, experimental data for above mentioned parameters
were obtained from literature (Ramos et al. 2009) for palm, rapeseed and sunflower
oil biodiesel. The other fuel property parameters were estimated using ANN models
developed in Chapter 3.
144 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
5.9.1 Kinematic viscosity
Kinematic viscosity (KV) is an important property of biodiesel since it plays a
dominant role in the fuel atomisation, fuel-air mixture formation and combustion
process particularly at cold weather when an increase in viscosity affects the fluidity
of fuel. The higher the KV, the higher is the pressure loss in the fuel line and
injection pump, therefore resulting in increases in engine deposits and shoot
formation, requiring more energy to pump the fuel and increasing wear on fuel pump
elements and injectors. On the other hand, low fuel KV is not desirable because it
will not provide sufficient lubrication for the precision fit of fuel injection pumps,
resulting in leakage or increased wear (Jahirul, Brown, Senadeera, #039, et al. 2013).
Therefore, the upper and lower limit of biodiesel KV is defined in all biodiesel
standards shown in Table 5-4. The KV of produced Beauty leaf biodiesels was 4.38 -
4.46 mm2/sec which were at the acceptable limit according to all biodiesel standards.
Table 5-4 also shows that the KV of Beauty leaf, palm, rapeseed and sunflower oil
biodiesels were quite similar at around 4.4 mm2/sec whereas soybean oil biodiesel
has the lowest (3.86 mm2/sec) and cottonseed oil biodiesel has the highest kinematic
(5.45 mm2/sec). The Beauty leaf biodiesel made from oil through oil press showed
slightly higher KV compared with chemical oil extraction, which may be due to the
higher viscosity of the same feedstock. Overall there is only minor variation in KV
was found between three different Beauty leaf biodiesels.
5.9.2 Density
Density is defined as mass per unit volume of the liquid fuel commonly expressed in
units of kg/m3. Density is an important property for automobile fuel because it
influences the amount of fuel injected in the engine cylinder. Changes in fuel density
will influence engine output power due to a different mass of fuel injected which
directly affects engine performance. Comparing crude vegetable oil (Table 5-2) and
vegetable oil methyl ester (Table 5-4), it can be seen that esterification process
reduces the density by 7 to 8%. Beauty leaf biodiesel produced from oil obtained
through mechanical extraction showed slighter higher density compared with that of
other Beauty leaf biodiesels. However, the densities of all biodiesels shown in Table
5-4 were in the acceptable range specified by Australian and European biodiesel
standards.
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 145
5.9.3 Higher heating value
Higher heating value (HHV) is important fuel property for identifying the suitability
of biodiesel as an engine fuel. It indicates energy content in the fuel per unit mass.
Therefore, the conventional unit of higher HHV is KJ/gm or MJ/kg. The HHV of
Beauty leaf biodiesels were found to vary from 40.85 to 40.96 MJ/kg with very little
fluctuation among the beauty oil source. These results indicate that the HHV of
vegetable oil methyl ester is about 4% higher than the crude vegetable oil as show in
Table 5-2. Table 5-4 shows that the biodiesel produced from non-edible Beauty leaf
oil produced has a HHV close to that of commercial biodiesel produced from edible
vegetable oil feedstocks.
5.9.4 Acid number
Acid number (AN) indicates the amount of carboxylic acid present, such as in fatty
acids. It is expressed as the amount of KOH (mg) required for neutralising 1 g of
fatty acid methyl ester or biodiesel. Fuel with high acid number can cause higher
level of lubricant degradation and severe corrosion in engine fuel systems (Haseeb et
al. 2011a). AN is set to a maximum of 0.5 KOH/g in both European (EN14214) and
American (ASTM D6751) biodiesel standards, whereas the Australian standard
allow slightly higher AN, setting the maximum value at 0.8 KOH/g. Naturally, the
AN of crude Beauty leaf oils were very high compared with the traditional edible
vegetable oil shown in Table 5-2. Table 5-4 shows that a significant reduction of this
acid value occurred in the two stage biodiesel production process utilised in this
study. However the acid number of Beauty leaf oil biodiesels remained high when
compared with other commercial biodiesels. Table 5-4 indicates that oils obtained
through oil press, n-Hexane and ASE produced biodiesel were 0.88, 0.76 and 1.00
KOH/g respectively. Although only biodiesel from n-Hexane oil met the Australian
biodiesel standard (Table 5-4), the other biodiesels were only slightly higher than the
standard. It is likely that with further optimisation, biodiesel from Beauty leaf oil
should be able to meet EN standards.
146 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
5.9.5 Oxidation stability
Oxidation stability (OS) is a fuel property which reflects the resistance of the fuel to
oxidation during long-term storage. Usually biodiesels show less oxidative stability
compared with petroleum diesel due to their different chemical composition and this
is one of the major issues that limits the wide spread use biodiesel as a fuel in
automobile engines. All of the biodiesels listed in Table 5-4 failed to meet the ASTM
standard in terms of oxidation stability, which is 3 hours minimum. Only Beauty leaf
and palm biodiesel were in the range of European standard of oxidation stability.
Beauty leaf biodiesels showed oxidation stability from 4.12 to 4.42 hours which was
much higher than all conventional biodiesel except palm oil biodiesel. This is
because oxidation is influenced by the presence of double bonds in the chains, that is,
feedstocks rich in polyunsaturated fatty acids are much more susceptible to oxidation
than the feedstocks rich in saturated or monounsaturated fatty acids. For the same
reason, Beauty leaf biodiesel obtained from oil press showed less oxidation stability
then other types. An overall oxidation estimation result confirms that Beauty leaf
biodiesel is a better fuel in terms of OS then most of the commercial biodiesel.
5.9.6 Iodine value
Iodine value (IV) is an important parameter in regard to fuel quality because higher
IV biodiesel leads to a higher rate of polymerisation of glyceride which results in
increasing fuel viscosity, causing the formation of engine deposits, thus adversely
affecting fuel injector spray patterns. This property is set to a maximum value of 120
g I2 /100g according to EN14214 standard. The IV results tabulated in Table 5-4
indicate that all biodiesels meet the EN14214 standard except the sunflower oil
methyl ester. The IV of Beauty leaf oil biodiesels (74.81 – 81.44 I2 /100g) were well
below the allowable limit and also below most of the commercial biodiesel. Only
palm oil biodiesel showed better result than Beauty leaf biodiesel in IV which was
estimated 57 I2 /100g. Oil press Beauty leaf biodiesel showed slightly higher IV due
to having a higher degree of unsaturation compared to other types.
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 147
5.9.7 Cetane Number
Cetane number (CN) is a widely used diesel fuel quality parameter, and is a
measurement of the combustion quality of diesel fuels during compression ignition.
It is related to the ignition delay (ID) time, that is, the time that passes between
injection of the fuel into the cylinder and the onset of ignition (Knothe 2005). A high
CN will help to ensure good cold start properties and will minimise the formation of
white smoke. On the other hand, lower CN may result in diesel knocking and an
increase in exhaust emissions. Australian and European biodiesel standard limit the
CN to a minimum value of 51 whereas ASTM standard limit it minimum value of 47
as shown in Table 5-4. Results indicate an excellent ignition quality of biodiesel
produced from Beauty leaf oil biodiesels. The CN of Beauty leaf biodiesels were
58.53 to 60.39; much higher than the minimum recommended value of 51. Moreover
the CN of Beauty leaf biodiesels were far better than most of the commercial
biodiesel produced from edible oil. The CN of palm oil biodiesel was found to be
61.80, slightly higher than Beauty leaf oil biodiesel. This is because palm oil contains
a higher percentage of saturated methyl ester. Oil press biodiesel showed a slightly
lower CN then the other Beauty leaf oil biodiesels which may due to the higher
linoleic acid content, which increases the degree of unsaturation and hence reduces
the CN.
5.9.8 Flash point temperature
The flash point (FP) is defined as the lowest temperature at which the fuel will start
to vaporise to form an ignitable mixture when it comes in contact with air. Australian
and European biodiesel specification required flash point temperature at least 120 °C,
whereas in the US the minimum requirement level is 93 °C. Table 5-4 shows that the
FP temperature of Beauty leaf biodiesels were between 143.06 and 145.64 °C, which
is higher the minimum requirement specified in the biodiesel standards. While
comparing with commercial biodiesel, Beauty leaf biodiesel showed lower flash
point temperature. It is noted that very high flash point temperature of automobile
fuel is not desirable because it can cause cold engine start-up problems, misfiring and
ignition delay, which increases carbon deposition in the combustion chamber
(Szybist et al. 2007). No significant variation in FP temperature was noted among the
different Beauty leaf biodiesel results shown in Table 5-4.
148 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
5.9.9 Cold filter plug point (CFPP)
One of the major problems associated with the use of biodiesel in countries with a
cold climate is their poor cold flow properties when compared with petroleum diesel
fuels. The parameter generally used to determine the cold flow property is cold-filter
plugging point (CFPP). The CFPP is defined as the lowest temperature at which a
fuel portion will pass through a standardised filtering device in a specified time
(Jahirul, et al. 2013). The cold temperature properties of biodiesel should be reported
according to the Australian, European and US although the limits are not specified.
However it is commonly understood that biodiesels with low CFPP, CP and PP are
better options for diesel engine fuels operating in cold weather condition. Table 5-4
shows that all cold temperature properties of Beauty leaf biodiesels were much
higher than that from most commercial biodiesel. Among the biodiesel shown in
Table 5-4, palm oil biodiesel showed highest CFPP temperature followed by Beauty
leaf oil biodiesel. The rapeseed oil biodiesel followed by soybean oil biodiesel
showed the lowest cold temperature properties. The average CFPP of Beauty leaf
biodiesel was found 3.5, 12.6 and -2.9 °C, respectively. Beauty leaf oil produced
through oil press showed slightly better cold weather properties due to having a
higher linolenic acid methyl ester content.
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 149
Table 5-4: Fuel properties of Beauty leaf oil biodiesel and commercial biodiesels
Property Unit Biodiesel Standard BLOME* SOME* COME* POME** ROME** SFOME**
Australian ASTM D6751
EN 14214 OP ASE nHX
Kinematic viscosity @40°C
mm2/sec 3.5-5 1.9-6 3.5-5 4.46 4.34 4.38 3.86 5.45 4.5 4.4 4.2
Density kg/m3 .860-900 n/a 860-900 0.894 0.892 0.893 0.863 0.871 0.874 0.877 0.880
HHV Mj/kg n/a n/a n/a 40.85 40.52 40.46 40.78 41.59 41.24 41.55 41.26
Acid number mg KOH/g .8, max 0.5, max 0.5, max 0.88 1.00 0.76 0.34 0.91 0.12 0.16 0.15
Oxidation stability hours n/a 3, min 6, min 4.14 4.44 4.42 2.71 3.21 5.31 3.09 1.88
Iodine value g iod/ 100g n/a n/a 120, max 81.44 74.81 75.24 119.47 107.07 57 109 132
Cetane Number - 51 min 47, min 51 min 58.53 60.42 60.39 47.89 49.16 61.80 52.02 44.90
Linolenic acid content
%(m/m) n/a n/a 12, max 0.28 0.17 0.19 6.9 5.23 0.3 7.9 0.2
Flash point °C 120, min 93, min 120, min 145.64 143.06 143.65 160.87 162.00 176 170 177
CFPP °C Report Report Report 2.45 4.11 3.92 -5.76 -2.94 10 -10 -3 *Experimental; **Literature (Ramos et al. 2009; Singh and Singh 2010; Hoekman et al. 2012)
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 150
5.10 VALIDATION OF BEAUTY LEAF BIODIESEL
To be an ideal source of biodiesel, Beauty leaf biodiesel should have suitable
chemical composition to ensure compliance with standard biodiesel properties. The
fuel properties of Beauty leaf biodiesel from three different extraction methods were
analysed and compared with five other commercially available biodiesels. To
determine the suitability of Beauty leaf biodiesels compared to other biodiesels based
on 14 criteria (fuel properties): CN, IV,OX, AN, HHV, KV, Density, FP, CFPP,
Linolenic acid, ACL, MUFA and PUFA, a multi-criteria decision method (MCDM)
software was used. In this study, the Preference Ranking Organisation Method for
Enrichment Evaluation (PROMETHEE) and Graphical Analysis for Interactive
Assistance (GAIA) were used because of their rational decision vector which
stretches towards the preferred solution compared to other MCDM (Brans and
Mareschal 1994).
Table 5-5: Variables and preference used in PROMETHEE-GAIA analysis
Variables Preference For PROMETHEE‐GAIA
Kinematic viscosity (KV) Min Density Min Higher heating value (HHV) Max Acid number (AN) Min Oxidation stability (OX) Max Iodine value (IV) Min Cetane Number (CN) Max Linolenic acid (LA) Min Flash point (FP) Max Cold filter plug point (CFPP) Min
In GAIA plane, the criteria which lie close to (±45°) are correlated, while those lying
in opposite directions (135–225°) are anti-correlated, and those in a roughly
orthogonal direction have no or less influence (Espinasse, Picolet and Chouraqui
1997). The preference function criteria (fuel property) were modelled as minimum
(i.e. lower values are preferred for good biodiesel) or maximum (higher values are
preferred for good biodiesel). The selection of preference function also influences the
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 151
direction of criteria. For example, IV and CN were inversely related but still showed
the same direction within ±45○. This is because the Cetane number were preferred to
maximum, but iodine number was preferred to minimum, as shown in Table 5-5,
which was suggested by Islam et al. (2013). Therefore, criteria which are in the same
preference (min/max) and lie close to ±45° are correlated. The direction and length
of criteria are indicative to their influence on decision vector (marked as red line in
Figure 5-10) (Islam et al. 2013), such that the very short length of some criteria, in
particular ‘Density’ and ‘HHV’, indicate the little effect on the decision vector.
The decision vector indicates the most preferable samples (i.e. those that align with
the direction of this vector) and the outermost criteria in the direction of the decision
vector are the most preferable (Figueira, Greco and Ehrgott 2005). In this study,
equally weighted criteria showed (Figure 5-10a) that POME was most aligned with
the decision vector and its farthest position from the centre gave it the highest
ranking.
Rank Biodiesel Phi
1 POME 0.16
2 ROME 0.05
3 BLOME_OP 0.02
4 BLOME_nHX 0.01
5 BLOME_ASE -0.01
6 COME -0.06
7 SFOME -0.08
8 SOME -0.10
(a) (b)
Figure 5-5-10: (a) Graphical Analysis for Interactive Assistance (GAIA) plot for
eight biodiesel showing 10 criteria and decision vector. (b) Corresponding ranking of
biodiesel on their outranking flow.
152 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
Figure 5-10(b) shows the overall ranking of the different biodiesel and the three
biodiesel from Beauty leaf, BLOME_OP, BLOME_nHX and BLOME_ASE, were
placed third, fourth and fifth, respectively, in the middle of the overall rankings. The
Phi value is the net flow score which could be negative or positive depending upon
the angular distance from the decision vector and the distance from the canter.
Biodiesel from soybean oil was at the bottom of the ranking compared with other
biodiesel. In can be seen from Figure 10 that the quality of Beauty leaf biodiesel in
terms of fuel properties did not depends on oil extraction methods. The results of this
analysis indicate the ability of Beauty leaf biodiesel to compete with commercially
available first- generation biodiesels.
Table 5-6: Comparative rank shift with different OS and CFPP weighting
OS CFPP Weighting 1‐3 4‐6 6‐10 1‐2 3‐4 5‐6 7‐10
POME 1 1 ‐ 1 ‐ 1 5 7 8
ROME 2 5 6 2 1 1 ‐ 1 ‐ BLOME_OP 3 2 2 3 6 6 ‐ 6 ‐
BLOME_nHX 4 3 3 4 7 8 7
BLOME_ASE 5 4 4 5 8 5 5 ‐
COME 6 6 ‐ 5 6 3 3 ‐ 3 ‐
SFOME 7 7 ‐ 7 ‐ 7 4 4 ‐ 4 ‐
SOME 8 8 ‐ 8 ‐ 8 2 2 ‐ 2 ‐
Black arrows upward: rank increase; Black arrows downward reduce rank; Hyphen: no ranking change
The quality ranking analyses of biodiesel shown in the previous section was
conducted with an equal weighting of all parameters. However, the importance of
some fuel properties depends on the country and place where it will be used and
stored. In tropical/sub-tropical regions, CFPP was not considered to be of importance
here. Elevated temperatures of these regions are, however, likely to affect oxidative
stability of the biodiesel. On the other hand, in the winter climate condition CFPP are
more important than oxidation stability. Therefore, ranking sensitivity analysis was
conducted for the fuel properties CFPP and OS by increasing the weighting from 1
(equal to other parameters) to 10, and the results are shown in the Table 5-6. A
significant change in ranking was found for both OS and CFPP. POME always
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 153
ranked 1 with the increasing of OS weighting. At the same time, the rank of Beauty
leaf biodiesels improved and was ranked just after POME. In contrast, the rank of
ROME dropped dramatically with the weighting increase of OS. On the other hand,
an opposite trend was observed when the weighting was increased for CFPP. Both
POME and Beauty leaf biodiesels dropped in rank and were placed at the bottom in
the ranking table. Therefore, as for palm oil biodiesel, Beauty leaf oil biodiesels are
unlikely to be suitable for cold climate conditions, especially in winter. These results
indicate that Beauty leaf biodiesels are a better choice for tropical/sub-tropical
regions than colder climate conditions.
5.11 CONCLUSION
Second-generation biodiesel is gaining more interest in the market as a sustainable
alternative of diesel fuel. However, to produce biodiesel from new sources and
continue to develop these in the market, various aspects must be examined. In this
study, the potential of Beauty leaf plant was evaluated as a source of second-
generation biodiesel. Oil was extracted from dry seed kernels using 3 different oil
extraction methods and oil properties have been analysed. Oil has been esterified to
produce biodiesel using a two-step esterification technique and the physico-chemical
properties were assessed. From the results obtained in this study the following
conclusion can be made.
By flowering two seasons in a year, Beauty leaf plant is able to produce a large
amount of seeds that contain non-edible oil. Due to the variability in size of the
seeds, having relatively soft and high moisture containing (about 32% by weight) oil
bearing kernels, special care need to be taken during the seed cracking process. This
will prevent damage to the kernels and reduce oil loss. The conventional seed
cracking methods using mallets and a stomper were able to produce good quality
kernels but those processes were found to be slow, labour-intensive and might be not
suitable for large scale production, processing approximately 2–3 kg of seeds per
operator per hour. Therefore, an automated seeds cracking device needs to be
designed for industrial scale production using Beauty leaf oil seeds.
154 Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production
This study also found that all oil extraction methods had several advantages and
disadvantages in terms of oil production from Beauty leaf oil seeds, which are
summarised in Table 1. The performance of Beauty leaf oil extraction using an oil
press resulted in a low oil yield. This drawback was overcome using chemical oil
extraction using n-hexane as oil solvent. Furthermore, the oil yield further increased
by 3-4% with high pressure and temperature extraction. The highest oil yield was
found on average 51.5% of dry kernels in ASE extraction method, which suggested
that Beauty leaf plant is able to produce about 1.56 tons of oil per year per hectare.
When comparing quality with edible vegetable oils, conventionally used as biodiesel
feedstock, in terms of acid value, density, kinematic viscosity, surface tension and
higher heating value, Beauty leaf oil showed much higher acid values resulting from
high free fatty acid contents. Chemical oil extraction under atmospheric conditions
produced oil containing relatively low levels of free fatty acids. However, those
results have illustrated that raw Beauty leaf oil may not suitable for direct use in
diesel engines. Another drawback of Beauty leaf oil is that conventional base-
catalysed transesterification cannot be used directly for biodiesel production.
Therefore, a two-step esterification process, involving acid-catalysed pre-
esterification and base-catalysed trans-esterification, was used in this study. during
the first stage of this process, the acid value was significantly reduced to the
acceptable limit for base-catalysed trans-esterification. The highest biodiesel
conversion efficiency was found to be 93.05% for the oil produced by chemical oil
extraction in atmospheric condition, whereas oil obtained from screw press and ASE
methods showed 75.74% and 83.76%, respectively, under similar reaction
conditions, which is due to variations in the acid value of the respective oils.
Beauty leaf oil biodiesels mostly comprise esters of saturated Hexadecanoic (C16:0)
and Octadecanoic (C18:0) acid, mono-unsaturated 9-Octadecenoic acid (C18:1) and
poli-unsaturated 9, 12-Octadecadienoic (C18:2). This biodiesel is rich in saturated
methyl esters compared with commercial biodiesels, except biodiesel from palm oil
and is also rich in long chain saturation factors. Like palm oil, this makes Beauty leaf
oil biodiesel better in terms of most of fuel properties, including kinematic viscosity,
density, higher heating value, oxidation stability, iodine value, cetane number, flash
Chapter 5: Pilot scale Beauty leaf (Calophyllum Inophyllum) biodiesel production 155
point, linoleic acid content. On the other hand, Beauty leaf biodiesels are perform
worse in terms of cold temperature properties and free fatty acid content. However,
Beauty leaf biodiesel is able to meet the American, European and Australian
biodiesel standards. The multivariate data analysis using PROMETHEE-GAIA
software indicated that biodiesel from Beauty leaf oil could be a better option for
automobile engine application compared with many other commercial biodiesel,
including biodiesel from cotton seed, sunflower and soybean oil, specially in
tropical/sub-tropical regions.
Chapter 6: Production process optimisation of biodiesel 157
Chapter 6: Production process optimisation of biodiesel
Biodiesel Production from Non-Edible Beauty Leaf
(Calophyllum inophyllum) Oil: Process Optimisation Using
Response Surface Methodology (RSM)
Md I. Jahirul , Wenyong Koh, Richard J. Brown , Wijitha Senadeera , Ian O’Hara
and Lalehvash Moghaddam
Publication: Journal of Energies. Vol. 7(8), pp. 5317-5331, 2014.
http://www.mdpi.com/1996-1073/7/8/5317
Author Contribution
Contributor Statement of Contribution
Jahirul M. I Conducted the experiments, performed the data analysis and drafted the manuscript Signature
Wenyong Koh Assisted with conducting the experiment, performed the data analysis and drafted the manuscript
Brown R. J Supervised the project, aided with the data analysis, development of the paper and extensively revised the manuscript
Senadeera W Supervised the project, aided with the development of the paper
Moghaddam L Assisted with conducting the experiment
Ian O’Hara Supervised the project and revised the manuscript
Principal Supervisor Confirmation
I have sighted email or other correspondence from all co-authors confirming
their certifying authorship.
Name
Dr Wijitha Senadeera
Signature
Date
158 Chapter 6: Production process optimisation of biodiesel
Abstract
In recent years, the Beauty leaf plant (Calophyllum Inophyllum) is being considered
as a potential 2nd generation biodiesel source due to high seed oil content, high fruit
production rate, simple cultivation and ability to grow in a wide range of climate
conditions. However, however, due to the high free fatty acid (FFA) content in this
oil, the potential of this biodiesel feedstock is still unrealised, and little research has
been undertaken on it. In this study, transesterification of Beauty leaf oil to produce
biodiesel has been investigated. A two-step biodiesel conversion method consisting
of acid catalysed pre-esterification and alkali catalysed transesterification has been
utilised. The three main factors that drive the biodiesel (fatty acid methyl ester
(FAME)) conversion from vegetable oil (triglycerides) were studied using response
surface methodology (RSM) based on a Box-Behnken experimental design. The
factors considered in this study were catalyst concentration, methanol to oil molar
ratio and reaction temperature. Linear and full quadratic regression models were
developed to predict FFA and FAME concentration and to optimise the reaction
conditions. The significance of these factors and their interaction in both stages was
determined using analysis of variance (ANOVA). The reaction conditions for the
largest reduction in FFA concentration for acid catalysed pre-esterification was 30:1
methanol to oil molar ratio, 10% (w/w) sulphuric acid catalyst loading and 75 °C
reaction temperature. In the alkali catalysed transesterification process 7.5:1
methanol to oil molar ratio, 1% (w/w) sodium methoxide catalyst loading and 55 °C
reaction temperature were found to result in the highest FAME conversion. The good
agreement between model outputs and experimental results demonstrated that this
methodology may be useful for industrial process optimisation for biodiesel
production from Beauty leaf oil and possibly other industrial processes as well.
Keywords: biodiesel; Beauty leaf; trans-esterification; response surface
methodology (RSM)
Chapter 6: Production process optimisation of biodiesel 159
6.1 INTRODUCTION
The current global energy supply is heavily dependent on finite reserves of fossil
fuels (oil, natural gas, coal) which represent 88% of total global energy consumption.
Based on current production scenarios, it is expected that the peak of global oil
production will occur between 2015 and 2030 (Jahirul, Brown, Senadeera, O'Hara, et
al. 2013). Therefore, fossil resources have practical limitations in their capacity to
supply future global energy requirements, and there are currently few large scale
alternatives available. Moreover, combustion of fossil fuels results in greenhouse gas
emissions and contributes to anthropogenic climate change. Despite global measures
such as the Kyoto Protocol and scientific innovation, atmospheric CO2 concentration
continues to increase and is exceeding benchmark levels much earlier than had
previously been predicted (Weitzman 2007).
With a growing world population, increasing energy consumption per capita, and the
impacts of global warming resulting from greenhouse gas emissions, the need for
long-term alternative energy source is acute (Jahirul et al. 2012; Jahirul et al. 2007;
Jahirul et al. 2010). Over the past few decades, biodiesel produced from oilseed
crops and animal fat is receiving much attention as a renewable and sustainable
alternative for automobile engine fuels, particularly for petroleum diesel (Reijnders
2006). It is currently produced in commercial quantities from edible
oil feedstocks such as soybean, palm, rapeseed and canola oil. Biodiesels produced
from these feedstocks are generally referred to as first-generation biodiesels (Rashid
and Anwar 2008b). Although biodiesels from these feedstocks offer reductions in
greenhouse gas emissions (GHG) and improve domestic energy security, first-
generation biodiesels are unlikely to be sustainable in the longer term due to land use
impacts and the price and social impacts associated with using a food-based
feedstock. Second-generation biodiesels produced from non-edible feedstocks have
the potential to overcome the disadvantages associated with first-generation
feedstocks, while addressing many of the climate change and energy availability
challenges (Posten and Schaub 2009).
160 Chapter 6: Production process optimisation of biodiesel
Vegetable oils are extremely viscous, ranging from 10 to 17 times higher viscosity
than that of petroleum diesel (Sinha, Agarwal and Garg 2008; Haseeb et al. 2011b).
This makes the raw oils unsuitable for direct use as a fuel in a modern diesel engine.
As a consequence, researchers and scientists have developed various methods to
reduce the viscosity of bio-oils to make them suitable for diesel engine use. Some of
these methods include dilution with other fuels, trans-esterification, micro-
emulsification, pyrolysis and catalytic cracking (Lin et al. 2011). Among these
techniques, transesterification is the most widely used solution due to its high
conversion efficiency, simplicity, low conversion cost and the good fuel qualities of
the product (Fernando et al. 2007).
Transesterification is a chemical reaction in which oils (triglycerides) react with
alcohols (e.g., methanol, ethanol) under acid or alkali catalysed conditions,
producing fatty acid alkyl esters and glycerol. A catalyst is used to improve the
reaction rate and ester yield. Because the transesterification reaction is reversible,
excess alcohol is used to shift the equilibrium to favour production of esters. After
the reaction is completed, glycerol is removed as a by-product and the esters are
purified into biodiesel (Fernando et al. 2007).
One limitation with the alkali catalysed transesterification process is that this process
is not suitable for vegetable oils containing high levels of free fatty acids (FFA). This
is because FFAs react with the catalyst to form soaps, resulting in emulsification and
separation problems (Rajendra, Jena and Raheman 2009). In addition excessive soap
formation reduces biodiesel yield and obstructs subsequent purification processes
including glycerol separation and water washing (Lam, Lee and Mohamed 2010).
However, the maximum limit of FFA in vegetable oil for alkali catalysed
transesterification is still uncertain with different benchmarks being reported. For
example, Van Gerpan reported that vegetable oils containing up to 5% FFA can be
trans-esterified using an alkali catalyst while Dorodo et al. (Dorado et al. 2002) and
Ramadhas et al. (Ramadhas, Jayaraj and Muraleedharan 2005) reported that FFA
content should not be greater than 3% and 2%, respectively. Many researchers have
also reported that FFA should be kept less than 1% for alkali catalysed
transesterification (Kumar Tiwari, Kumar and Raheman 2007; Ma and Hanna 1999;
Chapter 6: Production process optimisation of biodiesel 161
Zhang et al. 2003). In order to overcome the difficulties related to trans-esterifying
high FFA oils, a pre-esterification process can be used in which a homogeneous acid
catalysed process is used prior to transesterification (Lam, Lee and Mohamed 2010;
Zhang and Jiang 2008a).
The yield and quality of biodiesel are affected by several pre-esterification and
transesterification reaction parameters such as the quantity of alcohol, reaction
temperature, FFA content of the oil and the type and concentration of catalyst (Balat
and Balat 2008; Demirbas 2008a). For the stoichiometric transesterification reaction,
three moles of methanol are required per mole of triglyceride to yield three moles of
methyl esters and one mole of glycerol. The theoretical molar ratio of methanol to
triglyceride should, therefore, be 3:1 (Ma and Hanna 1999). However, the ratio of
alcohol to oil used in the reaction is much higher than this to promote complete
conversion of oils to FAME and varies with oil quality and the type of catalyst used.
For example, the molar ratio of alcohol to oil for alkali catalysed reactions is
typically 6:1, and for acid catalysed reactions it may be 15:1 or higher. An increase
in the concentration of catalyst generally increases the conversion of triglycerides
into fatty acid esters (Ma and Hanna 1999). Insufficient catalyst leads to an
incomplete conversion reaction and lower levels of fatty acid esters, whereas excess
catalyst has a negative impact on end product yield, because of the formation of
soaps.
On the other hand, a higher reaction temperature increases the reaction rate and
decreases the reaction time due to the reduction in viscosity of the oils. High reaction
temperatures above optimal levels, however, leads to a decrease in biodiesel yield, as
higher reaction temperatures accelerate the saponification of triglycerides (Leung and
Guo 2006b). Therefore, researches seek to optimise the important reaction
parameters for different biodiesel feedstock in order to achieve an efficient and
economical biodiesel production process.
The availability and price of feedstock are significant factors as feedstock cost
represents approximately 75%–88% of the total biodiesel production cost (Bozbas
162 Chapter 6: Production process optimisation of biodiesel
2008; Haas et al. 2006). However, there are vast areas of grazing (e.g., cleared) and
degraded (e.g., mined) land on which biodiesel crops can be successfully established
for complementing fuel supplies. In a recent study, a number of species have been
found suitable for growth on degraded land which has the capacity for producing a
considerable amount of non-edible oil for biodiesel production (Ashwath 2010b).
Among these species, Beauty leaf has been identified as the most suitable feedstock
for future generation biodiesel (Ashwath 2010b; Jahirul, Brown, Senadeera,
Ashwath, et al. 2013). It is a moderately sized tree that grows between 8–20 m tall
and is most notable for its decorative leaves and fragrant flowers. The tree grows in
tropical and sub-tropical climates close to sea level. It is a moderately quick growing
tree reaching up to 1 m tall within a year. It has also been seen to flourish even with
the presence of weeds and other species, so the plant can be grown in mixed cultures.
The Beauty leaf tree has the ability to produce about 4800 kg of non-edible oil per
year per hectare (Jahirul, Brown, Senadeera, Ashwath, et al. 2013). However, the
potential of Beauty leaf as a source of future generation biodiesel is yet to be
established in part due to a lack of knowledge of its optimum production process.
Response surface methodology (RSM) is a collection of mathematical and statistical
techniques that are useful for modelling, analysis and optimisation problems in
which the response of interest is influenced by several factors (Jeong and Park 2009;
Vicente et al. 1998). In this technique, a well-designed experiment can substantially
reduce the number of tests, and yet provide the essential information required for
process optimisation. RSM uses statistical methods for experimental design to
identify important factors by characterising the response surface using a polynomial
model (Ferella et al. 2010). In the practical application of RMS it is necessary to
develop a statistically valid approximating model for the true response surface. The
relationship of the response variable Y and the dependent variables X1, X2, … Xk, in
the RSM application is generally expressed as in the following equation (Saidur et al.
2008):
, , …………… . . Equation 6-1
Where, E is the noise or error observed in the response Y; and f is the response
surface.
Chapter 6: Production process optimisation of biodiesel 163
This study aims to investigate the effect of several reaction parameters on the
production of biodiesel from high FFA vegetable oil obtained from the Beauty leaf
seed. This study also implemented RSM in developing linear and full quadratic
polynomial equations for predicting FFA and FAME content and predicting the
optimum reaction condition for pre-esterification and transesterification processes.
6.2 MATERIALS AND METHOD
6.2.1 Beauty leaf oil extraction method
The Beauty leaf oil used in this study was obtained through a chemical oil extraction
methods using n-hexane as a solvent. In this process, dried seed kernels were ground
using a blender and coffee grinder to obtain a fine consistency to maximise particle
surface area. The ground kernels were then put into conical flasks into which n-
hexane was added at a ratio of 1.6:1 by weight (n-hexane:seed kernels). The mixture
was given an initial stir to ensure that all kernels were wetted with hexane. The
conical flask openings were covered with aluminium foil and placed on laboratory
scale orbital mixer in a fume hood, and the samples were extracted for at least 8 h
with 150 rpm shaking speed. Following extraction, the hexane/oil mixtures were
collected, filtered and decanted into aluminium foil containers for solvent
evaporation, and placed into the fume hood for 8 to 10 h. Hexane was again added to
the conical flask of kernels, but at a ratio of 8:1 (weight) for the second extraction,
and a similar procedure was followed for recovery of the oil. When it was
determined that the hexane had been fully evaporated, the raw Beauty leaf oil was
collected.
6.2.2 Analysis methods
FFA content of the Beauty leaf oil was analysed using D5555-95 (2011) standard test
method. Ester content of the FAME was analysed by gas chromatography and flame
ionisation detection (GC-FID) in accordance with EN 14103 standards. The gas
chromatograph (GC) was a Hewlett-Packard 6890 System fitted with Varian
164 Chapter 6: Production process optimisation of biodiesel
Select™ 30 m × 0.32 mm × 0.25 µm columns. Oil density and surface tension were
analysed following ASTM D1298 and ASTM D971-12 standard test methods using a
KSV Sigma 702 Tensiometer Viscosity was measured using Brookfield DV-III
Rheometer and following the ASTM D445 standard test method. The fatty acid
compositions of the oils were analysed using a Hewlett Packard Plus 6890 series GC-
FID and a capillary column of acidified polyethylene glycol (HP-INNOWax
19091N-133, 30 m × 250 μm × 0.25 μm).
6.2.3 Pre-esterification and transesterification methods
Both acid-catalysed pre-esterification and base-catalysed transesterification were
conducted in a 500 mL triple neck bottom flask reactor (Figure 6-1a). An oil quantity
of 40 g was used for the acid-catalysed pre-esterification experiments, and 30 g was
used for each base-catalysed transesterification trial. For each experiment, oil was
carefully transferred into the reaction flask and preheated in an oil bath to the
reaction temperature. For acid-catalysed esterification trials, sulphuric acid (H2SO4)
was used as catalyst. The sulphuric acid and methanol solution were freshly prepared
and added to the preheated oil, and the mixture was agitated for two hours. At the
completion of the two hours, the mixture was centrifuge in a self-sanding tube for 7
min to separate the methanol-water and esterified oil phases as shown in Figure 6-1b.
The majority of the excess methanol, sulphuric acid and impurities were separated
into the top phase. The bottom phase containing the oil was collected for base-
catalysed trans-esterification.
Chapter 6: Production process optimisation of biodiesel 165
Figure 6-6-1: (a) Esterification and transesterification reactor; (b) Layer of
Methanol-Water (top) and oil (bottom) after acid-catalysed pre-esterification; (c)
Layer of Beauty leaf oil methyl ester (top) and glycerol (bottom) after base-catalysed
Trans-esterification.
Table 6-1: Experimental range and levels of independent variables
Variables Unit Symbol coded Range & levels
−1 0 1
Acid-catalysed pre-esterification
MeOH: Oil mole M 10:1 20:1 30:1 H2SO4 wt% oil C 5 10 15
Temperature °C T 45 60 75
Base-catalysed transesterification
MeOH: Oil mole M' 4:1 6:1 8:1 CH3ONa wt% oil C' 0.6 0.8 1
Temperature °C T' 45 60 75
Table 6-2: Coded experimental design
Run
Acid-catalysed pre-esterification
Base-catalysed trans-esterification
M C T M' C' T'
1 0 −1 1 0 0 0 2 −1 −1 0 1 0 −1 3 −1 0 −1 −1 −1 0 4 1 −1 0 −1 1 0 5 0 0 0 −1 0 −1 6 1 1 0 −1 0 1 7 0 1 1 1 1 0 8 0 0 0 0 1 1 9 0 0 0 0 −1 −1
10 0 −1 −1 0 −1 1
(a) (b) (c)
166 Chapter 6: Production process optimisation of biodiesel
11 0 1 −1 1 0 1 12 −1 0 1 1 −1 0 13 1 0 −1 0 0 0 14 −1 1 0 0 0 0
15 1 0 1 0 1 -1
In the based-catalysed transesterification trials, sodium methoxide (NaOCH3) was
used as a catalyst with a reaction time of 1.5 h. Similarly to the acid-catalysed pre-
esterification trials, the phases of the transesterification product were separated using
a centrifuge and the bottom layer drained using a separation funnel as shown in the
Figure 6-1c. The top layer containing Beauty leaf methyl ester was collected for
analysis.
Experiments were carried out according to a Box-Behnken response surface design
which involves 3 factors and requires 3 levels and a total of 15 runs. The factors and
the ranges and levels used in this study are shown in Table 6-1. The Minitab 16
statistical software package was used to randomly generate runs orders of the
experiments which are shown in Table 6-2.
Since the key focus of acid-catalysed esterification reaction is on reducing the free
fatty acid content to be <3–5 wt%, a full quadratic model was used for statistical
analysis in order to correlate the %FFA with the operating variables. The form of the
full quadratic model for the first step is as shown in Equation (6-2):
2 2 20 1 1 2 2 3 3 1,2 1 2 1,3 1 3 2,3 2 3 1,1 1 2,2 2 3,3 3Y X X X X X X X X X X X X
………………………..Equation 6-2
Where, Y is the %FFA; β0 is a constant; β1, β2, β3 are regression coefficients and X1,
X2, X3 are independent variables. For the base-catalysed transesterification reaction,
the same form of the full quadratic model is used; however, in this case Y is the ester
content of the biodiesel (%).
Chapter 6: Production process optimisation of biodiesel 167
6.3 RESULTS AND DISCUSSION
6.3.1 Beauty leaf oil characterisations
The quality of Beauty leaf oil has been characterised in terms of chemical
composition and physical properties in order to identify its suitability as a feedstock
for biodiesel for diesel engine combustion. The results of analyses of the chemical
composition and properties of the crude Beauty leaf oil used in this study are shown
in Tables 6-3 and 6-4. The compositional analysis shows that Beauty leaf oil contains
high levels of stearic (C18:0), oleic (C18:1) and linoleic (C18:2) acids (Table 6-3).
This indicates the potential for high combustion quality and hence suitability of
Beauty leaf oil as a fuel (Refaat 2009b). However, kinematic viscosity and surface
tension are significantly higher than other oils which may lead to poor atomisation
and volatility characteristics. Therefore, Beauty leaf oil may not be suitable as fuel
for direct use in conventional diesel engines (Balat and Balat 2008).
Table 6-3: Fatty acid composition of Beauty leaf oil
Fatty Acid Weight percentage
Palmitic, C16:0 13.66 Palmitoleic, C16:1 0.24 Heptadecanoic, C17:0 0.15 Heptadecanoic, C17:1 0.06 Stearic, C18:0 16.55 Oleic, C18:1 42.48 Linoleic, C18:2 25.56 Linolenic, C18:3 0.20 Arachidic, C20:0 0.87 Arachidonic, C20:1 0.23
Table 6-4: Properties of Beauty leaf oil.
Properties Values
Density, kg/m3 936 Surface tension (mN/s) 35.6 Kinematic Viscosity @40 °C, cSt 40.05 Free fatty acid (wt%) 12
168 Chapter 6: Production process optimisation of biodiesel
On the other hand, the FFA content of Beauty leaf oil was 12% (w/w) of oil (Table 6-
4) which is much higher than the recommended FFA content of vegetable oil for the
base catalysed transesterification (Dorado et al. 2002; Ramadhas, Jayaraj and
Muraleedharan 2005). To overcome this high FFA level, biodiesel production was
conducted in two processing steps as described. For both pre-esterification and trans-
esterification, final acid value and production yield has been optimised, and a
statistical production model has been developed.
6.3.2 Acid-catalysed pre-esterification
Table 6-5 summarises the experimental conditions and results from each pre-
esterification experimental run. The results indicate a significant reduction in FFA
content of the Beauty leaf oil following acid-catalysed pre-esterification. FFA
content of the pre-esterified samples ranged from 3.25 to 1.83. The minimum FFA
content resulted from the esterification condition with 30:1 MeOH to oil molar ratio,
10% weight concentration of catalyst and 75 °C reaction temperature (Test 15). An
FFA content of less than 2% was also achieved for Test 6 (30:1 MeOH to oil molar
ratio, 15% weight concentration of catalyst and 60 °C reaction temperature). Based
on the experimental results, a linear and a quadratic equations have been developed
using Minitab 16 software in order to predict the FFA percentages as a function of
methanol to oil molar ratio, catalyst concentration and reaction temperature in acid
catalyst esterification. The developed quadratic models equations are shown in
Equations (6-3) and (6-4):
FFA(%) 3.97617 0.0528 0.0055 0.00542M C T ……………..Equation 6-3
2 2 2
FA(%) 1.493 0.00498 0.02487 0.06596 0.000085 0.0007083
0.0002233 0.0001554 0.001558 0.000458
M C T MC MT
CT M C T
……….Equation 6-4
Chapter 6: Production process optimisation of biodiesel 169
Table 6-5: Experimental conditions and results for acid-catalysed pre-esterification
Test MeOH: Oil Molar
Ratio H2SO4 Conc.
(wt%) Temp (°C) FFA (wt%)
1 20 5 75 2.39 2 10 5 60 3.25 3 10 10 45 2.88 4 30 5 60 2.10 5 20 10 60 2.61 6 30 15 60 1.97 7 20 15 75 2.39 8 20 10 60 2.61 9 20 10 60 2.53
10 20 5 45 2.61 11 20 15 45 2.67 12 10 10 75 3.02 13 30 10 45 2.12 14 10 15 60 3.10 15 30 10 75 1.83
Table 6-6 summarises the regression coefficient generated using Minitab 16
software. The significance of each coefficient in this equation was evaluated by
ANOVA using Minitab software in terms of the p-value. Low p-value indicates that
the corresponding coefficient is significant. In the linear model shows that the
methanol to oil ratio (MeOH:Oil) was the most significant with p-value of 0 followed
by temperature and catalyst concentration (H2SO4) with the p-value of 0.07 and
0.511 respectively. For the full quadratic model, temperature was the most
significant with p-value of 0.069. Following this, the interaction effect between
methanol to oil molar ratio and temperature was most significant with p-value of
0.08. Finally, the quadratic effect of temperature was most significant with p-value of
0.097. Table 6-6 also shows that all p-values were fairly high, considerably above
0.05 which infer that the coefficients in full quadratic model are not statistically
insignificant in 95% confidence interval.
The accuracy of the prediction model obtained by the regression analysis was
verified by a scattered diagram (Figure 6-2), where experimental results for FFA
were compared to predicted values from the model. In linear model, the regression
170 Chapter 6: Production process optimisation of biodiesel
coefficient (R2) and the adjusted regression coefficient (R2 (adj)) of 0.983 and 0.949
shows a good fit between actual and predicted results whereas in full quadratic
model, those parameters were 0.981 and 0.949 respectively. Therefore, it is apparent
that the linear model is statistically more appropriate than that of full quadratic
model. Moreover, the full quadratic model is over-specified as none of the
coefficients are significant in 95% confident interval.
Table 6-6: Regression coefficients for %FFA prediction
Predictor Linear Full quadratic
Coefficient p-value Coefficient p-value
Constant 3.97617 0 1.493 0.212 MeOH:Oil (M) −0.0528 0 −0.00498 0.874
H2SO4 (C) −0.0055 0.511 −0.02487 0.695 Temp (T) −0.00542 0.07 0.06596 0.069
MeOH:Oil × H2SO4 (MC) 0.000085 0.934 MeOH:Oil × Temp (MT) −0.0007083 0.080
H2SO4 × Temp (CT) −0.0002233 0.744 MeOH:Oil × MeOH:Oil (M2) −0.0001554 0.771
H2SO4 × H2SO4 (C2) 0.001558 0.476
Temp × Temp (T2) −0.000458 0.097
Figure 6-6-2: Scatter diagram of experimental FFA (%) and predicted FFA (%) of a
linear model.
Chapter 6: Production process optimisation of biodiesel 171
(a)
(b)
Figure 6-6-3: Response surface of FFA content against
(a) methanol to oil molar ratio and reaction temperature at 10% acid catalyst
(H2SO4); (b) against catalyst concentration and reaction temperature at 30:1 methanol to oil molar ratio.
172 Chapter 6: Production process optimisation of biodiesel
Figures 6-3a and 6-3b show the surface plots generated using the model equation on
the effect of each variable on FFA content. Methanol to oil molar ratio has a strong
effect on FFA reduction which is evident in Figure 6-3a. One the other hand, H2SO4
concentration has only a minimal impact on FFA reduction within the range of
catalyst concentrations used, suggesting low linear effect on FFA. In addition, FFA
content increases correspondingly after 10 wt% H2SO4, implying that further
increase in H2SO4 concentration will have adverse effects. This might be the result of
oil decomposition at high acid concentrations. Reaction temperature has a small
effect on FFA content at lower values and effect on FFA reduction is significant after
65 °C, which was more evident in Figure 6-3b.
Table 6-7: Experimental data for base-catalysed trans-esterification.
Run MeOH: oil molar ratio NaOCH3 (wt%) Temp (°C) FAME
(%)
1 6 0.8 60 89.21 2 8 0.8 45 88.51 3 4 0.6 60 74.15 4 4 1.0 60 85.51 5 4 0.8 45 81.92 6 4 0.8 75 78.02 7 8 1.0 60 89.41 8 6 1.0 75 87.70 9 6 0.6 45 64.18 10 6 0.6 75 75.76 11 8 0.8 75 84.78 12 8 0.8 60 63.02 13 6 0.8 60 87.72 14 6 0.8 60 87.20 15 6 1.0 45 90.76
6.3.3 Base-catalysed transesterification of pre-esterified Beauty leaf oil
All samples produced from the acid-catalysed esterification process were thoroughly
mixed to produce a homogenous feedstock for trans-esterification. The FFA content
of the mixture was found to be 2.46% (w/w). Similarly to the acid-catalysed pre-
esterification trials, 15 experimental runs were undertaken based on Box-Behnken
design as shown in Table 6-2. The results obtained from those experiments are
shown in Table 6-7. In the transesterification experiments, the ester content of the
Chapter 6: Production process optimisation of biodiesel 173
FAME ranged from 63.02% to 90.76% with the highest content resulting from
reaction conditions with 6:1 methanol to oil molar ratio, 1 wt% NaOCH3 and 45 °C
temperature.
As from the experimental data processed using Minitab 16 software to generate the
linear and quadratic model for statistical prediction of ester content as a function of
methanol to oil molar ratio, catalyst concentration and reaction temperature. The
linear and quadratic model equation resulting from this is shown in Equations (6-5)
and (6-6) respectively and Table 6-8 summarises the resulting regression coefficients
and corresponding p-value:
%FAME 40.9817 0.3825 ' 47.6688 ' 0.00742M C T ………Equation 6-5
2 2 2
%FAME 129.28 2.256 ' 339 ' 1.817 9.392 ' ' 0.0014 ' '
1.2203 ' ' 0.7892 ' 171.57 ' 0..007011 '
M C T M C M T
C T M C T
…………………..Equation 6-6
Table 6-8: Regression coefficients for FAME (%) prediction
Predictor Linear Full quadratic
Coefficient p-value Coefficient p-value
Constant 40.9817 0.008 −129.28 0.141
MeOH:Oil (M') 0.3825 0.727 2.256 0.809
NaOCH3 (C') 47.6688 0.001 339 0.021
Temp (T') 0.00742 0.959 1.817 0.239
MeOH:Oil × NaOCH3 (M'C') 9.392 0.133
MeOH:Oil × Temp (M'T') 0.0014 0.985
NaOCH3 × Temp (C'T') −1.2203 0.141
MeOH:Oil × MeOH:Oil (M'2) −0.7892 0.0207
NaOCH3 × NaOCH3 (C'2) −171.57 0.025
Temp × Temp (T'2) −0.007011 0.502
The regression coefficient (R2) and the adjusted regression coefficient (R2 (adj)) of
linear model were 0.6465 and 0.55 demonstrated that the linear model may not be
suitable for estimate FAME in given reaction condition. Whereas full quadratic
model with 0.9224 of regression coefficient (R2) and 0.8413 of adjusted regression
174 Chapter 6: Production process optimisation of biodiesel
coefficient (R2 (adj)) shows better model for FAME estimation. Figure 6-4 shows the
accuracy of the prediction model in a scattered plot between experimental and
predicted ester contents. All points are close to straight line demonstrate a good
agreement between experimental results and those ones calculated by the model.
Figure 6-6-4: Scatter diagram of experimental and calculated FAME (%) of full
quadratic model.
More detailed analysis of the effect of base-catalysed transesterification reaction
parameters on Beauty leaf FAME ester content are shown in Figures 6-5a and 6-5b.
These figures predict that an optimal methanol to oil molar ratio would be 7.5:1,
however, further increase in methanol would not have a positive effect on ester
content. On the other hand, NaOCH3 concentration has a strong effect on ester
content of the FAME with corresponding increment with agreement in terms of
linear and quadratic effects. Temperature had a less significant effect than methanol
to oil molar ratio and catalyst concentration. Figure 6-5 shows that the optimum
temperature of transesterification was 65 °C. These figures illustrated that although
all parameters are not statistically significant at 95% confident level but the
relationship still contains useful information for some biodiesel production purposes.
Chapter 6: Production process optimisation of biodiesel 175
(a)
(b)
Figure 6-6-5: Response surface ester content against catalyst concentration vs. (a)
methanol to oil molar ratio at 55 °C; (b) reaction temperature at 7.5:1 methanol to oil
molar ratio.
176 Chapter 6: Production process optimisation of biodiesel
6.4 CONCLUSIONS
A response surface method based a Box-Behnken design was employed to determine
a feasible experimental plan to optimise the Beauty leaf oil to biodiesel conversion
procedure. Due to the high FFA content of Beauty leaf oil (12 wt%), a two-step
process was employed utilising sulphuric acid catalysed pre-esterification followed
by sodium methoxide catalysed trans-esterification. Effects of reaction parameters
such as methanol to oil molar ratio, catalyst loading and reaction temperature were
statistically investigated on the reduction of FFA content in pre-esterification and
ester content in trans-esterification. The optimal conditions for pre-esterification
were 30:1 methanol to oil molar ratio, 10 wt% sulphuric acid catalyst and 75 °C
reaction temperature which reduced the FFA content to 1.8 wt%. With the aid of
statistical modelling, the predicted optimal conditions for transesterification
methanol to oil molar ratio, catalyst concentration and reaction temperature were
7.5:1, 1% and 55 °C respectively. Based on these conditions, the highest achievable
ester content of FAME predicted by the model was found to be approximately 93%.
However a higher result may be achievable by future lowering FFA content of
Beauty leaf oil. In terms of a linear effect on FFA reduction for the first step,
methanol to oil molar ratio was found to be highly significant and reaction
temperature moderately significant. For trans-esterification, catalyst concentration
was found be the most dominant variable in achieving high ester contents. The
limitation of the developed response surface model is that all the p-values are greater
than 0.05. Therefore, the developed models might be over-specified and some that
terms can be omitted. However, the information contained in the model and
experiment in this study is very significant in industrial biodiesel production.
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 177
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
Particle emissions from biodiesels with different
physicochemical properties
M. M. Rahman, A. M. Pourkhesalian , M. I. Jahirul , S. Stevanovic , P. X. Pham,
H. Wang , A.R. Masri , R. J. Brown and Z. D. Ristovski
Publication: Journal of Fuel, Vol. 134, pp. 201-208, 2014
http://dx.doi.org/10.1016/j.fuel.2014.05.053 0016-2361/2014 Elsevier Ltd. All rights reserved
Author Contribution
Contributor Statement of Contribution
M. M. Rahman Conducted the experiments, performed the data analysis and drafted the manuscript
A. M. Pourkhesalian Assisted with conducting the experiment and data analysis
M. I. Jahirul Conducted the experiments, performed the data analysis and drafted the manuscript Signature
S. Stevanovic Assisted with conducting the experiment
P. X. Pham Assisted with conducting the experiment
H. Wang Assisted with conducting the experiment
A.R. Masri Supervised the project and revised the manuscript
R. J. Brown Supervised the project and revised the manuscript
Z. D. Ristovski Supervised the project and revised the manuscript
Principal Supervisor Confirmation
I have sighted email or other correspondence from all co-authors confirming
their certifying authorship.
Name
Dr Wijitha Senadeera
Signature
Date
178 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
Abstract
Biodiesels produced from different feedstocks usually have wide variations in their
fatty acid methyl ester (FAME) so that their physical properties and chemical
composition are also different. The aim of this study is to investigate the effect of the
physico-chemical properties of biodiesels on engine exhaust particle emissions.
Alongside with neat diesel, four biodiesels with variations in carbon chain and
degree of unsaturation have been used at three blending ratio (B100, B50, B20) in a
common rail engine. It is found that particle emission increased with the increase of
carbon chain length and degree of unsaturation in FAME. However, for similar
carbon chain length, particle emissions from totally unsaturated biodiesel is found to
be slightly less than that of partially (about 50%) unsaturated biodiesel. Particle size
is also found to be dependent on fuel type. The fuel or fuel mix responsible for
higher PM and PN emissions is also found responsible for lager particle median size.
Particle emissions reduced consistently with fuel oxygen content regardless of the
proportion of biodiesel in the blends, whereas it increased with fuel viscosity and
surface tension only for higher diesel-biodiesel blend percentages (B100, B50).
However, since fuel oxygen content increases with the decreasing carbon chain
length, it is not clear which of these factors drives the lower particle emission.
Rather, overall, it is evident from the results presented here that chemical
composition of biodiesel is more important than its physical properties in controlling
exhaust particle emissions.
Keywords: Biodiesel, particle emissions, fuel physical properties, fuel chemical
composition
Highlights
Four biodiesels were used to investigate their influence on particle emissions.
Particle emission increased with the increase of biodiesel carbon chain length.
Particle emissions reduced consistently with fuel oxygen content.
Particle median size found dependent on the type of fuel used.
Biodiesel chemical composition found more important than physical properties.
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 179
7.1 INTRODUCTION:
Compression Ignition (CI) engines are increasing in popularity due to their higher
thermal efficiency. They power a wide range of land and sea transport as well as
provide electrical power, used in farming, construction and industrial applications.
Tail pipe emissions of diesel engines, especially particulate matter (PM) are still a
matter of concern due to its harmful effects both on human health and the
environment(Brito et al. 2010; Jacobson 2001). Exposure to diesel particulate matter
(DPM) can cause pulmonary diseases such as asthma, bronchitis and lung
cancer(Brito et al. 2010) and because of these adverse effects, the International
Agency for Research on Cancer (IARC) included DPM as carcinogenic to human
health.
The harmful effects caused by DPM are related to both the physical properties and
chemical composition of the particles. The physical properties that influence
respiratory health include particle mass, surface area, mixing status of particles,
number and size distribution (Ristovski et al. 2012). The particles deposit in different
parts of the lung depending on their size. The smaller the particles the higher the
deposition efficiency (Broday and Rosenzweig 2011) and the greater the chance of
them penetrating deep into the lung. The smaller particles stay suspended in the
atmosphere for longer thus have a higher probability of being inhaled and
consequently deposited deep in the alveolar region of the lung. Particle number
governs the ability of particles to grow larger in size by coagulation while particle
surface area determines the ability of the particles to carry toxic substances. Recent
studies reveal that DPM surface area and organic compounds play a significant role
in initiating various cellular and chemical processes responsible for respiratory
disease (Giechaskiel, Alfoldy and Drossinos 2009; Ristovski et al. 2012). In addition
to this, a large fraction of DPM is black carbon, which is considered the second most
potential greenhouse warming agent after carbon dioxide (Jacobson 2001). After
treatment devices (ATD) like diesel particulate filters (DPF) and diesel oxidation
catalysts (DOC) aid in reducing DPM (Herner et al. 2011). Alternative fuels are
another potential emission reducing source (Bakeas, Karavalakis and Stournas 2011).
Of these fuels, biodiesel is considered one of the more promising for diesel engines
180 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
as it (Varuvel et al. 2012; Xue, Grift and Hansen 2011) produces less PM and other
gaseous emissions (Lapuerta, Armas and Rodríguez-Fernández 2008b; Xue, Grift
and Hansen 2011; Surawski, Miljevic, Ayoko, Roberts, et al. 2011). Biodiesel in
diesel engines has the potential to greatly reduce carbon emissions and is a renewable
source of energy.
Biodiesel is a mixture of fatty acid esters with physicochemical properties that
mostly depend on the structure of this molecule. Biodiesel can be produced from a
variety of feedstock sources such as vegetable oil, animal fat, municipal and
industrial waste and some from insects (Salvi and Panwar 2012; Sharma, Singh and
Upadhyay 2008; Morshed et al. 2011; Alptekin, Canakci and Sanli 2012). An
extensive range of fatty acid profiles exist among these feedstocks (Moser 2014),
with some being within the same feedstock; which can be controlled. Physical
properties and chemical composition of biodiesel varies among different feedstocks,
which can have a noticeable influence on engine performance and emissions
(Hoekman et al. 2012; Mccormick, Graboski, Alleman and Harrin 2001).
McCormick et al. (McCormick, Graboski, Alleman, Herring, et al. 2001) reported
constant PM emissions from different biodiesel feedstocks when the density was less
than 0.89 g/cm3 or cetane number was greater than about 45, but increase of NOx
emissions with the increase of biodiesel density and iodine number. In contradiction
to these findings, a difference in particle emissions from biodiesel from different
feedstocks has also been reported (Surawski, Miljevic, et al. 2011a; Allan, Williams
and Rogerson 2008). Lapuerta et al.(2008b) reported a 10% increase of NOx and
20% decrease of particle emissions by unsaturated biodiesel. Benjumea et al.(2011)
found that the degree of unsaturation in biodiesel doesn’t significantly affect the
engine performance but increases smoke opacity and THC emissions. Kravalkis et al.
(2011) reported noticeable influence of biodiesel origin on particle emissions,
especially particles associated with PAH and carbonyl emissions. Very recently
Salamanaca et al.(2012) reported increased PM and HC emissions from biodiesel
that contains more unsaturated compounds that favour soot precursor formation.
There is no distinction however, that exists in the literature, which indicates whether
chemical composition of biodiesel, physical properties or a combination of these is
responsible for this variation in engine performance and emissions. This study
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 181
therefore, aims to investigate the effect of biodiesel physical properties and chemical
composition on engine exhaust particle emissions. It is an extension of the previous
study (Pham et al. 2013) where results from the same experiments were presented for
the engine performance characteristics and emission of pollutants including some
preliminary results for the particle emission, particularly for pure biodiesel. It should
be noted that the results for B100 are reproduced here for comparison purposes.
Furthermore, the paper elaborates on these findings and presents new analysis in
terms of the physico-chemical properties of the fuels and their blends.
Table 7-1: Test engine specification
Model Cummins ISBe220 31
Cylinders 6 in-line
Capacity (L) 5.9
Bore x Stroke (mm) 102 x 120
Maximum power (kW/rpm) 162/2500
Maximum torque (Nm/rpm) 820/1500
Compression ratio 17.3
Aspiration Turbocharged & after cooled
Fuel injection Common rail
After treatment systems None
Emissions certification Euro III
7.2 MATERIALS AND METHODS
7.2.1 Engine and fuel specification
This experimental study was performed in a heavy duty 6 litres, six cylinders,
turbocharged after cooled, common rail diesel engine typically used in medium size
trucks. Test engine is the same as used in Pham at el. (2013). Table 7-1 shows
specification of the test engine. Engine was coupled to a water brake dynamometer,
and both of them are connected to an electronic control unit (ECU). Engine was
operated at 1500 rpm (maximum torque speed) and at 2000 rpm (intermediate
speed), and four different loads 25%, 50%, 75%, & 100% for each engine speed.
182 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
Maximum load at any particular engine speed depends upon the type of fuel used,
therefore for each fuel at first maximum load was measured when engine was in full
throttle for a particular speed. This measured load is then considered as 100% load
for that speed and other loads were determined based upon measured 100% load.
An ultra-low sulphur diesel (sulphur content < 6ppm) and four biodiesels with
different physicochemical properties were used to run the engine. All four biodiesels
were used at three blending ratio i.e. 100% biodiesel (B100), blends of 50% diesel
and 50% biodiesel (B50), and blends of 80% diesel and 20% biodiesel (B20). 7-2
shows the fatty acid profile of used biodiesels as found using gas chromatography
mass spectrometry (GCMS) analysis. Biodiesel samples were analysed using Perkin
Elmer clarus 580GC-MS equipped with Elite 5MS 30m x 0.25mm x 0.25um column
with a flow rate of 1mL/min. Before analysing, each biodiesel was diluted with n-
hexane (1:100 v/v). Initial temperature was 120 0C for 0.5 minutes, then raised to 310 0C for 2 minutes at 10 0C/min and kept at 310 0C for 2 minutes. The mass selective
detector was optimised using calibrating standards with reference masses at m/z (35-
40). Among four biodiesels, C810 is fully saturated and composed of 52% and 46%
caprylic acid and capric acid ester respectively. C1214 is also dominated by saturated
compounds but has comparatively longer carbon chain length fatty acid ester i.e.
48% lauric, 19% % myristic, 10% palmitic and 18% oleic acid ester. On the other
hand both C1618 and C1822 are dominated by long chain unsaturated fatty acid
esters. C1618 is composed of 21% palmitic, 9% stearic, 58% cis-oleic and 10%
linoleic acid ester where C1822 has 10% more oleic and linoleic acid ester. C1822
also has small amount (4%) of trans- oleic acid ester.
Some important properties of all used fuels related to combustion and emissions are
shown in Table 7-3. Among the used biodiesels physical properties varied with the
variation in chemical composition i.e. carbon chain length and degree of
saturation/unsaturation. Viscosity, heating value, iodine value and oxygen content
increased with the increase of carbon chain length and degree of unsaturation, where
saponification value decreased. Density of all four biodiesels was found higher than
diesel, and no trend observed among biodiesels either with the carbon chain length or
degree of unsaturation. Surface tension also increased with the carbon chain length in
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 183
biodiesels, although no significant change was observed with the degree of
unsaturation. For example, there is almost no difference in surface tension between
C1618 & C1822, although C1822 contains much higher percentage of unsaturated
compounds compared to C1618. Surface tension and cetane value of diesel were
found to be lower than all four used biodiesels where calorific value was higher.
Viscosity of diesel was higher than C810 but lower than the rest of the three
biodiesels.
Table 7-2: Fatty acid profile of used biodiesels
Common name Lipid number Biodiesels
C810 C1214 C1618 C1822
Caprylic acid C8:0 52.16 0 0 0 Capric acid C10:0 46.38 0.17 0 0 Lauric acid C12:0 1.38 47.8 0.1 0 Myristic acid C14:0 0 18.89 0.06 0.03 Pentadecylic acid C15:0 0 0 0.03 0.02 Palmitic acid C16:0 0 10.19 21 4.45 Palmitoleic acid C16:1 0 0 0 0.12 Margaric acid C17:0 0 0 0.06 0 Stearic acid C18:0 0 2.55 9.47 2.53 Oleic acid C18:1cis 0 18.53 58.72 68.13 Elaidic acid C18:1trans 0 0 0 3.96 Linoleic acid C18:2 0 1.76 9.98 18.69 Arachidic acid C20:0 0 0.08 0.3 0.49 Gadoleic acid C20:1 0 0 0.24 1.03 Behenic acid C22:0 0 0.03 0.03 0.17 Glycerol 0.08 0 0 0
7.2.2 Exhaust sampling and measurement system
The Dekati ejector diluter was used to partly sample raw exhaust from the engine
exhaust pipe and then dilute it with particle free compressed air. A second Dekati
diluter was connected in series with the first one to further increase the dilution ratio
in order to further decrease concentration. A HEPA filter was used to provide
particle free compressed air for the diluters. The purpose of the dilution was to bring
down the temperature as well as the concentration of gases and PM within the
measuring range of the instruments. Diluted exhaust was then sent to different
gaseous and particle measuring instruments. A CAI 600 series CO2 analyser was
184 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
used to measure the CO2 concentration directly from the raw exhaust. A second CO2
meter (SABLE, CA-10) connected via a three way valve between the two diluters
was used to record the CO2 concentration from the diluted exhaust. Background
corrected CO2 was used as tracer gas to calculate the dilution ratio for each stage.
After first stage dilution, CAI 600 series CLD NOx analyser was used to measure the
NOx. PM2.5 emissions were measured by a TSI DustTrak (Model 8530). DustTrak
readings were converted into a gravimetric measurement by using the tapered
element oscillating microbalance to DustTrak correlation for diesel particles
published by Jamriska et al.(2004). It is worth noting this conversion can introduce
significant uncertainties if the optical properties of particles significantly changes.
The particle number size distribution for C810, C1214, C1618 and their blends was
measured by a scanning mobility particle sizer (SMPS). This SMPS consisted of a
TSI 3080 electrostatic classifier (EC) and a TSI 3025 butanol based condensation
particle counter (CPC). Due to technical problems a new SMPS had to be used for
the reference diesel and C1822. This SMPS system consisted of a 3085 classifier
with a nano-DMA (differential mobility analyser). As the measurement range of the
two SMPS’s used was different, we have used a fitting procedure (see Section 7.3.2)
to recalculate the total PN and make the measurements comparable. A TSI 3089 nm
aerosol sampler (NAS) was used in conjunction with a Tandem Differential Mobility
Analyser (TDMA) to collect preselected particles on Transmission Electron
Microscopic (TEM) grids for morphological analysis. The EC in the TDMA
preselected the size of the particles, which deposited on the TEM grid in the NAS.
An Aethalometer (Magee Scientific) was also connected after second stage dilution
for black carbon (BC) measurement. Results from TEM analysis and BC data will be
published in a separate paper.
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 185
Table 7-3: Important physicochemical properties of tested fuels
Relevant properties Fuels
C810 C1214 C1618 C1822 Diesel
Average formula C9.5H19.7O2 C9.5H19.7O2 C9.5H19.7O2 C9.5H19.7O2 CxHya
Average unsaturation (AU) 0 0.22 0.7892 1.11 ‐ Oxygen content (wt%) 18.72 13.25 10.74 10.83 0 a
Stoichiometric air fuel ratio 11.12 12.05 12.50 12.48 14.5 Relative density(kg/l) 0.877 0.871 0.873 0.879 0.8482 Viscosity (mm2/sec) 1.95 4.37 4.95 5.29 3.148 Surface tension (mN/m) 26.184 28.41 29.9 29.966 26 Cetane value 62.96 65.57 61.06 53.65 48.5 Iodine number 1 max 8 65 105 Saponication value 330 233 195 185 Acid value 0.9 0.4 0.8 0.4 <0.05 Boiling point (0C) 190 >150 165.6 >150 >190 a
Gross Calorific value(MJ/kg)
35.335 38.409 37.585 39.825 44.365
Sulphur content(mg/kg) 0 0 0 0 2.5 a Values with superscripts have been taken from literature(Surawski, Miljevic, et al.
2011c) and (Lapuerta, Armas and Rodriguez-Fernandez 2008).
Figure 7-7-1: Schematic diagram of used engine exhaust measurement system
186 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
7.3 RESULTS AND DISCUSSION
7.3.1 Specific PM emissions
All four biodiesels that were used, disregarding the variations in physical properties
and chemical composition, reduced PM emissions in comparison to petroleum diesel.
Figure 7-2a and b shows brake specific PM emissions at engine operating speeds of
1500 rpm and 2000 rpm, respectively. It was found that as the biodiesel percentage
in the diesel-biodiesel blends increased, PM emissions decreased consistently. The
maximum reduction in PM was observed for 100% biodiesel blends, an observation
common in the literature (Lapuerta, Armas and Rodríguez-Fernández 2008a;
Surawski, Miljevic, et al. 2011b; Xue, Grift and Hansen 2011). Noticeable variations
in PM emissions were also observed among the four biodiesels and their blends. In
the case of using 100% biodiesel, a massive 98% reduction in PM was observed for
biodiesel C810, where C1214, C1618 and C1822 reduced PM 83%, 70% and 76%
respectively. Similar trends in PM emissions were also found for B50 and B20
blends although there was a difference of PM reduction proportion in these. For the
B50 blend, PM reduction among biodiesels C810, C1214, C1618 and C1822 was
88%, 75%, 70% and 76% respectively. B20 was slightly lower, measuring 66%,
57%, 42% and 48% respectively. PM emissions from other tested engine loads are
shown in the appendix (Figure A7-1). Similar trends in PM emissions were also
observed for these loads at 2000 rpm engine speed, although at 1500 rpm, PM
emissions from B20 (C1618) were found to be slightly higher than for the diesel.
These variations in PM emissions among biodiesels could be due to either their
chemical composition or their physical properties. Among the biodiesels, PM
emissions increased consistently with biodiesel carbon chain length with the
exception of C1822. This blends carbon chain length was similar to C1618 but its
degree of unsaturation was higher and its PM emissions were less. Pinzi et al.(2013)
also reported reduction of PM emission with the increase of degree of unsaturation
but same carbon chain length in FAME. Opposing observations were also reported in
the literature, which suggests that unsaturated compounds have a tendency to act as
soot precursor (Salamanca et al. 2012; Benjumea, Agudelo and Agudelo 2011). In
addition, important physical properties of C1822 in regards to particle emissions i.e.
viscosity and surface tension were also higher. This slight reduction in PM emissions
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 187
from C1822 might be attributed to its high iodine or low cetane value. Fuels with low
cetane value undergo prolonged premixed combustion phases that are responsible for
less soot formation. In addition, NOx emissions from C1822 were highest among the
fuels that were favourable for soot oxidation. This could also be responsible for
comparatively low PM emissions from C1822.
Figure 7-7-2: Brake specific PM emission at
(a) 1500 rpm 100% load and (b) 2000 rpm 100% load. The PM emissions shown are
calculated based on DustTrak measurement.
7.3.2 Specific PN emissions
Variations that were observed in PN emissions were similar to the fuels used. There
were however, slight differences in proportion compared to PM emissions. All PN
emissions were calculated for the size range from 10.2-514 nm. As the measurements
for neat diesel and C1822 were done using the nano-DMA, in the size range from
4.6nm-156nm, a fitting procedure was used to recalculate the PN concentration to the
same size range as used in the other measurements (Heintzenberg 1994). As shown
in Figure 7-3, PN emissions from B100 were found to be lower than diesel for all
biodiesels. Among the biodiesels used, C810 reduced PN most and C1618 reduced
188 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
PN the least compared to neat diesel, at 90% and 20% respectively. Reductions from
C1214 and C1822 were measured at 60% and 35% respectively. For B50, in the case
of C810, C1214 and C1822, the PN emissions remained lower than diesel although
C1618 increased approximately 10%. Similar to B100, the lowest PN emissions were
observed with C810 for B50 with C1214 and C1822 following the trend for B100.
PN emissions from C1214 increased 15% with a large standard error at 2000 rpm,
while at 1500 rpm it remained almost same to C1822. Apart from B100 and B50, PN
emissions from B20 were found to be slightly less than diesel and almost the same
among the biodiesels with the exception of around 15% increase from diesel at 1500
rpm. Brake specific PN emissions from other engine loads at both rpm are shown in
appendix (Figure A7-2). PN emissions from all other lodes and engine speeds
showed a similar trend with the exception of 1500 rpm 50% load where PN emission
from B20 appeared to have a different trend as compared to the rest of the results.
Figure 7-7-3: Brake specific PN emissions at 1500 rpm 100% load (a) and 2000 rpm
100% load (b).
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 189
7.3.3 Particle number size distribution
Particle size distribution (PSD) was always found to be unimodal with a single peak
in the accumulation mode despite the variations in the fuel and the condition of
engine operation (appendix: Figures A7-3 to A7-5). Variations in PSD among the
biodiesels were more prevalent for B100, followed by B50. Comparatively, PSD
from B20 was found to be similar to petroleum diesel regardless of the variations of
biodiesel. Another important feature is that biodiesel reduced a higher proportion of
large particles (mobility diameter >100 nm) compared to nanoparticles (mobility
diameter <50 nm). Nanoparticle emissions from biodiesel however, did not exceed
that of diesels, which have been reported in few studies (Shi et al. 2010; Surawski,
Miljevic, et al. 2011a). The presence of second peak (nucleation mode) in PSD is
responsible for increased nanoparticle emission which we didn’t observe in this
study. Presence of excessive volatiles and semi volatiles in exhaust which partitioned
into particles upon cooling down are the primary contributor to nucleation mode
peak. In addition, impurities in biodiesel especially glycerol doesn’t undergo
complete combustion due to their high viscosity, poor atomisation and mixing
property. They form partially oxidised volatiles and semi volatiles which can be a
major contributor to nucleation mode peak. Biodiesels used in this study was free
from glycerol and other impurities, which might facilitate the absence of nucleation
mode peak in PSD.
7.3.4 Particle median size
Particle size also varied among used fuels in a similar way to PM and PN. In case of
diesel, the median size of the particles in the SMPS size distribution was 61 nm and
56 nm at 1500 and 2000 rpm respectively. For 100% biodiesel, particle median size
was always found to be smaller than for neat diesel and diesel-biodiesel blends
(Figure 7-4a and 7-4b). Among four used biodiesels, C810 produced the smallest
particle median size i.e. 40 nm and 43 nm at 1500 and 2000 rpm respectively,
followed by C1822 which was 53 nm and 44 nm. C1214 and C1618 gave almost the
same particle size as neat diesel with slight difference between two engines operating
speeds. Particle size from B50 was found to be larger than B100 but smaller than
B20 blends. Interestingly, particle emitted from all B20 blends were found to be
larger than diesel with the largest particle median size observed for C1618 B20
190 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
blend. Similar trends in particle size were also observed at other engine loads which
are shown in appendix (Figures A7-6 and A7-7). To gain some insight of the
variation in particle sizes among different biodiesels and its blends, particle median
size from all measurement was plotted against the particle number concentrations. As
can be seen on Figure 7-4c a moderate positive correlation, with a Pearson
correlation coefficient of 0.61 was found between particle median size and total
number concentration. This indicates that total PN number concentration through
coagulation could be one of the key parameters influencing the overall particle size.
Higher the particle number emissions, larger the particle size can be. The other
factors may be the biodiesel viscosity, surface tension and especially oxygen
contents which ensure the presence of more oxygen functional groups on the surface
of particles responsible for enhance particle oxidation and subsequent size reduction
(Wang et al. 2009; Zhu, Cheung and Huang 2011).
Figure 7-7-4: Variations in particle median size among used fuels at 1500 rpm 100%
load (a) and 2000 rpm 100% load (b), while (c) shows particle median size variation
with total number concentration.
7.3.5 NOx emissions
NOx emissions were also found to be dependent on biodiesel carbon chain length
and degree of unsaturation (see Figure. 7-5). Biodiesels with higher degree of
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 191
saturation and shorter carbon chain length emitted less NOx than biodiesels with
relatively longer carbon chain and higher degree of unsaturation. Interestingly NOx
emissions from C810 and C1214 were found to be less than for diesel especially at
higher blend percentages. An interesting trend in NOx emissions was also observed
among the different blends used. The usual trend, as reported in most of the
literature, is to observe the increase in NOx emissions with the increase of biodiesel
blend percentage (Bakeas, Karavalakis and Stournas 2011; Hoekman and Robbins
2012). While we have observed a similar trend for long chained biodiesels with
higher degree of unsaturation i.e. C1618 and C1822, however for saturated and short
chained biodiesels i.e. C810 the opposite trend was observed. NOx formation mostly
depends on the duration of premixed combustion phase and in cylinder temperature.
Biodiesels with higher degree of unsaturation have low cetane number, which leads
to prolonged premixed combustion favourable for thermal NOx formation. So the
higher NOx emissions from C1618 and C1875 are expected and are due to their
higher unsaturation as well as their higher heating value. On the other hand higher
degree of saturation, higher cetane number and lower heating value of C810 and
C1214, may cause shorter premixed combustion and lower in cylinder temperature
responsible for less NOx emissions. Therefore the discrepancy in reported (Redel-
Macías et al. 2012) NOx emissions, among different biodiesel studies in the
literature, may be due to biodiesel chemical composition. The generally adopted
concept of increase in NOx emissions for biodiesels does not always stand. Rather,
whether biodiesels will increase or reduce NOx emissions depends upon their
chemical composition.
192 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
Figure 7-7-5: Brake specific NOx emission at (a) 1500 rpm 100% load and (b) 2000
rpm 100% load.
7.3.6 Influence of fuel physical properties and chemical composition on particle emissions
To understand what is the relative influence of fuel physical properties as well as
chemical composition on particle emissions, PM and PN emissions for all used fuels
were plotted against fuel viscosity, surface tension and oxygen contents. Variations
in PM and PN emissions with fuel viscosity, surface tension and oxygen contents at
100% load are shown in Figure 7-6, where the other loads are shown in appendix
(Figures A7-8 to A7-10). As shown in Figure 7-6, particle emissions increased with
the increase of fuel viscosity and surface tension but only within a specific blend. For
higher blend percentages (B50 and B100) there was almost a linear relationship
between surface tension, viscosity and particle emissions. On the other hand, for the
same viscosity and surface tension, particle emissions also found significantly
different among fuel/fuel mix. It is evident from the literature, both viscosity and
surface tensions have noticeable influence on fuel atomisation process (Ejim, Fleck
and Amirfazli 2007), which is a key parameter relative to in-cylinder soot formation.
Lower the viscosity and surface tension of fuel, more easily they evaporate, atomise
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 193
and mix into in-cylinder air, and more complete their combustion are (Lee et al.
2002; Chen et al. 2013). In this case, the used engine was employed with common
rail injection system where fuel injection pressure was high (around 200 bars). Such
a high injection pressure might be minimised the effect of small variation in fuel
viscosity and surface tension on fuel atomisation and subsequent particle emissions.
On the other hand, a more consistent negative relationship was observed between
fuel oxygen content and particle emissions. This relationship did not depend on the
blend percentage. Similar reduction in particle emissions with fuel oxygen content
has also been reported in the literature (S.S. Gilla 2011; Rahman et al. 2013; Xue,
Grift and Hansen 2011). Therefore, this is a clear indication that the fuel chemical
composition, particularly the oxygen content, could be more important than its
physical properties in terms of engine exhaust particle emissions.
Figure 7-7-6: Variation in specific PM and PN emissions with used fuel surface
tension, viscosity and oxygen content
((a), (b), (c) for PM and (d), (e), (f) for PN), Ordinate of (a), (b), (c) and (d), (e), (f) are same where abscissa of (a), (d) and (b), (e) and (c), (f) are same.
194 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
7.3.7 Comparison of engine performance and particle emissions among used biodiesels
A comparison between engine performance parameters and particle emissions is
shown in Figure 7-7. In this figure, the vertical axis represents the percentage change
of engine power, break specific fuel consumption and specific PM/PN emissions
while the horizontal axis indicates biodiesel proportion in the blends. Neat diesel was
used as a reference fuel to calculate the percentage changes. There is a significant
difference among all four biodiesels used. For example, C810 provides the highest
reduction in particle mass and number but the penalty for that is also highest, around
25% reduction in engine brake power and an additional 25% increase of specific fuel
consumption due to that reduced engine power. This fuel and power penalty is lowest
for C1618 but particle mass and number reduction is also the lowest in this fuel
blend. Therefore it is necessary to make a trade-off between particle emission
reduction, fuel and power penalty, ensuring maximum benefit; not just for emission
levels but for engine power and fuel economy as well. Considering the
aforementioned factors, C1822 seems to have advantage over the rest of the fuels, as
it maintains the lowest power and fuel penalty regardless of the blending ratio to
diesel and a reasonable reduction in engine exhaust particle emissions. The evidence
suggests that biodiesels with a longer carbon chain length and higher degree of
unsaturation might be a solution to reduce particle emissions to a certain extent with
less fuel and engine power penalty.
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 195
Figure 7-7-7: Comparison of engine performance (power, BSFC) and particle
emissions (PM, PN) among biodiesels and their blends where petroleum diesel was
used as a reference fuel.
7.4 CONCLUSIONS
In conclusion, biodiesel fuels with shorter carbon chain lengths and higher degrees of
saturation have more potential to decrease engine exhaust particle emissions. With
the increase of carbon chain length and degree of unsaturation, particle emissions
also increase. Particle size also depends on type of fuel used. Fuel or fuel mix
responsible for higher PM and PN emissions was also found to be to have a larger
particle median size. This indicates that Coagulation plays a role in overall engine
exhaust particle size. Particle emissions increase linearly with fuel viscosity and
surface tension for higher diesel- biodiesel blend percentages (B100, B50). It reduces
consistently with fuel oxygen content regardless of the proportion of biodiesel in the
blends. High fuel injection pressure by common rail injection systems might
minimise the effects of small variation in fuel viscosity and surface tension on
particle emissions. Fuel oxygen content increases with the decrease of FAME carbon
chain length; therefore it is not clear whether FAME carbon chain length or oxygen
content is the driving force that decreases particle emission. The results support the
view that chemical composition of biodiesel is more important than its physical
properties in regards to reducing engine exhaust particle emissions.
Power BSFC PM PN
-40
-30
-20
-10
0
10
20
30
40
50
60
70
80
90
100
Incr
ease
d(%
)R
educ
ed(%
)
B100
Power BSFC PM PN
B50
Power BSFC PM PN
B20
C810 C1214 C1618 C1822
196 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
APENDIX A7
Particle emissions from biodiesels with different physical properties and chemical
composition
Figure A7-1: Brake specific PM emissions at 75%, 50% and 25% loads respectively
while the engine operated at 1500 and 2000 rpm respectively.
C810 C1214 C1618 C18220.00
0.02
0.04
0.06
0.08
0.10
Diesel
Bra
ke s
peci
fic P
M (
g/kW
-hr)
Biodiesel type
B20 B50 B100
1500 rpm 75% load
C810 C1214 C1618 C18220.00
0.01
0.02
0.03
0.04
0.05Diesel
2000 rpm 75% load
Bra
ke s
peci
fic P
M (
g/kW
-hr)
Biodiesel type
B20 B50 B100
C810 C1214 C1618 C18220.00
0.02
0.04
0.06Diesel
1500 rpm 50% load
Bra
ke s
peci
fic P
M (
g/kW
-hr)
Biodiesel type
B20 B50 B100
C810 C1214 C1618 C18220.00
0.01
0.02
0.03
0.04
0.05
Diesel
2000 rpm 50% load
Bra
ke s
peci
fic P
M (
g/kW
-hr)
Biodiesel type
B20 B50 B100
C810 C1214 C1618 C18220.00
0.02
0.04
0.06
0.08
0.10
0.12
Diesel
1500 rpm 25% load
Bra
ke s
peci
fic P
M (
g/kW
-hr)
Biodiesel type
B20 B50 B100
C810 C1214 C1618 C18220.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
Diesel
2000 rpm 25% load
Bra
ke s
peci
fic P
M (
g/kW
-hr)
Biodiesel type
B20 B50 B100
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 197
Figure A7-2: Brake specific PN emissions at 75%, 50% and 25% loads respectively
while the engine speed was 1500 and 2000 rpm
C810 C1214 C1618 C18220.0
3.0x1013
6.0x1013
9.0x1013
1.2x1014
1.5x1014
1.8x1014
Bra
ke s
peci
fic P
N(#
/kW
-hr)
Biodiesel Type
B20 B50 B100
1500 rpm 75% load
Diesel
C810 C1214 C1618 C18220.0
2.0x1013
4.0x1013
6.0x1013
8.0x1013
1.0x1014
1.2x1014
1.4x1014
Bra
ke s
peci
fic P
N(#
/kW
-hr)
Biodiesel Type
B20 B50 B100
2000 rpm 75% load
Diesel
C810 C1214 C1618 C1822
2.0x1013
4.0x1013
6.0x1013
8.0x1013
1.0x1014
1.2x1014
1.4x1014
Bra
ke s
peci
fic P
N(#
/kW
-hr)
Biodiesel Type
B20 B50 B100
1500 rpm 50% load
C810 C1214 C1618 C18220.0
4.0x1013
8.0x1013
1.2x1014
1.6x1014
2.0x1014
Bra
ke s
peci
fic P
N(#
/kW
-hr)
Biodiesel Type
B20 B50 B100
2000 rpm 50% load
Diesel
C810 C1214 C1618 C18220.0
5.0x1013
1.0x1014
1.5x1014
2.0x1014
2.5x1014
3.0x1014
Bra
ke s
peci
fic P
N(#
/kW
-hr)
Biodiesel Type
B20 B50 B100
1500 rpm 25% laod
Diesel
C810 C1214 C1618 C18220
1x1014
2x1014
3x1014
4x1014
5x1014
Bra
ke s
peci
fic P
N(#
/kW
-hr)
Biodiesel Type
B20 B50 B100
2000 rpm 25% load
Diesel
198 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
Figure A7-3: Particle number size distribution for B100, B50, and B20 at 1500 rpm
100% load (a, b, c respectively) and 2000 rpm 100% load (d, e, and f respectively)
Figure A7-4: Particle number size distribution for B100, B50, and B20 at 1500 rpm
75% load (a, b, c respectively) and 2000 rpm 75% load (d, e, and f respectively)
1010 20 30 40 50 60 70 8090100 200 300
2.40x1064.80x1067.20x1069.60x1061.20x1071.44x1071.68x1071.92x1072.16x107
dN/d
logd
p(cm
-3)
1010 20 30 40 50 60 70 8090100 200 300200 300
Diesel C810 C1214 C1618 C1822
1010 20 30 40 50 60 70 8090100 200 300
2.40x1064.80x1067.20x1069.60x1061.20x1071.44x1071.68x1071.92x1072.16x107
(b)
10 20 30 40 50 60 70 8090 200 300
(e)
dN
/d lo
gdp(
cm-3)
1010 20 30 40 50 60 70 8090100 200 300
2.40x1064.80x1067.20x1069.60x1061.20x1071.44x1071.68x1071.92x1072.16x107
(c)
Particle electrical mobility diameter(nm)Particle electrical mobility diameter(nm)1010 20 30 40 50 60 708090100 200 300
(f)
(d)dN
/d lo
gd p
(cm
-3)
(a)
10 20 40 60 80 100 2000.0
5.0x106
1.0x107
1.5x107
2.0x107
2.5x107
3.0x107
dN/d
log
d p(#
/cm
3)
10 20 40 60 80 100 2000.0
5.0x106
1.0x107
1.5x107
2.0x107
2.5x107
3.0x107
(f)
(e)(b)
(d)
Diesel C810 C1214 C1618 C1822
(a)
10 20 40 60 80 100 2000.0
5.0x106
1.0x107
1.5x107
2.0x107
2.5x107
3.0x107
dN/d
log
d p(#
/cm
3)
10 20 40 60 80 100 2000.0
5.0x106
1.0x107
1.5x107
2.0x107
2.5x107
3.0x107
10 20 40 60 80 100 2000.0
5.0x106
1.0x107
1.5x107
2.0x107
2.5x107
3.0x107
(c)
dN/d
log
d p(#
/cm
3)
Particle electrical mobility diameter(nm)
10 20 40 60 80 100 2000.0
5.0x106
1.0x107
1.5x107
2.0x107
2.5x107
3.0x107
Particle electrical mobility diameter(nm)
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 199
Figure A7-5: Particle number size distribution for B100, B50, and B20 at 1500 rpm
25% load (a, b, c respectively) and 2000 rpm 25% load (d, e, and f respectively)
Figure A7-6: Variations in particle median size among fuels at 1500 rpm 75%, 50%
and 25% loads
10 20 40 60 80 100 2000.00
7.60x106
1.52x107
2.28x107
3.04x107
3.80x107
dN/d
log
d p(#/c
m3 )
10 20 40 60 80 100 2000.00
7.60x106
1.52x107
2.28x107
3.04x107
3.80x107
(f)
(e)(b)
(d)
Diesel C810 C1214 C1618 C1822
(a)
10 20 40 60 80 100 2000.00
7.60x106
1.52x107
2.28x107
3.04x107
3.80x107
dN/d
log
d p(#/c
m3 )
10 20 40 60 80 100 2000.00
7.60x106
1.52x107
2.28x107
3.04x107
3.80x107
10 20 40 60 80 100 2000.00
7.60x106
1.52x107
2.28x107
3.04x107
3.80x107
(c)
dN/d
log
d p(#
/cm
3)
Particle electrical mobility diameter(nm)
10 20 40 60 80 100 2000.00
7.60x106
1.52x107
2.28x107
3.04x107
3.80x107
Particle electrical mobility diameter(nm)
200 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
Figure A7-7: Variations in particle median size among fuels at 2000 rpm 75%, 50%
and 25% loads
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 201
Figure A7-8: Variation in specific PM emissions with fuel oxygen content at 75%,
50% and 25% load
202 Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition
Figure A7-9: Variation in specific PM emissions with fuel surface tension at 75%,
50% and 25% load
Chapter 7: Diesel engine testing with biodiesel of controlled chemical composition 203
Figure A7-10: Variation in specific PM emissions with fuel viscosity at 75%, 50%
and 25% load
Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel 205
Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel
8.1 INTRODUCTION
Based on the work reported in this thesis, a number of Australian native plants were
found to be potential candidates for second-generation biodiesel feedstock. However,
this potential is still largely unexploited, mainly due to uncertainties related to the
quality of second-generation biodiesel as engine fuel and concern regarding engine
warranties and performance. The majority of current vehicle engines are not
optimised for the use of biodiesel and the different chemical and physical properties
of biodiesel (compared to petroleum diesel) will eventually effect fuel combustion
performance, exhaust emissions and engine durability. Therefore, biodiesel from new
feedstocks may not be suitable for direct use in conventional diesel engines and
engine system modification may be required. Moreover, quality standards are
becoming more crucial in relation to the commercial use of any fuel product, for the
assessment of safety risks and environmental pollution. As a result, any new
biodiesel needs to satisfy these quality standards and allow for smooth operation in
conventional diesel engines, before it can be considered as a sustainable fuel.
Therefore, the main aim of the work reported in this chapter was to experimentally
investigate the properties BOME, and to explore its suitability for use in unmodified
conventional automotive diesel engines, in terms of engine performance and
emissions.
206 Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel
8.2 INSTRUMENTATION AND METHODOLOGY
In this study, Beauty leaf biodiesel (BOME) was tested and compared with neat
regular petroleum diesel. The BOME was blended with petroleum diesel in two
different ratios by volume: 5% biodiesel with 95% diesel and 10% biodiesel with
90% diesel. The preparation of biodiesel from Beauty leaf seeds and its chemical
composition are presented in the previous chapters. The physical properties of diesel
and BOME were tested according to standard test procedures and the results are
presented in Table 8-1. In order to determine the fuel properties of diesel and BOME,
tests were conducted in the BERF fuel testing facility at QUT, Karlsruhe Institute of
Technology (KIT), Germany, and Caltex Refinery Laboratory in Wynnum, Brisbane.
The calorific value (HHV) of BOME was 13.3% lower than that of petroleum diesel.
Moreover, the higher kinematic viscosity, density, cetane number, surface tension,
flash point temperature and acid value of BOME compared to petroleum diesel were
also investigated as a possible cause of differences in engine performance and
emissions.
Table 8-1: Properties of Beauty leaf fatty acid methyl ester (BOME) and petroleum diesel
Properties Unit Test Method
Biodiesel Standard Diesel BOME Australian ASTM
D6751-12 EN14214
Kinematic Visocity @40 oC mm2/sec ASTM D445
3.5-5 1.9-6 3.5-5 2.64 4.54
Density @15 oC g/cm3 ASTM D4052
0.86-0.90 n/a 0.86-0.90 0.838 0.881
Higher Heating Value Mj/kg - n/a n/a n/a 45.93 39.82
Cetane Number - DIN 51773
51 min 47, min 51 min 50.5 64.03
Surface tension mPa.s - - - 26 29.85
Acid number 0.8 max 0.5 max 0.5 max 0 1.39
Flash point temperature °C ASTM D93
120, min 93, min 120, min 71 157
Lubricity @60 oC (wsd 1.4) mm IP 450 - - - 0.406
Cloud Point °C IP 309 Report Report Report 4 10.4
Sulphur mg/kg ASTM D7039
10 max 15 max 10 max 5.9
Experiments were conducted in the engine testing laboratory at the University of
Queensland (UQ) using a typical four-cylinder, turbocharged diesel car engine. The
Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel 207
engine specifications are shown in Table 8-2 and the experimental setup is shown in
the Figure 8-1. The speed and load of the test engine was controlled by a Froude
Hofmann AG 150 eddy current dynamometer. The dynamometer was embedded with
TEXCEL-Vl2 software for precise digital control and data acquisition. The
measurement accuracy for the torque and speed were ±1.25Nm (± 0.25% of full scale
load) and ± l rpm, respectively. A Kistlerpiezostar (Type 6056A42) pressure
transducer was used to measure in-cylinder gas pressure, which was recorded for
every 0.5 °CA. The variation of cycle-to-cycle cylinder pressure was recorded for
100 consecutive cycles and the mean was used for analysis. The engine was operated
at 2000 rpm with four different loads: 25%, 50%, 75% and 100%. The maximum
load at any particular engine speed depends on the type of fuel used, therefore, first
maximum load was measured for each fuel when engine was in full throttle for a
particular speed. This measured load was then considered as 100% load for that
speed and other load conditions were determined based upon the measured 100%
load. For each engine operating condition, the gaseous exhaust NOx, particle mass
(PM) and particle number (PN) were measured following a similar procedure and
using the same equipment as described in Section 7-2-2.
Figure 8-8-1: Experimental setup
Compressed air Dekati dilutor
4-cylinder DI diesel engine
Dynamometer
Sable
DustTrak
Crank angle encoder
Pressure transducer
Control
Compressor
Turbocharger
Air filter
DMS 500
Exhaust
Exhaust gas analyser
Computer
Computer
208 Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel
Table 8-2: Test engine specification
Model Peugeot 308 2.0 HDi
Cylinders 4 in-line
Capacity (L) 2
Bore x Stroke (mm) 85 x 88
Maximum power (kW/rpm) 100/4000
Maximum torque (Nm/rpm) 32/2000
Compression ratio 18
Aspiration Turbocharged & Inter cooled
Fuel injection Common rail (Multiple fuel injection)
Injection pressure (bar) 160
8.3 RESULTS AND DISCUSSION
8.3.1 Engine power
Variations in brake power and indicated power for the different fuels under full and
part loads are shown in Figure 8-2a and 8-2b, respectively. The results show that the
measured engine power of Beauty leaf biodiesel was less than that of diesel fuel and
as biodiesel percentage in the diesel-biodiesel blends increased, engine power
consistently decreased. The maximum reduction in engine brake power for 5% and
10% BOME with diesel blends at 100% load were 1.3 kW and 2.8 kW, whereas for
indicated power, the reductions were 1.4 kW and 3.3 kW, respectively. These results
were expected due to the lower heating value of BOME compared to diesel. Similar
results were reported in the literature (Utlu and Koçak 2008; Karabektas 2009;
Hansen, Gratton and Yuan 2006; Kaplan, Arslan and Sürmen 2006; Murillo et al.
2007) where reductions in engine output power were also explained by the lower
heating value of biodiesel. A detailed explanation for variations in power output from
the use of biodiesel in diesel engines has already been given in Section 2.3.1.
Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel 209
(a) (b)
Figure 8-8-2: Variation of power output for neat diesel and biodiesel blends (a) brake
power; (b) indicated power
8.3.2 BTE and BSFC
Brake thermal efficiency (BTE) and brake specific fuel consumption (BSFC) are
important indicators for evaluating diesel engine performance. These parameters
indicate the effectiveness of the engine in transferring the chemical energy of fuel for
use by the engine. BTE is the ratio of brake power of engine output shaft in a given
time to the input energy of the fuel supplied during the same time, whereas BSFC is
the fuel consumption, in mass, to produce unit energy output for a given period of
time. Figure 8-3a and 8-3b shows the respective variation in BTE and BSFC with
engine load for neat diesel engine and Beauty leaf fuel blends. It can be seen from
Figure 8-3a that the BTE of BOME-diesel blends and neat diesel was quite similar at
lower engine loads. At higher loads, the BTE decreased with the increase of biodiesel
percentage in the blend. The maximum reduction in engine BTE for 10% and 5%
BOME with diesel blends at 100% load were 1.43% and 0.55%, respectively. The
small reduction in BTE with the increase of BOME percentage might be due to the
higher viscosity, higher volatility, poor air fuel mixture, poor spray characteristics
and lower heating value at high loads (Nabi, Rahman and Akhter 2009;
Suryanarayanan et al. 2008). However, at a low load, this decrease was not observed,
probably due to the increased lubricity of the biodiesel blends compared to diesel
fuel (Ramadhas, Muraleedharan and Jayaraj 2005). In contrast to BTE, the BSFC
decreased with higher engine loads and increased with the biodiesel blend ratio, as
shown in Figure 8-3b. This trend was likely due to the fact that biodiesel mixtures
210 Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel
have a lower heating value than neat diesel fuel, and thus, more of the biodiesel
mixture was required for the maintenance of a constant power output (Nabi, Rahman
and Akhter 2009). At full load, the BSFCs increased by 2.5% and 5.6% for the
blending of 5% and 10% BOME compared to neat diesel. Apart from the lower
heating value of biodiesel, the increase in BSFC with biodiesel may be due to a
change in combustion timing caused by biodiesels’ higher cetane number, as well as
the injection timing change (Buyukkaya 2010). Similar observations were also
reported by many authors while conducting experiments with various biodiesel fuels
in diesel engines (Qi et al. 2010; Nabi, Rahman and Akhter 2009; Ilkılıç et al. 2011;
Utlu and Koçak 2008; Kumar 2009a; Haşimoğlu et al. 2008; Canakci 2007; Lin and
Li 2009a; Lapuerta, Armas and Rodriguez-Fernandez 2008; Zheng et al. 2008;
Raheman and Phadatare 2004; Labeckas and Slavinskas 2006).
(a) (b)
Figure 8-8-3: (a) Brake thermal efficiency (BTE) and (b) Brake specific fuel
consumption (BSFC) for neat diesel and biodiesel blends
8.3.3 Cylinder pressure
The variation of cylinder pressure with crank angle for diesel and BOME blends at
different engine loads is shown in Figure 8-4(a-d). It can be seen from the figures
that the peak cylinder pressure decreased with the increase of BOME in the diesel
and biodiesel blend, which is more visible for lower engine loads. The peak cylinder
pressure of 18,025, 17,860 and 17,850 kPa were recorded for neat diesel, 5% BOME
and 10% BOME blends at full load, respectively. Although the combustion process
Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel 211
for the test fuels was similar and consisted of a premixed combustion phase
followed by of diffusion combustion phase, the physical properties of the fuel also
play an important roles in atomisation rate and air fuel mixing quality (Senthil
Kumar et al. 2005; Canakci, Ozsezen and Turkcan 2009; Devan and Mahalakshmi
2009). Therefore, the lower peak cylinder pressure of Beauty leaf biodiesel is likely
due to its higher viscosity and lower volatility compared to standard diesel. Under
higher engine load conditions, the difference in cylinder pressure may be due to
better combustion of the oxygen content in biodiesel at high loads.
(a)
(b)
(c) (d)
Figure 8-8-4: Engine cylinder pressure for diesel and Beauty leaf biodiesel blends,
(a) full load; (b) 75% load; (c) 50% load; and (d) 25% load
212 Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel
8.3.4 Nitrogen oxide (NOx) emission
The variation in specific NOX emissions from the addition of 5% and 10% BOME in
petroleum diesel are shown in Figure 8-5. It can be seen from the figure that NOX
emissions increased for both increases in engine load and percentage of BOME in
diesel. On average, the addition of 5% and 10% BOME in diesel increased specific
NOX emissions by 24.41% and 32.73%, respectively. This increase in NOX is not
unusual, with most researches reporting that the main reason behind such an increase
is related to the fuel injection mechanism (Senatore et al. 2000; Tat and Van Gerpen
2003b; Szybist, Kirby and Boehman 2005; Monyem and H Van Gerpen 2001;
Cardone et al. 2002; Yamane, Ueta and Shimamoto 2001). Because of the higher
viscosity, higher density and lower compressibility of biodiesel, the pressure rises
more quickly in the pump and also progress more quickly towards the injector.
Therefore, an advanced fuel injection (early needle opening) is observed with
biodiesel compared with diesel, resulting in a higher pack temperature and NOx
formation rate. However, for common-rail engines, where physical properties have a
significant effect on fuel injection, researchers have reported another explanation for
the increase in NOX emissions from biodiesel. When conducting an experiment with
soybean oil biodiesel and regular diesel, whilst maintaining constant fuel injection
timing, Cheng et. al. (2006) attributed the observed increase in NOx emissions with
biodiesel to the reduction of shoot formation with biodiesel, which reduced radiation
heat dissipation, resulting in a higher adiabatic flame temperature and increased NOx
formation. The other arguments frequently put forward to explain the higher NOx
emissions from biodiesel include: (1) higher oxygen availability in the combustion
chamber favours the NOx formation reaction when using biodiesel (Schmidt and Van
Gerpen 1996; Song et al. 2004) and (2) higher cetane number which serves to
advance combustion by shortening the ignition delay (Monyem and H Van Gerpen
2001).
Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel 213
Figure 8-8-5: NOX emission for diesel and BOME blend for different engine load
conditions
8.3.5 Particle mass (PM) and particle number (PN)
Figure 8-6 shows the variation in particle emissions, in terms of specific particle
mass (PM) and specific particle number (PN) at different loads, for diesel and
BOME blends. A clear trend of decreasing particle emissions was observed with the
increase in BOME percentage and also for engine load (Figure 8-6a). The maximum
reduction of specific particle mass was found to be 54.6% with the engine running at
full load using 10% BOME with diesel. When the engine operated under part load
conditions, the reduction of particle emissions was 33-37% for the same BOME
blend. The higher oxygen content in BOME compared to neat diesel might be the
main reason for reductions in PM emissions. For example, Frijters and Baert (2006)
found a good correlation between PM emission reduction and the oxygen content of
fuel when conducting experiments using various biodiesel blends. The excess oxygen
allows for more complete combustion and also promotes oxidation of the already
formed soot. Several other authors reported similar reasons to explain the reduction
in PM emissions when using biodiesel or biodiesel blends (Lapuerta, Armas and
Ballesteros 2002; Wang et al. 2000; Schmidt and Van Gerpen 1996; Rakopoulos et
al. 2008). Apart from the excess oxygen effect, the absence of soot precursors, such
as aromatics and sulphur content in biodiesel, may be another reason for the
reduction in PM (Choi, Bower and Reitz 1997; Wang et al. 2000; Schmidt and Van
Gerpen 1996; Chang and Van Gerpen 1997). On the other hand, no significant
change was found for specific particle number with the addition of BOME in diesel
214 Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel
fuel. However, at high engine loads (50%, 75% and 100%), a small increase in
specific PN was observed. Therefore, the reduction in particle mass and increase in
particle number shown in Figure 8-6 indicates the production of smaller particles
which using BOME compared with that of pure diesel. Many authors also reported
an increase in the number of small particles while testing diesel engines with various
biodiesel fuels (Krahl et al. 1999; Lapuerta, Armas and Rodriguez-Fernandez 2008;
Young et al. 2012). A brief explanation of small particle formation when using
biodiesel as a diesel engine fuel is presented in Section 7.3.3.
(a) (b)
Figure 8-8-6: Particle emission for diesel and BOME blend in different engine load
condition (a) Brake specific particle mass (PM); (b) brake specific particle number
8.4 CONCLUSION
The performance and emissions of a four-cylinder common rail diesel engine were
experimentally investigated using neat diesel and biodiesel produced from Beauty
leaf oil. The important performance and emission indicators of diesel engines, in
terms of brake power, indicated power, brake thermal efficiency, brake specific fuel
consumption, cylinder pressure, NOx emission, specific PM emission and specific
PN emissions were measured and presented graphically. Results indicate that 5% and
10% blends of Beauty leaf oil biodiesel with diesel fuel can be used in conventional
diesel engines without engine modification. However, variations in engine
performance and emissions were observed, due to the different physicochemical
Chapter 8: Automobile diesel engine testing with Beauty leaf biodiesel 215
properties of commercial diesel and biodiesel produced from Beauty leaf oil, as well
as the fact that the test engine was designed for use with petroleum diesel only.
Beauty leaf biodiesel reduced the engine power, as well as brake thermal efficiency
and increased brake specific fuel consumption at higher engine loads. The cylinder
peak pressure decreased with increasing Beauty leaf biodiesel in the blend and this
was more visible for lower engine loads. Specific PM decreased sharply with the
increase in Beauty leaf biodiesel under all engine load conditions. On the other hand,
slightly higher specific PN was found with Beauty leaf biodiesel compared with neat
diesel. In addition, specific NOx emissions also increased when the engine was run
with Beauty leaf biodiesel. The variations in engine emission and performance using
neat diesel and Beauty leaf biodiesel blends were not surprising and similar findings
have been reported in the literature when testing diesel engines with commercially
available biodiesels. However, the use of higher Beauty leaf biodiesel blends or pure
biodiesel is recommended in order to obtain a more in-depth analysis.
Chapter 9: Conclusions 217
Chapter 9: Conclusions
9.1 CONCLUSIONS ARISING FROM THIS THESIS
Biodiesel is a potential source of renewable alternative energy that is capable of
replacing the conventional petroleum fuel currently used for CI engine applications.
It is commonly anticipated that biodiesel will play a significant role in providing
energy requirements for transportation fuel in the near future, due to the many socio-
economic advantages it has over fossil fuel. Therefore, numerous research studies
have been conducted over the past few decades, with the aim of improving biodiesel
technology. As a consequence, biodiesels are now produced globally on an
industrials scale, most of which are obtained from edible vegetable oil feedstocks
such as soybean, palm, canola, sunflower etc. In recent years, these types of biodiesel
have been the subject of discussions in relation to its impact on rising food prices and
creating pressure on agricultural land, and therefore, it is considered unsustainable in
the long term. There is now an urgent need to promote further research, in order to
overcome the drawbacks of current biodiesel technology and investigate the
suitability of second-generation biodiesel obtained from non-edible feedstocks. To
that end, this research assessed the potential of selected Australian native plants as a
source of second-generation biodiesel for use in internal combustion engines. The
aim of this study was achieved by using a systematic approach, starting with a
detailed literature review on current biodiesel technology and culminating in an
experimental investigation of the performance of conventional diesel engines fuelled
with second-generation biodiesel obtained from the native Australian Beauty leaf
plant. In order to understand its physico-chemical characteristics, a range of
experimental investigations were carried out using conventional biodiesels and
artificially prepared biodiesel with a controlled composition. In addition, various
numerical tools were used to achieve the objectives of this study, including: artificial
neural networks (ANN), principle component analysis (PCA), preference ranking
organisation methods and geometrical analysis for interactive assistance
(PROMETHEE-GAIA) design of experiment (DOE), the response surface
methodology (RSM) and analysis of variance (ANOVA). This research has provided
218 Chapter 9: Conclusions
enough information to reach a number of innovative conclusions, which are briefly
described in the following paragraphs.
In general, biodiesels are methyl esters of fatty acids produced through the
transesterification of vegetable oils or animal fats. However, clear differences in the
chemical composition of biodiesel, in terms of fatty acid structure, have been found
within feedstocks, as well as from one feedstock to the next, which ultimately
determines many important fuel properties of biodiesel. Fuel quality, which is known
to effect fuel combustion performance, exhaust emissions and engine durability, is
more sensitive in modern diesel engines, as the use of high pressure (about 75,000
bars) common rail fuel injection systems has increased. Therefore, biodiesel from
any feedstock needs to fulfil the quality standards for fuel properties, in order to
qualify for commercial use as an IC engine fuel. Internationally recognised standards
include EN14214 (Europe) and ASTM D-6751 (USA), whereas many other countries
have defined their own standards, which are frequently derived from these two. They
serve as guidelines for production, assure customers that they are buying high-quality
fuels, and provide authorities with approved tools for a common approach to
transport, storage and handling. The most important indicators of a fuels’ properties
include kinetic viscosity, density, cetane number, calorific value, flash point,
oxidation stability, cold temperature properties and iodine value. Experimental
investigations of these properties require specialised equipment, skilled technicians
and biodiesel production on a pilot scale. These things are often costly, especially for
biodiesel from new sources, which is one barrier to the establishment of biodiesel
from new feedstock. Therefore, numerical modelling is a possible alternative to
replace costly experimental studies and hence, accelerate the development of second-
generation biodiesel technology.
Multivariate data analysis (using PCA tools) confirmed a complex relationship
between the chemical composition of biodiesel and individual fuel properties. This
study also indicated that individual methyl esters of fatty acids, average chain length
(ACL), average number of double bonds (ANBD), weight percentages of oxygen
(O), hydrogen (H), carbon (C), saturated fatty acids, mono-unsaturated fatty acids
(MUFA), poly-unsaturated fatty acids (PUFA), mono-glycerol and free fatty acids
Chapter 9: Conclusions 219
(FFA) in the chemical composition of biodiesel have a different level of influence on
fuel properties. For instance, ACL has a very positive influence on cetane number,
whereas it has a moderate or lesser influence on other fuel properties, except
oxidation stability, where no significant influence was found. The most influential
parameters of chemical composition that effected all biodiesel properties in this study
were the presence of PUFAs and ANBDs in the chemical composition of biodiesels.
The findings of the correlation study between chemical composition and biodiesel
properties assisted in the selection of input variables for a particular fuel property
when developing the ANN models. The parameters of chemical composition, which
have a significant influence on this property, were used in the input layer of the ANN
model. MatLab R2012a software was used to train, validate and simulate the ANN
model on a personal computer and the developed ANN prediction models ran
simulations using test data sets, in order to evaluate the estimation performance.
From the simulation results, it was found that the absolute fraction of variance (R2)
was close to unity, ranging from 0.8932 to 0.9622, with Root-Mean-Squared (RMS)
errors ranging from 0.011 to 4.171 and a maximum average error percentage
(MAEP) ranging from 1.86% to 5.53%. The results of this study also show that
ANNs have the ability to learn and generalise a wide range of experimental
conditions for biodiesel.
The reason for developing the ANN models was to estimate the fuel properties of
biodiesel obtained from Australian native plants based on its chemical composition,
instead of having to conduct costly and time consuming experimental studies. For
this purpose, biodiesels were produced on a laboratory scale from eleven non-edible
oil seeds plants, which were Beauty leaf, Candle nut, Blue berry lily, Queen palm,
Castor, Bidwilli, Karanja, Whitewood, Cordyline, Flame tree and Chinese rain. Most
of the native plant biodiesel contained much higher free fatty acids (FFA) compared
to the recommended level (5% by weight) from alkali-trans-esterification. The flame
oil contained 36.7% FFAs, which was highest among the bio-oils, followed by
Beauty leaf (22%), Queen palm (15%) and Blue berry lily (13.1%). Therefore, the
FFA content in non-edible plant feedstock could be one of the major issues inhibiting
the success of biodiesel production from native species. The chemical composition of
biodiesel obtained from native oil seeds were similar to that of conventional
220 Chapter 9: Conclusions
biodiesel and they were mostly rich in triglycerides of Oleic (C18:1) acid, followed
by Stearic (C18:0), Linoleic (C18:2), Palmitic (C16:0) and Linolenic (C18:3) fatty
acids, with the exception of Queen palm and Castor. Queen palm was rich in short
chain saturated fatty acids (C8:0, C10:0, C12:0 and C8:0) and Castor biodiesel was
rich in unsaturated long chain fatty acids (C20:1). Queen palm contained an
exceptionally higher amount of oxygen, accounting for 14.19% on a per weight
basis. The oxygen content of other methyl esters ranged from 10.25 to 11.82%.
Based on their chemical composition, the important fuel properties of biodiesels
were estimated using ANN models and compared with Australian, European and
American biodiesel standards. Results indicated that most of the biodiesel
investigated in this study performed well against biodiesel standards for all quality
indicators, except oxidation stability (OS). Queen palm biodiesel was the only
sample that fulfilled biodiesel standards in terms of OS, due to its very high saturated
fatty acid content. Based on the dry seed production capability, level of difficulty
processing seeds, bio-oil content in the seed kernel, amount of free fatty acids and
estimated fuel properties of biodiesel, the native species feedstock were evaluated
and compared with each other using a multi-criteria decision method (MCDM)
software PORMETHEE-GAIA. In addition, sensitivity analysis of the ranking of
native plant species was conducted by changing the weighting of three important
criteria, being OY, OS and CFPP, which produced significant changes in the ranking
of feedstocks. This study found that Beauty leaf was the top ranked feedstock for
biodiesel production, followed by Queen palm, Castor and Karanja. Overall, Beauty
leaf and Queen palm biodiesel were found to be a good choice for second-generation
biodiesel production in tropical/sub-tropical regions, however, the opposite is true for
cold weather conditions, where Castor, Cordyline or Flame tree might be a better
choice.
As Beauty leaf was found to be one of the most promising feedstock, a detailed
experimental investigation of the species was undertaken at the Central Queensland
University (CQU), using about 140 kg of ground-dried Beauty leaf seeds. Overall, it
was found that Beauty leaf bio-oil extraction using an oil press resulted in a low oil
yield. This drawback was overcome by using a chemical oil extraction method with
n-hexane as the oil solvent. Furthermore, the oil yield increased 3-4% when using
accelerated solvent extraction (ASE) methods, with high pressure and temperature
Chapter 9: Conclusions 221
extraction. The highest oil yield was found to be 51.5% for dry kernels. Overall,
when comparing the quality of non-edible bio-oil with edible vegetable oils, in terms
of acid value, density, kinematic viscosity, surface tension and higher heating value,
Beauty leaf oil showed much higher acid values, resulting from its high free fatty
acid content. Although this made Beauty leaf oil inappropriate for direct base-
catalysed trans-esterification, a two-step esterification process was used in this study:
acid-catalysed pre-esterification and base-catalysed trans-esterification. Due to the
lack of samples, a response surface method (RSM) based on a Box-Behnken design
was employed in order to determine an experimental plan to optimise the Beauty leaf
oil to biodiesel conversion procedure. The effect of reaction parameters such as
methanol to oil molar ratio, catalyst loading and reaction temperature were
investigated in terms of the reduction of FFA content in pre-esterification and ester
content in trans-esterification. The optimal conditions for pre-esterification were 30:1
methanol to oil molar ratio, 10 wt% sulphuric acid catalyst and 75 °C reaction
temperature, which reduced the FFA content to 1.8 wt%. With the aid of statistical
modelling, predicted optimal conditions for the transesterification methanol to oil
molar ratio, catalyst concentration and reaction temperature were 7.5:1, 1% and 55
°C, respectively. Based on these conditions, the highest achievable ester content of
FAME predicted by the model was found to be approximately 93%. In terms of a
linear effect on FFA reduction for the first step, methanol to oil molar ratio was
found to be highly significant and reaction temperature moderately significant. For
trans-esterification, catalyst concentration was found be the most dominant variable
for achieving high ester contents.
Beauty leaf biodiesels mostly comprise esters of saturated Hexadecanoic (C16:0) and
Octadecanoic (C18:0) acid, mono-unsaturated 9-Octadecenoic acid (C18:1) and
poly-unsaturated 9, 12-Octadecadienoic (C18:2) acid. This biodiesel is rich in
saturated methyl esters compared with commercial biodiesels, except the biodiesel
from palm oil. This makes Beauty leaf oil biodiesel preferable in terms of most fuel
properties, including kinematic viscosity, density, higher heating value, oxidation
stability, iodine value, cetane number, flash point and linoleic acid content. On the
other hand, Beauty leaf biodiesels perform relatively poorly in terms of cold
temperature properties and free fatty acid content. However, Beauty leaf biodiesel
foes meet the American, European and Australian biodiesel standards. The
222 Chapter 9: Conclusions
multivariate data analysis using PROMETHEE-GAIA software indicated that
biodiesel produced from Beauty leaf bio-oil could be a better option for use in
automobile engines, compared with many other commercial biodiesels, including
biodiesel from cotton seed, sunflower and soybean oil, especially in tropical/sub-
tropical regions.
The influence of a biodiesels’ chemical composition and physical properties on
diesel particle emissions was experimentally investigated using four biodiesels with
different fatty acid carbon chain length and degrees of unsaturation. The aims of this
investigation were to correlate the findings of these results with the second-
generation biodiesel produced from Australian native plants. This study found a
consistent reduction in particle emissions with increased fuel oxygen content.
Therefore, it can be expected that particle emissions may be lower for an engine
running on Queen palm biodiesel, because of its significantly higher oxygen content
(14.19% by weight). Although the oxygen content of other native plant biodiesels
was found to be similar, the ACL of Castor and Bidwilli biodiesel was found to be
19.90 and 18.52, respectively, which is much higher compared with other biodiesels.
Therefore, these two biodiesel may produce a higher amount of particle emissions.
This study also found that particle emissions decreased linearly with kinematic
viscosity. Overall, it is evident from this section of the study that the chemical and
physical properties of biodiesel may be important parameters impacting on the
performance and emission characteristics of automobile engines when native plant
biodiesels. The last chapter of this thesis (Chapter 9) went on to investigate the use of
Beauty leaf biodiesel in a four-cylinder automobile diesel engine. The important
performance and emission indicators for diesel engines, in terms of brake power,
indicated power, brake thermal efficiency, brake specific fuel consumption, cylinder
pressure, NOx emissions, specific PM emissions and specific PN emissions were
measured and presented graphically. Results indicated that 5% and 10% blends of
Beauty leaf oil biodiesel with diesel fuel can used in conventional diesel engines
without engine modification. Beauty leaf biodiesel reduced the engine power, brake
thermal efficiency, cylinder peak pressure and specific nitrogen oxide (NOx) particle
mass (PM) emissions. At the same time, brake specific fuel consumption and particle
emissions (in terms of number) were found to be higher for Beauty leaf biodiesel
compared to conventional diesel. The variation in engine emissions and performance
Chapter 9: Conclusions 223
using neat diesel versus Beauty leaf biodiesel blends was not surprising and similar
findings have been reported in the literature when testing diesel engines with
commercially available biodiesels. Nevertheless, for a better understanding of the
performance and emission characteristics of Beauty leaf and other native biodiesels,
it may be necessary to run automobile diesel engines using the biodiesels in a pure
form.
9.2 LIMITATIONS AND RECOMMENDATIONS FOR FUTURE WORK
The higher fuel economy, better performance and durability of diesel engines are
making them a dominant power source in both mobile and stationary applications. In
addition, the use of alternative fuels (i.e. biodiesel) is also expanding around the
world. Therefore, more inclusive research initiatives, with collaboration between
different branches of science and engineering, are needed to develop sustainable
biodiesel technologies, in order to produce a secure energy supply in the future.
Having a vast land area and naturally grown non-edible oil seed crops, Australia has
the unique opportunity to become a major supplier of second-generation biodiesel. In
order to take full advantage of this opportunity, multi-disciplinary research needs to
be undertaken. This study explored the potential of eleven Australian native plants as
a feedstock for second-generation biodiesel production from an engineering and
application point of view. For the further development and initiation of the
commercial production of biodiesels, a socio-economic assessment of the use of
these plant species will be required. Investigations from an agricultural point of view
and proper planning for effective land use for oil seed plants will also be essential.
Therefore, more extensive research studies should be undertaken, in collaboration
with plant scientists, engineers, economists and policy makers.
This study found that seed processing methods significantly influenced oil extraction
and yield. Manual seed cracking and kernel extraction is a very time-consuming and
laborious task and given that the size and physical condition of the seeds varies
224 Chapter 9: Conclusions
significantly, even for the same species, further research is required to design
automated seed crushers for native oil seeds. The physical properties of dry seeds
also need to be investigated in terms of geometry, and tensile and compression
strength. After analysing these physical properties, a 3D Cad model can be
developed and simulated for optimising the design. The design of automated seed
crushers should be optimised in such a way that it would not destroy or smash the oil
bearing kernels.
Since the current study only focused on the bio-oil content of Australian native oil
seeds, further research should be conducted for investigating the by-products
obtained during oil extraction. For instance, the seed husk and kernel residue can also
be used for bio-oil, bio-gas and bio-char production through pyrolysis. This would
certainly add the more value to the use of Australian native plants in fuel production.
Moreover, a life cycle assessment of those plants, in terms of energy production and
greenhouse gas reductions, would also produce interesting outcomes.
This study found a good relation between the chemical composition of biodiesel and
various important fuel quality parameters. This research can be further extended for
additional fuel properties such as lubricity, copper strip corrosion, distillation
temperature, fire point, cloud point, pour point etc. Moreover, the physico-chemical
properties investigated in this study indicated that a particular feedstock may be
better for certain quality parameters but not for others. Therefore, biodiesels from
several feedstocks could be blended in order to obtain the desired fuel quality and it
would be interesting to find out the optimum blend ratios for achieving a higher level
of fuel quality.
The developed ANN models for estimating important biodiesel properties showed
good estimation accuracy. However, the performance of ANN models can be further
improved by providing additional data sets while training the ANN networks. The
data sets should cover as much of a range as possible, in order to ensure the most
robust ANN models are developed. Moreover, further research would be worthwhile
Chapter 9: Conclusions 225
for developing a universal ANN model that would be able to predict the combustion
performance of versatile automobile engines and fuel types.
This thesis found that, when using biodiesels and their blends, particle emissions
decreased with the decrease in ACL and increase in oxygen content of biodiesel.
However, biodiesel ACL and oxygen content are interrelated. Oxygen content
increases with a decrease in ACL and vice versa. Therefore, whether oxygen content
or FAME carbon chain length is the dominant reason for particle emission reductions
is not evident from the results reported in this thesis. It would be interesting to design
experiments that will evaluate the effect biodiesel carbon chain length keeping the
oxygen content constant or vice versa.
Due to the limited number and type of samples, this study conducted engine
experiments fuelled with 5% and 10% Beauty leaf biodiesel blended with petroleum
diesel. However, Beauty leaf biodiesel in its pure form needs to be used in diesel
engines, in order to gain for better understanding of its combustion performance. It is
also recommended to test automobile engine performance using the pure form of
other native plant biodiesels.
6 Bibliography 227
6 Bibliography
(APAC), Asia Pacific. 2009. Biofuel Consultants. Australian Biofuels. Edited by EnergyQuest and Ecco Consulting. Canberra, Australia.
Agarwal, Avinash Kumar. 1999. "Performance evaluation and tribological studies on a
biodiesel-fuelled compression ignition engine, Ph. D. thesis, Center for Energy Studies, Indian Institute of Technology, Delhi, India.
Agarwal, Madhu, Kailash Singh and SP Chaurasia. 2010. "Prediction of biodiesel
properties from fatty acid composition using linear regression and ANN techniques." Indian Chemical Engineer 52 (4): 347-361.
Ahmad, AL, NH Yasin, CJC Derek and JK Lim. 2011. "Microalgae as a sustainable
energy source for biodiesel production: a review." Renewable and Sustainable Energy Reviews 15 (1): 584-593.
Ahmad, M, K Ullah, MA Khan, S Ali, M Zafar and S Sultana. 2011. "Quantitative and
qualitative analysis of sesame oil biodiesel." Energy Sources, Part A: Recovery, Utilization, and Environmental Effects 33 (13): 1239-1249.
Albuquerque, MCG, YL Machado, AEB Torres, DCS Azevedo, CL Cavalcante Jr, LR
Firmiano and EJS Parente Jr. 2009. "Properties of biodiesel oils formulated using different biomass sources and their blends." Renewable Energy 34 (3): 857-859.
Alcantara, R, J Amores, L t Canoira, E Fidalgo, MJ Franco and A Navarro. 2000.
"Catalytic production of biodiesel from soy-bean oil, used frying oil and tallow." Biomass and bioenergy 18 (6): 515-527.
Ali Ahmadi, Mohammad, Sohrab Zendehboudi, Ali Lohi, Ali Elkamel and Ioannis
Chatzis. 2013. "Reservoir permeability prediction by neural networks combined with hybrid genetic algorithm and particle swarm optimization." Geophysical Prospecting 61 (3): 582-598.
Ali, Rehab Farouk Mohammed. 2013. "Preparation and characterization of protein
isolate and biodiesel from garden cress seed." European Journal of Chemistry 4 (2): 85-91.
Ali, Yusuf, Milford A Hanna and Susan L Cuppett. 1995a. "Fuel properties of tallow
and soybean oil esters." Journal of the American Oil Chemists’ Society 72 (12): 1557-1564.
228 6 Bibliography
Ali, Yusuf, MilfordA Hanna and SusanL Cuppett. 1995b. "Fuel properties of tallow and soybean oil esters." Journal of the American Oil Chemists’ Society 72 (12): 1557-1564.
Alkım, Erdem, Emre Gürbüz and Erdal Kılıç. 2012. "A fast and adaptive automated
disease diagnosis method with an innovative neural network model." Neural Networks 33: 88-96.
Allan, Robert, John Williams and John Rogerson. 2008. "Effect of the Molecular
Structure of Individual Fatty Acid Alcohol Esters (Biodiesel) on the Formation of Nox and Particulate Matter in the Diesel Combustion Process." SAE Int. J. Fuels Lubr. 1 (1): 849-872.
Alleman, TL and Robert L McCormick. 2006. "Analysis of coconut-derived biodiesel
and conventional diesel fuel samples from the Philippines." US DOE, NREL/MP-540-38643.
Alptekin, Ertan and Mustafa Canakci. 2008. "Determination of the density and the
viscosities of biodiesel–diesel fuel blends." Renewable Energy 33 (12): 2623-2630.
Alptekin, Ertan, Mustafa Canakci and Huseyin Sanli. 2012. "Evaluation of leather
industry wastes as a feedstock for biodiesel production." Fuel 95 (0): 214-220.
Altıparmak, Duran, Ali Keskin, Atilla Koca and Metin Gürü. 2007. "Alternative fuel
properties of tall oil fatty acid methyl ester–diesel fuel blends." Bioresource Technology 98 (2): 241-246.
Amaro, Helena M, A Guedes and F Xavier Malcata. 2011. "Advances and perspectives
in using microalgae to produce biodiesel." Applied Energy 88 (10): 3402-3410.
Anastopoulos, G, E Lois, D Karonis, S Kalligeros and F Zannikos. 2005. "Impact of
oxygen and nitrogen compounds on the lubrication properties of low sulfur diesel fuels." Energy 30 (2): 415-426.
Anastopoulos, G, E Lois, F Zannikos, S Kalligeros and C Teas. 2001. "Influence of
aceto acetic esters and di-carboxylic acid esters on diesel fuel lubricity." Tribology International 34 (11): 749-755.
Antolın, G, FV Tinaut, Y Briceno, V Castano, C Perez and AI Ramırez. 2002.
"Optimisation of biodiesel production by sunflower oil transesterification." Bioresource technology 83 (2): 111-114.
6 Bibliography 229
Aransiola, EF, TV Ojumu, OO Oyekola and DIO Ikhuomoregbe. 2012. "A Study of Biodiesel Production from Non-Edible Oil Seeds: A Comparative Study." In The Open Conference Proceedings Journal, M1.
Arcaklioğlu, Erol and İsmet Çelıkten. 2005. "A diesel engine's performance and exhaust
emissions." Applied energy 80 (1): 11-22.
Armas, Octavio, Kuen Yehliu and André L Boehman. 2010. "Effect of alternative fuels
on exhaust emissions during diesel engine operation with matched combustion phasing." Fuel 89 (2): 438-456.
Ashwath, Nanjappa. 2010a. Evaluating Biodiesel Potential of Australian Native and
Naturalised Plant Canberra, Australia: Rural Industries Research and Development Corporation (RIRDC).
Ashwath, Nanjappa. 2010b. Evaluating biodiesel potential of Australian native and
naturalised plant species: Rural Industries Research and Development Corporation.
Atabani, AE, AS Silitonga, HC Ong, TMI Mahlia, HH Masjuki, Irfan Anjum Badruddin
and H Fayaz. 2013. "Non-edible vegetable oils: a critical evaluation of oil extraction, fatty acid compositions, biodiesel production, characteristics, engine performance and emissions production." Renewable and Sustainable Energy Reviews 18: 211-245.
Atapour, Mehdi and Hamid-Reza Kariminia. 2011. "Characterization and
transesterification of Iranian bitter almond oil for biodiesel production." Applied Energy 88 (7): 2377-2381.
Aydin, Hüseyin and Hasan Bayindir. 2010. "Performance and emission analysis of
cottonseed oil methyl ester in a diesel engine." Renewable Energy 35 (3): 588-592.
Bajpai, Divya and VK Tyagi. 2006. "Biodiesel: source, production, composition,
properties and its benefits." Journal of Oleo Science 55 (10): 487-502.
Bakeas, Evangelos, Georgios Karavalakis and Stamoulis Stournas. 2011. "Biodiesel
emissions profile in modern diesel vehicles. Part 1: Effect of biodiesel origin on the criteria emissions." Science of the total environment 409 (9): 1670-1676.
Balabin, Roman M, Ekaterina I Lomakina and Ravilya Z Safieva. 2011. "Neural
network (ANN) approach to biodiesel analysis: analysis of biodiesel density, kinematic viscosity, methanol and water contents using near infrared (NIR) spectroscopy." Fuel 90 (5): 2007-2015.
230 6 Bibliography
Balat, Mustafa and Havva Balat. 2008. "A critical review of bio-diesel as a vehicular fuel." Energy conversion and management 49 (10): 2727-2741.
Balat, Mustafa and Havva Balat. 2010. "Progress in biodiesel processing." Applied
energy 87 (6): 1815-1835.
Bamgboye, AI and AC Hansen. 2008. "Prediction of cetane number of biodiesel fuel
from the fatty acid methyl ester (FAME) composition." International Agrophysics 22 (1): 21.
Banapurmath, NR, PG Tewari and RS Hosmath. 2008. "Performance and emission
characteristics of a DI compression ignition engine operated on Honge, Jatropha and sesame oil methyl esters." Renewable energy 33 (9): 1982-1988.
Banković-Ilić, Ivana B, Olivera S Stamenković and Vlada B Veljković. 2012. "Biodiesel
production from non-edible plant oils." Renewable and Sustainable Energy Reviews 16 (6): 3621-3647.
Bannikov, MG. 2011. "Combustion and emission characteristics of Mustard biodiesel."
In 6th International Advanced Technologies Symposium (IATS’11), Turkey, edited, 1-5.
Barnwal, BK and MP Sharma. 2005. "Prospects of biodiesel production from vegetable
oils in India." Renewable and sustainable energy reviews 9 (4): 363-378.
Baroutian, S.; Aroua, M.K. 2008. "Estimation of vegetable oil-based ethyl esters
biodiesel densities using artificial neural networks." Journal of Applied Science 8: 3005-3011.
Behçet, Rasim. 2011. "Performance and emission study of waste anchovy fish biodiesel
in a diesel engine." Fuel Processing Technology 92 (6): 1187-1194.
Benjumea, P., J. R. Agudelo and A. F. Agudelo. 2011. "Effect of the degree of
unsaturation of biodiesel fuels on engine performance, combustion characteristics, and emissions." Energy and Fuels 25 (1): 77-85.
Benjumea, Pedro, John Agudelo and Andres Agudelo. 2008. "Basic properties of palm
oil biodiesel–diesel blends." Fuel 87 (10): 2069-2075.
Benjumea, Pedro, John R Agudelo and Andrés F Agudelo. 2010. "Effect of the degree
of unsaturation of biodiesel fuels on engine performance, combustion characteristics, and emissions." Energy & Fuels 25 (1): 77-85.
6 Bibliography 231
Berrios, M, MC Gutiérrez, MA Martín and A Martin. 2009. "Application of the factorial design of experiments to biodiesel production from lard." Fuel Processing Technology 90 (12): 1447-1451.
Betiku, Eriola and Tunde Folorunsho Adepoju. 2013. "Methanolysis optimization of
sesame (Sesamum indicum) oil to biodiesel and fuel quality characterization." International Journal of Energy and Environmental Engineering 4 (1): 1-8.
Bhuyan, S, S Sundararajan, L Yao, EG Hammond and T Wang. 2006. "Boundary
lubrication properties of lipid-based compounds evaluated using microtribological methods." In STLE/ASME 2006 International Joint Tribology Conference, edited, 151-157: American Society of Mechanical Engineers.
Bioenergy, IEA. 2008. "From 1st-to 2nd-Generation BioFuel technoloGies." An
overview of current industry and RD&D activities. IEA-OECD.
Blangino, E, AF Riveros and SD Romano. 2008. "Numerical expressions for viscosity,
surface tension and density of biodiesel: analysis and experimental validation." Physics and Chemistry of Liquids 46 (5): 527-547.
Boudy, Frédéric and Patrice Seers. 2009. "Impact of physical properties of biodiesel on
the injection process in a common-rail direct injection system." Energy Conversion and Management 50 (12): 2905-2912.
Bozbas, Kahraman. 2008. "Biodiesel as an alternative motor fuel: Production and
policies in the European Union." Renewable and Sustainable Energy Reviews 12 (2): 542-552.
BP. 2012. "Energy outlook 2030". London: BP.
Bradsher, K. 2008. "The Other Oil Shock: Vegetable Oil Prices Soar". In The New York
Times, World Business. New York, USA.
Brans, J Pierre. 2002. "Ethics and decision." European Journal of Operational Research
136 (2): 340-352.
Brans, Jean-Pierre and Bertrand Mareschal. 1994. "The PROMCALC & GAIA decision
support system for multicriteria decision aid." Decision support systems 12 (4): 297-310.
Brito, J. M., L. Belotti, A. C. Toledo, L. Antonangelo, F. S. Silva, D. S. Alvim, P. A.
Andre, P. H. N. Saldiva and D. H. R. F. Rivero. 2010. "Acute cardiovascular and inflammatory toxicity induced by inhalation of diesel and biodiesel exhaust particles." Toxicological Sciences 116 (1): 67-78.
232 6 Bibliography
Broday, David M. and Ravid Rosenzweig. 2011. "Deposition of fractal-like soot aggregates in the human respiratory tract." Journal of Aerosol Science 42 (6): 372-386.
Buyukkaya, Ekrem. 2010. "Effects of biodiesel on a DI diesel engine performance,
emission and combustion characteristics." Fuel 89 (10): 3099-3105.
Canakci, M. 2005a. "Performance and emissions characteristics of biodiesel from
soybean oil." Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering 219 (7): 915-922.
Canakci, M and H Sanli. 2008. "Biodiesel production from various feedstocks and their
effects on the fuel properties." Journal of industrial microbiology & biotechnology 35 (5): 431-441.
Canakci, M.; Sanli, H. 2005b. "An assessment about the reasons of NOx rise in
biodiesel’s exhaust emissions." J. Naval Sci. Eng 3: 81–92.
Canakci, Mustafa. 2007. "Combustion characteristics of a turbocharged DI compression
ignition engine fueled with petroleum diesel fuels and biodiesel." Bioresource technology 98 (6): 1167-1175.
Canakci, Mustafa, Ahmet Necati Ozsezen, Erol Arcaklioglu and Ahmet Erdil. 2009.
"Prediction of performance and exhaust emissions of a diesel engine fueled with biodiesel produced from waste frying palm oil." Expert systems with Applications 36 (5): 9268-9280.
Canakci, Mustafa, Ahmet Necati Ozsezen and Ali Turkcan. 2009. "Combustion analysis
of preheated crude sunflower oil in an IDI diesel engine." Biomass and Bioenergy 33 (5): 760-767.
Canakci, Mustafa and Jon H Van Gerpen. 2003. "Comparison of engine performance
and emissions for petroleum diesel fuel, yellow grease biodiesel, and soybean oil biodiesel." Transactions of the ASAE 46 (4): 937-944.
Candeia, RA, MCD Silva, JR Carvalho Filho, MGA Brasilino, TC Bicudo, IMG Santos
and AG Souza. 2009. "Influence of soybean biodiesel content on basic properties of biodiesel–diesel blends." Fuel 88 (4): 738-743.
Cardone, Massimo, Maria Vittoria Prati, Vittorio Rocco, Maurizia Seggiani, Adolfo
Senatore and Sandra Vitolo. 2002. "Brassica carinata as an alternative oil crop for the production of biodiesel in Italy: engine performance and regulated and unregulated exhaust emissions." Environmental science & technology 36 (21): 4656-4662.
6 Bibliography 233
Carraretto, C, A Macor, A Mirandola, A Stoppato and S Tonon. 2004. "Biodiesel as alternative fuel: Experimental analysis and energetic evaluations." Energy 29 (12): 2195-2211.
Carrillo, Snaider, Jim Harkin, Liam McDaid, Sandeep Pande, Seamus Cawley, Brian
McGinley and Fearghal Morgan. 2012. "Advancing interconnect density for spiking neural network hardware implementations using traffic-aware adaptive network-on-chip routers." Neural networks 33: 42-57.
Cay, Yusuf, Adem Cicek, Fuat Kara and Selami Sağiroğlu. 2012. "Prediction of engine
performance for an alternative fuel using artificial neural network." Applied Thermal Engineering 37: 217-225.
Cecrle, Eric, Christopher Depcik, Andrew Duncan, Jing Guo, Michael Mangus, Edward
Peltier, Susan Stagg-Williams and Yue Zhong. 2012. "Investigation of the effects of biodiesel feedstock on the performance and emissions of a single-cylinder diesel engine." Energy & Fuels 26 (4): 2331-2341.
Çelik, Veli and Erol Arcaklioğlu. 2005. "Performance maps of a diesel engine." Applied
Energy 81 (3): 247-259.
Chang, David YZ and Jon H Van Gerpen. 1997. Fuel properties and engine
performance for biodiesel prepared from modified feedstocks: SAE Technical Paper.
Chapman, E.; Boehman, A.L. 2006. "Emissions Characteristics of a Light Duty Diesel
Engine Fuelled with a Hydrogenated Biodiesel Fuel." In 231st National Meeting of the American Chemical Society, Atlanta, GA, USA, 26–30 March 2006, edited.
Cheenkachorn, Kraipat. 2004. "Predicting properties of biodiesels using statistical
models and artificial neural networks." In Proceedings of the Joint International Conference on Sustainable Energy and Environment, Hua Hin, Thailand, edited, 1-3.
Chen, He, Baoxiang Peng, Dezheng Wang and Jinfu Wang. 2007. "Biodiesel production
by the transesterification of cottonseed oil by solid acid catalysts." Frontiers of Chemical Engineering in China 1 (1): 11-15.
Chen, Pin-Chia, Wei-Cheng Wang, William L. Roberts and Tiegang Fang. 2013. "Spray
and atomization of diesel fuel and its alternatives from a single-hole injector using a common rail fuel injection system." Fuel 103 (0): 850-861.
Cheng, AS, A Upatnieks and CJ Mueller. 2006. "Investigation of the impact of biodiesel
fuelling on NOx emissions using an optical direct injection diesel engine." International Journal of Engine Research 7 (4): 297-318.
234 6 Bibliography
Chhetri, AB and KC Watts. 2012a. "Densities of canola, jatropha and soapnut biodiesel
at elevated temperatures and pressures." Fuel 99: 210-216.
Chhetri, AB and KC Watts. 2012b. "Viscosities of canola, jatropha and soapnut
biodiesel at elevated temperatures and pressures." Fuel 102: 789-794.
Chhetri, Arjun B, Martin S Tango, Suzanne M Budge, K Chris Watts and M Rafiqul
Islam. 2008. "Non-edible plant oils as new sources for biodiesel production." International journal of molecular sciences 9 (2): 169-180.
Chhetri, Arjun B, K Chris Watts and M Rafiqul Islam. 2008. "Waste cooking oil as an
alternate feedstock for biodiesel production." Energies 1 (1): 3-18.
Chisti, Yusuf. 2007. "Biodiesel from microalgae." Biotechnology advances 25 (3): 294-
306.
Choi, CY, GR Bower and Rolf D Reitz. 1997. Effects of biodiesel blended fuels and
multiple injections on DI diesel engines: SAE Technical Paper.
Choi, CY and RD Reitz. 1999. "A numerical analysis of the emissions characteristics of
biodiesel blended fuels." Journal of engineering for gas turbines and power 121 (1): 31-37.
Choi, Y and J-Y Chen. 2005. "Fast prediction of start-of-combustion in HCCI with
combined artificial neural networks and ignition delay model." Proceedings of the Combustion Institute 30 (2): 2711-2718.
Choudhury, S and PK Bose. 2008. Jatropha derived Biodiesel–its suitability as CI
engine fuel: SAE Technical Paper.
Chuck, Christopher J, Chris D Bannister, J Gary Hawley, Matthew G Davidson, Ivan La
Bruna and Alex Paine. 2009. "Predictive model to assess the molecular structure of biodiesel fuel." Energy & Fuels 23 (4): 2290-2294.
Chung, Kyong-Hwan. 2010. "Transesterification of Camellia japonica and Vernicia
fordii seed oils on alkali catalysts for biodiesel production." Journal of Industrial and Engineering Chemistry 16 (4): 506-509.
Costa, Marcelo Azevedo, Antonio Padua Braga and Benjamin Rodrigues De Menezes.
2012. "Convergence analysis of sliding mode trajectories in multi-objective neural networks learning." Neural Networks 33: 21-31.
6 Bibliography 235
Crabbe, Edward, Cirilo Nolasco-Hipolito, Genta Kobayashi, Kenji Sonomoto and Ayaaki Ishizaki. 2001. "Biodiesel production from crude palm oil and evaluation of butanol extraction and fuel properties." Process Biochemistry 37 (1): 65-71.
CSIRO. 2010. Calophyllum inophyllum,” Australian Tropical Rainforest Plants, and.
Cursaru, Diana, Mihaela Neagu and Liana Bogatu. 2013. "Investigations on the
Oxidation Stability of Biodiesel Synthesized from Different Vegetable Oils." REVISTA DE CHIMIE 64 (4): 438-441.
Dantas, MB, AR Albuquerque, AK Barros, MG Rodrigues Filho, NR Antoniosi Filho,
FSM Sinfrônio, R Rosenhaim, LEB Soledade, IMG Santos and AG Souza. 2011. "Evaluation of the oxidative stability of corn biodiesel." Fuel 90 (2): 773-778.
Darnoko, D and Munir Cheryan. 2000. "Kinetics of palm oil transesterification in a
batch reactor." Journal of the American Oil Chemists' Society 77 (12): 1263-1267.
Davis, JP, D Geller, WH Faircloth and TH Sanders. 2009. "Comparisons of biodiesel
produced from unrefined oils of different peanut cultivars." Journal of the American Oil Chemists' Society 86 (4): 353-361.
de Oliveira, Débora, Marco Di Luccio, Carina Faccio, Clarissa Dalla Rosa, João Paulo
Bender, Nádia Lipke, Cristiana Amroginski, Cláudio Dariva and José Vladimir de Oliveira. 2005. "Optimization of alkaline transesterification of soybean oil and castor oil for biodiesel production." In Twenty-Sixth Symposium on Biotechnology for Fuels and Chemicals, edited, 553-560: Springer.
Deh Kiani, M Kiani, B Ghobadian, T Tavakoli, AM Nikbakht and G Najafi. 2010.
"Application of artificial neural networks for the prediction of performance and exhaust emissions in SI engine using ethanol-gasoline blends." Energy 35 (1): 65-69.
Deligiannis, A, A Papazafeiropoulou, G Anastopoulos and F Zannikos. 2011. "Waste
coffee grounds as an energy feedstock." In Proceedings of the 3rd International CEMEPE & SECOTOX Conference Skiathos, ISBN, edited, 978-960.
Demirbas, A. 2009a. "Potential resources of non-edible oils for biodiesel." Energy
Sources, Part B 4 (3): 310-314.
Demirbas, Ayhan. 2003. "Chemical and fuel properties of seventeen vegetable oils."
Energy Sources 25 (7): 721-728.
236 6 Bibliography
Demirbas, Ayhan. 2005. "Biodiesel production from vegetable oils via catalytic and non-catalytic supercritical methanol transesterification methods." Progress in energy and combustion science 31 (5): 466-487.
Demirbas, Ayhan. 2007. "Progress and recent trends in biofuels." Progress in Energy
and Combustion Science 33 (1): 1-18.
Demirbas, Ayhan. 2008a. "Comparison of transesterification methods for production of
biodiesel from vegetable oils and fats." Energy Conversion and Management 49 (1): 125-130.
Demirbas, Ayhan. 2008b. "Relationships derived from physical properties of vegetable
oil and biodiesel fuels." Fuel 87 (8): 1743-1748.
Demirbas, Ayhan. 2009b. "Biodiesel from waste cooking oil via base-catalytic and
supercritical methanol transesterification." Energy Conversion and Management 50 (4): 923-927.
Demirbaş, Ayhan. 2002. "Biodiesel from vegetable oils via transesterification in
supercritical methanol." Energy conversion and management 43 (17): 2349-2356.
Demirbas, Ayhan and M Fatih Demirbas. 2011. "Importance of algae oil as a source of
biodiesel." Energy Conversion and Management 52 (1): 163-170.
Devan, PK and NV Mahalakshmi. 2009. "Study of the performance, emission and
combustion characteristics of a diesel engine using poon oil-based fuels." Fuel processing technology 90 (4): 513-519.
Dias, Joana M, Maria Alvim-Ferraz and Manuel F Almeida. 2009. "Production of
biodiesel from acid waste lard." Bioresource technology 100 (24): 6355-6361.
Do, Linh D, Vinay Singh, Lixia Chen, Tohren CG Kibbey, Sub R Gollahalli and David
A Sabatini. 2011. "Algae, Canola, or Palm oils—Diesel microemulsion fuels: Phase behaviors, viscosity, and combustion properties." International Journal of Green Energy 8 (7): 748-767.
Dorado, M Pilar, Evaristo Ballesteros, Francisco J López and Martin Mittelbach. 2004.
"Optimization of alkali-catalyzed transesterification of Brassica C arinata oil for biodiesel production." Energy & Fuels 18 (1): 77-83.
Dorado, MP, E Ballesteros, JM Arnal, J Gomez and FJ Lopez Gimenez. 2003. "Testing
waste olive oil methyl ester as a fuel in a diesel engine." Energy & Fuels 17 (6): 1560-1565.
6 Bibliography 237
Dorado, MP, E Ballesteros, JA De Almeida, C Schellert, HP Löhrlein and R Krause. 2002. "An alkali-catalyzed transesterification process for high free fatty acid waste oils." Transactions of the ASAE 45 (3): 525-529.
Dos Santos, ICF, SHV De Carvalho, JI Solleti, W Ferreira de La Salles, K Teixeira da
Silva de La Salles and SMP Meneghetti. 2008. "Studies of< i> Terminalia catappa</i> L. oil: Characterization and biodiesel production." Bioresource technology 99 (14): 6545-6549.
Duncan, Andrew M, Azita Ahosseini, Reece McHenry, Christopher D Depcik, Susan M
Stagg-Williams and Aaron M Scurto. 2010. "High-pressure viscosity of biodiesel from soybean, canola, and coconut oils." Energy & Fuels 24 (10): 5708-5716.
Dunn, RO and BR Moser. 2010. "Cold flow properties and performance of biodiesel."
The Biodiesel Handbook, 2nd edn, AOCS Press, Urbana: 147-203.
Durrett, Timothy P, Christoph Benning and John Ohlrogge. 2008. "Plant triacylglycerols
as feedstocks for the production of biofuels." The Plant Journal 54 (4): 593-607.
Ejim, C. E., B. A. Fleck and A. Amirfazli. 2007. "Analytical study for atomization of
biodiesels and their blends in a typical injector: Surface tension and viscosity effects." Fuel 86 (10–11): 1534-1544.
El Diwani, G. and S. El Rafie. 2008. "Modification of thermal and oxidative properties
of biodiesel produced from vegetable oils." International Journal of Environmental Science & Technology 5 (3): 391-400.
Ellabban, Omar, Haitham Abu-Rub and Frede Blaabjerg. 2014. "Renewable energy
resources: Current status, future prospects and their enabling technology." Renewable and Sustainable Energy Reviews 39 (0): 748-764.
Encinar, Jose M, Juan F Gonzalez and Antonio Rodríguez-Reinares. 2005. "Biodiesel
from used frying oil. Variables affecting the yields and characteristics of the biodiesel." Industrial & Engineering Chemistry Research 44 (15): 5491-5499.
Espinasse, Bernard, Guy Picolet and Eugène Chouraqui. 1997. "Negotiation support
systems: A multi-criteria and multi-agent approach." European Journal of Operational Research 103 (2): 389-409.
Fazal, MA, ASMA Haseeb and HH Masjuki. 2011. "Biodiesel feasibility study: An
evaluation of material compatibility; performance; emission and engine durability." Renewable and Sustainable Energy Reviews 15 (2): 1314-1324.
Feitosa, Filipe X, Marcelo de L Rodrigues, Carolina B Veloso, Célio L Cavalcante Jr,
Mônica CG Albuquerque and Hosiberto B de Sant’Ana. 2010. "Viscosities and
238 6 Bibliography
densities of binary mixtures of coconut+ colza and coconut+ soybean biodiesel at various temperatures." Journal of Chemical & Engineering Data 55 (9): 3909-3914.
Ferella, F, G Mazziotti Di Celso, I De Michelis, V Stanisci and F Vegliò. 2010.
"Optimization of the transesterification reaction in biodiesel production." Fuel 89 (1): 36-42.
Ferguson, Colin R and Alan T Kirkpatrick. "Internal combustion engines, 2001": John
Wiley & Sons, New York, New York.
Fernando, Sandun, Prashanth Karra, Rafael Hernandez and Saroj Kumar Jha. 2007.
"Effect of incompletely converted soybean oil on biodiesel quality." Energy 32 (5): 844-851.
Figueira, Josâe, Salvatore Greco and Matthias Ehrgott. 2005. Multiple criteria decision
analysis: state of the art surveys. Vol. 78: Springer.
Friday, JB and Dana Okano. 2006. "Calophyllum inophyllum (kamani)." Species
Profiles for Pacific Island Agroforestry 2 (1): 1-17.
Frijters, PJM and RSG Baert. 2006. "Oxygenated fuels for clean heavy-duty diesel
engines." International journal of vehicle design 41 (1): 242-255.
Fröhlich, A and B Rice. 2005. "Evaluation of Camelina sativa oil as a feedstock for
biodiesel production." Industrial Crops and Products 21 (1): 25-31.
Gao, Daqi, Zeping Yang, Chaoqian Cai and Fangjun Liu. 2012. "Performance
evaluation of multilayer perceptrons for discriminating and quantifying multiple kinds of odors with an electronic nose." Neural Networks 33: 204-215.
Garshick, Eric, Francine Laden, Jaime E Hart, Bernard Rosner, Thomas J Smith,
Douglas W Dockery and Frank E Speizer. 2004. "Lung cancer in railroad workers exposed to diesel exhaust." Environmental Health Perspectives: 1539-1543.
Geller, Daniel P and John W Goodrum. 2004. "Effects of specific fatty acid methyl
esters on diesel fuel lubricity." Fuel 83 (17): 2351-2356.
Genever, Matt, John Grindrod and Bryce Barker. 2003. "Holocene palynology of
Whitehaven Swamp, Whitsunday Island, Queensland, and implications for the regional archaeological record." Palaeogeography, Palaeoclimatology, Palaeoecology 201 (1): 141-156.
6 Bibliography 239
Gerpen, Jon Van. 2005. "Biodiesel processing and production." Fuel Processing Technology 86 (10): 1097-1107.
Ghadge, Shashikant Vilas and Hifjur Raheman. 2005. "Biodiesel production from
mahua ( Madhuca indica) oil having high free fatty acids." Biomass and Bioenergy 28 (6): 601-605.
Giechaskiel, B., B. Alfoldy and Y. Drossinos. 2009. "A metric for health effects studies
of diesel exhaust particles." Journal of Aerosol Science 40 (8): 639-651.
Giwa, Solomon and Clement Ogunbona. 2014. "Sweet almond (Prunus amygdalus"
dulcis") seeds as a potential feedstock for Nigerian Biodiesel Automotive Project." Revista Ambiente & Água 9 (1): 37-45.
Godiganur, Sharanappa, CH Suryanarayana Murthy and Rana Prathap Reddy. 2009.
"6BTA 5.9 G2-1 Cummins engine performance and emission tests using methyl ester mahua (Madhuca indica) oil/diesel blends." Renewable Energy 34 (10): 2172-2177.
Goodrum, John W and Daniel P Geller. 2005. "Influence of fatty acid methyl esters
from hydroxylated vegetable oils on diesel fuel lubricity." Bioresource Technology 96 (7): 851-855.
Gopinath, A, Sukumar Puhan and G Nagarajan. 2009. "Relating the cetane number of
biodiesel fuels to their fatty acid composition: a critical study." Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering 223 (4): 565-583.
Gouveia, Luisa and Ana Cristina Oliveira. 2009. "Microalgae as a raw material for
biofuels production." Journal of industrial microbiology & biotechnology 36 (2): 269-274.
Graboski, Michael S and Robert L McCormick. 1998. "Combustion of fat and vegetable
oil derived fuels in diesel engines." Progress in energy and combustion science 24 (2): 125-164.
Graboski, MS, RL McCormick, TL Alleman and AM Herring. 2003. "The effect of
biodiesel composition on engine emissions from a DDC series 60 diesel engine." National Renewable Energy Laboratory (Report No: NREL/SR-510-31461).
Gumus, M and S Kasifoglu. 2010. "Performance and emission evaluation of a
compression ignition engine using a biodiesel (apricot seed kernel oil methyl ester) and its blends with diesel fuel." Biomass and bioenergy 34 (1): 134-139.
240 6 Bibliography
Gunstone, Frank. 2011. Vegetable oils in food technology: composition, properties and uses: John Wiley & Sons.
Guzatto, Rafael, Tiago Luis De Martini and Dimitrios Samios. 2011. "The use of a
modified TDSP for biodiesel production from soybean, linseed and waste cooking oil." Fuel Processing Technology 92 (10): 2083-2088.
Haagenson, Darrin M, Rachel L Brudvik, Hongjian Lin and Dennis P Wiesenborn.
2010. "Implementing an in situ alkaline transesterification method for canola biodiesel quality screening." Journal of the American Oil Chemists' Society 87 (11): 1351-1358.
Haas, Michael J, Andrew J McAloon, Winnie C Yee and Thomas A Foglia. 2006. "A
process model to estimate biodiesel production costs." Bioresource technology 97 (4): 671-678.
Haas, Michael J. 2005. "Improving the economics of biodiesel production through the
use of low value lipids as feedstocks: vegetable oil soapstock." Fuel Processing Technology 86 (10): 1087-1096.
Hansen, AC, MR Gratton and W Yuan. 2006. "Diesel engine performance and NOX
emissions from oxygenated biofuels and blends with diesel fuel." Transactions of the ASABE 49 (3): 589-595.
Haseeb, A. S. M. A., M. A. Fazal, M. I. Jahirul and H. H. Masjuki. 2011a.
"Compatibility of automotive materials in biodiesel: A review." Fuel 90 (3): 922-931.
Haseeb, ASMA, MA Fazal, MI Jahirul and HH Masjuki. 2011b. "Compatibility of
automotive materials in biodiesel: a review." Fuel 90 (3): 922-931.
Haşimoğlu, Can, Murat Ciniviz, İbrahim Özsert, Yakup İçingür, Adnan Parlak and M
Sahir Salman. 2008. "Performance characteristics of a low heat rejection diesel engine operating with biodiesel." Renewable Energy 33 (7): 1709-1715.
Heintzenberg, Jost. 1994. "Properties of the Log-Normal Particle Size Distribution."
Aerosol Science and Technology 21 (1): 46-48. doi: 10.1080/02786829408959695.
Herner, J. D., S. H. Hu, W. H. Robertson, T. Huai, M. C. O. Chang, P. Rieger and A.
Ayala. 2011. "Effect of Advanced Aftertreatment for PM and NO(x) Reduction on Heavy-Duty Diesel Engine Ultrafine Particle Emissions." Environmental science & technology 45 (6): 2413-2419.
6 Bibliography 241
Heywood, John B. 1988. Internal combustion engine fundamentals. Vol. 930: Mcgraw-hill New York.
Hoda, N. 2010. "Optimization of biodiesel production from cottonseed oil by
transesterification using NaOH and methanol." Energy Sources, Part A: Recovery, Utilization, and Environmental Effects 32 (5): 434-441.
Hoekman, S Kent and Curtis Robbins. 2012. "Review of the effects of biodiesel on NOx
emissions." Fuel Processing Technology 96: 237-249.
Hoekman, S. Kent, Amber Broch, Curtis Robbins, Eric Ceniceros and Mani Natarajan.
2012. "Review of biodiesel composition, properties, and specifications." Renewable and Sustainable Energy Reviews 16 (1): 143-169.
Holser, Ronald Alan and Rogers Harry-O’Kuru. 2006a. "Transesterified milkweed (
Asclepias) seed oil as a biodiesel fuel." Fuel 85 (14): 2106-2110.
Holser, Ronald Alan and Rogers Harry-O’Kuru. 2006b. "Transesterified milkweed
(Asclepias) seed oil as a biodiesel fuel." Fuel 85 (14–15): 2106-2110.
Hosen, Mohammad Anwar, Abbas Khosravi, Saeid Nahavandi and Douglas Creighton.
2014. "Prediction interval-based neural network modelling of polystyrene polymerization reactor–A new perspective of data-based modelling." Chemical Engineering Research and Design.
Hossain, ABM Sharif, Aishah Salleh, Amru Nasrulhaq Boyce, Partha Chowdhury and
Mohd Naqiuddin. 2008. "Biodiesel fuel production from algae as renewable energy." American journal of biochemistry and biotechnology 4 (3): 250-254.
Hu, Jianbo, Zexue Du, Changxiu Li and Enze Min. 2005. "Study on the lubrication
properties of biodiesel as fuel lubricity enhancers." Fuel 84 (12): 1601-1606.
Hu, Qiang, Milton Sommerfeld, Eric Jarvis, Maria Ghirardi, Matthew Posewitz, Michael
Seibert and Al Darzins. 2008. "Microalgal triacylglycerols as feedstocks for biofuel production: perspectives and advances." The Plant Journal 54 (4): 621-639.
Hyland, BPM, T Whiffin, DC Christophel, B Gray and RW Elick. 2003. "Australian
Tropical Rainforest Plants." Trees, Shrubs and Vines. CSIRO, Melbourne.
Ikwuagwu, OE, IC Ononogbu and OU Njoku. 2000. "Production of biodiesel using
rubber [Hevea brasiliensis (Kunth. Muell.)] seed oil." Industrial crops and products 12 (1): 57-62.
242 6 Bibliography
Ilkılıç, Cumali, Selman Aydın, Rasim Behcet and Hüseyin Aydin. 2011. "Biodiesel from safflower oil and its application in a diesel engine." Fuel Processing Technology 92 (3): 356-362.
Islam, Muhammad, Marie Magnusson, Richard Brown, Godwin Ayoko, Md. Nabi and
Kirsten Heimann. 2013. "Microalgal Species Selection for Biodiesel Production Based on Fuel Properties Derived from Fatty Acid Profiles." Energies 6 (11): 5676-5702.
Issariyakul, Titipong, Mangesh G. Kulkarni, Ajay K. Dalai and Narendra N. Bakhshi.
2007. "Production of biodiesel from waste fryer grease using mixed methanol/ethanol system." Fuel Processing Technology 88 (5): 429-436.
Jacobson, M. Z. 2001. "Strong radiative heating due to the mixing state of black carbon
in atmospheric aerosols." Nature 409 (6821): 695-697.
Jahirul, M I, HH Masjuki, R Saidur, MA Kalam, MH Jayed and MA Wazed. 2010.
"Comparative engine performance and emission analysis of CNG and gasoline in a retrofitted car engine." Applied Thermal Engineering 30 (14): 2219-2226.
Jahirul, M Islam, J Richard Brown, W Senadeera, N Ashwath, C Laing, J Leski-Taylor
and MG Rasul. 2013. "Optimisation of Bio-Oil Extraction Process from Beauty Leaf (Calophyllum Inophyllum) Oil Seed as a Second Generation Biodiesel Source." Procedia Engineering 56: 619-624.
Jahirul, M.I.; Brown, R.J.; Senadeera, W.; O'Hara, I.M.; Ristovski, Z.D. 2013. "The Use
of Artificial Neural Networks for Identifying Sustainable Biodiesel Feedstocks. ." Energies 6: 3764 - 3806.
Jahirul, MI, R Saidur, M Hasanuzzaman, HH Masjuki and MA Kalam. 2007. "A
comparison of the air pollution of gasoline and CNG driven car for Malaysia." International Journal of Mechanical and Materials Engineering 2 (2): 130-138.
Jahirul, Mohammad I, Mohammad G Rasul, Ashfaque Ahmed Chowdhury and
Nanjappa Ashwath. 2012. "Biofuels production through biomass pyrolysis—a technological review." Energies 5 (12): 4952-5001.
Jahirul, Mohammed, Richard Brown, Wijitha Senadeera, #039, Ian Hara and Zoran
Ristovski. 2013. "The Use of Artificial Neural Networks for Identifying Sustainable Biodiesel Feedstocks." Energies 6 (8): 3764-3806.
Jahirul, Mohammed I, Richard J Brown, Wijitha Senadeera, Ian M O'Hara and Zoran D
Ristovski. 2013. "The use of artificial neural networks for identifying sustainable biodiesel feedstocks." Energies 6 (8): 3764-3806.
6 Bibliography 243
Jain, Siddharth and MP Sharma. 2012. "Application of thermogravimetric analysis for thermal stability of Jatropha Curcas biodiesel." Fuel 93: 252-257.
Jamriska, M., L. Morawska, S. Thomas and C He. 2004. "Diesel bus emissions
measured in a tunnel study." Environ. Sci. Technol. 38 (24): 6701–6709.
Jayed, MH, HH Masjuki, R Saidur, MA Kalam and Mohammed I Jahirul. 2009.
"Environmental aspects and challenges of oilseed produced biodiesel in Southeast Asia." Renewable and Sustainable Energy Reviews 13 (9): 2452-2462.
Jeong, Gwi-Taek and Don-Hee Park. 2009. "Optimization of biodiesel production from
castor oil using response surface methodology." Applied biochemistry and biotechnology 156 (1-3): 1-11.
Jham, Gulab N, Bryan R Moser, Shailesh N Shah, Ronald A Holser, Onkar D Dhingra,
Steven F Vaughn, Mark A Berhow, Jill K Winkler-Moser, Terry A Isbell and Ray K Holloway. 2009. "Wild Brazilian mustard (Brassica juncea L.) seed oil methyl esters as biodiesel fuel." Journal of the American Oil Chemists' Society 86 (9): 917-926.
Joshi, Rushang M and Michael J Pegg. 2007a. "Flow properties of biodiesel fuel blends
at low temperatures." Fuel 86 (1): 143-151.
Joshi, Rushang M. and Michael J. Pegg. 2007b. "Flow properties of biodiesel fuel
blends at low temperatures." Fuel 86 (1–2): 143-151.
Kalam, M. A. and H. H. Masjuki. 2002a. "Biodiesel from palmoil - an analysis of its
properties and potential." Biomass & Bioenergy 23 (6): 471-479.
Kalam, MA and HH Masjuki. 2002b. "Biodiesel from palmoil—an analysis of its
properties and potential." Biomass and Bioenergy 23 (6): 471-479.
Kalligeros, S, F Zannikos, S Stournas, E Lois, G Anastopoulos, CH Teas and F
Sakellaropoulos. 2003. "An investigation of using biodiesel/marine diesel blends on the performance of a stationary diesel engine." Biomass and Bioenergy 24 (2): 141-149.
Kalogirou, Soteris A. 2003. "Artificial intelligence for the modeling and control of
combustion processes: a review." Progress in Energy and Combustion Science 29 (6): 515-566.
Kapdan, Ilgi Karapinar and Fikret Kargi. 2006. "Bio-hydrogen production from waste
materials." Enzyme and microbial technology 38 (5): 569-582.
244 6 Bibliography
Kapilan, N and RP Reddy. 2008. "Evaluation of methyl esters of mahua oil (Madhuca indica) as diesel fuel." Journal of the American Oil Chemists' Society 85 (2): 185-188.
Kaplan, Cafer, Ridvan Arslan and Ali Sürmen. 2006. "Performance characteristics of
sunflower methyl esters as biodiesel." Energy Sources 28 (8): 751-755.
Karabektas, Murat. 2009. "The effects of turbocharger on the performance and exhaust
emissions of a diesel engine fuelled with biodiesel." Renewable Energy 34 (4): 989-993.
Karavalakis, G., V. Boutsika, S. Stournas and E. Bakeas. 2011. "Biodiesel emissions
profile in modern diesel vehicles. Part 2: Effect of biodiesel origin on carbonyl, PAH, nitro-PAH and oxy-PAH emissions." Science of the Total Environment 409 (4): 738-747.
Karavalakis, George, Stamoulis Stournas and Evangelos Bakeas. 2009. "Light vehicle
regulated and unregulated emissions from different biodiesels." Science of the Total Environment 407 (10): 3338-3346.
Karmakar, Aninidita, Subrata Karmakar and Souti Mukherjee. 2010. "Properties of
various plants and animals feedstocks for biodiesel production." Bioresource technology 101 (19): 7201-7210.
Karonis, D, E Lois, F Zannikos, A Alexandridis and H Sarimveis. 2003. "A neural
network approach for the correlation of exhaust emissions from a diesel engine with diesel fuel properties." Energy & fuels 17 (5): 1259-1265.
Kaul, Savita, R. C. Saxena, Ajay Kumar, M. S. Negi, A. K. Bhatnagar, H. B. Goyal and
A. K. Gupta. 2007a. "Corrosion behavior of biodiesel from seed oils of Indian origin on diesel engine parts." Fuel Processing Technology 88 (3): 303-307.
Kaul, Savita, RC Saxena, Ajay Kumar, MS Negi, AK Bhatnagar, HB Goyal and AK
Gupta. 2007b. "Corrosion behavior of biodiesel from seed oils of Indian origin on diesel engine parts." Fuel Processing Technology 88 (3): 303-307.
Kaya, Canan, Candan Hamamci, Akin Baysal, Osman Akba, Sait Erdogan and
Abdurrahman Saydut. 2009. "Methyl ester of peanut (Arachis hypogea L.) seed oil as a potential feedstock for biodiesel production." Renewable Energy 34 (5): 1257-1260.
Kesgin, Ugur. 2004. "Genetic algorithm and artificial neural network for engine
optimisation of efficiency and NOx emission." Fuel 83 (7): 885-895.
6 Bibliography 245
Khan, Shakeel A, Mir Z Hussain, S Prasad and UC Banerjee. 2009. "Prospects of biodiesel production from microalgae in India." Renewable and Sustainable Energy Reviews 13 (9): 2361-2372.
Kinast, JA. 2003a. "Production of biodiesels from multiple feedstocks and properties of
biodiesels and biodiesel/diesel blends." NREL Final Report, SR-510-31460.
Kinast, JA. 2003b. Production of Biodiesels from Multiple Feedstocks and Properties of
Biodiesels and Biodiesel/Diesel Blends: Final Report; Report 1 in a Series of 6: National Renewable Energy Lab., Golden, CO.(US).
Kingsbury, K. 2012. "After the Oil Crisis, a Food Crisis".
Kingsbury, K. 2007. "After the oil crises, a food crisis?". In Time. New York City, USA:
Time-Warner publication.
Kinoshita, E, T Myo, K Hamasaki and S Nishi. 2007. Combustion characteristics of
diesel engine with coconut oil ethyl ester: SAE Technical Paper.
Knothe, Gerhard. 2005. "Dependence of biodiesel fuel properties on the structure of
fatty acid alkyl esters." Fuel processing technology 86 (10): 1059-1070.
Knothe, Gerhard. 2008. "“Designer” biodiesel: optimizing fatty ester composition to
improve fuel properties†." Energy & Fuels 22 (2): 1358-1364.
Knothe, Gerhard. 2009. "Improving biodiesel fuel properties by modifying fatty ester
composition." Energy & Environmental Science 2 (7): 759-766.
Knothe, Gerhard and Robert O Dunn. 2003. "Dependence of oil stability index of fatty
compounds on their structure and concentration and presence of metals." Journal of the American Oil Chemists' Society 80 (10): 1021-1026.
Knothe, Gerhard, Andrew C Matheaus and Thomas W Ryan III. 2003. "Cetane numbers
of branched and straight-chain fatty esters determined in an ignition quality tester." Fuel 82 (8): 971-975.
Knothe, Gerhard and Kevin R Steidley. 2005. "Kinematic viscosity of biodiesel fuel
components and related compounds. Influence of compound structure and comparison to petrodiesel fuel components." Fuel 84 (9): 1059-1065.
Knothe, Gerhard and Kevin R Steidley. 2007. "Kinematic viscosity of biodiesel
components (fatty acid alkyl esters) and related compounds at low temperatures." Fuel 86 (16): 2560-2567.
246 6 Bibliography
Koçak, Mevlüt Süreyya, Erol Ileri and Zafer Utlu. 2007. "Experimental study of emission parameters of biodiesel fuels obtained from canola, hazelnut, and waste cooking oils." Energy & fuels 21 (6): 3622-3626.
Korres, DM, G Anastopoulos, E Lois, A Alexandridis, H Sarimveis and G Bafas. 2002.
"A neural network approach to the prediction of diesel fuel lubricity." Fuel 81 (10): 1243-1250.
Kousoulidou, Marina, Georgios Fontaras, Leonidas Ntziachristos and Zissis Samaras.
2010. "Biodiesel blend effects on common-rail diesel combustion and emissions." Fuel 89 (11): 3442-3449.
Krahl, J, K Baum, U Hackbarth, HE Jeberien, A Munack, C Schütt, O Schröder, N
Walter, J Bünger and MM Müller. 1999. "Gaseous compounds, ozone precursors, particle number and particle size distributions, and mutagenic effects due to biodiesel." In ASAE/CSAE-SCGR Annual International Meeting, Toronto, Ontario, Canada, 18-21 July, 1999., edited: American Society of Agricultural Engineers (ASAE).
Kumar, A.S.; Maheswar, D.; Reddy, K.V.K. . 2009a. "Comparison of diesel engine
performance and emission from neat and transesterified cotton seed oil." Jordan J. Mech. Ind.Eng 3: 190–197.
Kumar, Ashwani and Satyawati Sharma. 2011. "Potential non-edible oil resources as
biodiesel feedstock: an Indian perspective." Renewable and Sustainable Energy Reviews 15 (4): 1791-1800.
Kumar, Gajendra, D Kumar, Shailandra Singh, S Kothari, Sumit Bhatt and Chandra P
Singh. 2010. "Continuous low cost transesterification process for the production of coconut biodiesel." Energies 3 (1): 43-56.
Kumar, J; Bansal, A; Jha, MK. "Comparison of Statistical and Neural Network
Techniques in Predicting Physical Properties of Various Mixtures of Diesel and Biodiesel." Paper presented at the World Congress on Engineering and Computer Science, San Francisco, USA, October 24-26, 2000.
Kumar, Jatinder and Ajay Bansal. 2010. "Application of artificial neural network to
predict properties of diesel-biodiesel blends." Kathmandu University Journal of Science, Engineering and Technology 6 (2): 98-103.
Kumar, Jatinder, Ajay Bansal and MK Jha. 2007. "Comparison of statistical and neural
network techniques in predicting physical properties of various mixtures of diesel and biodiesel." In Proceedings of the World Congress on Engineering and Computer Science, San Francisco, CA, USA, edited, 24-26: Citeseer.
6 Bibliography 247
Kumar, K Vasanth. 2009b. "Neural network prediction of interfacial tension at crystal/solution interface." Industrial & Engineering Chemistry Research 48 (8): 4160-4164.
Kumar Tiwari, Alok, Akhilesh Kumar and Hifjur Raheman. 2007. "Biodiesel production
from jatropha oil ( Jatropha curcas ) with high free fatty acids: An optimized process." Biomass and Bioenergy 31 (8): 569-575.
Kyriakidis, Nikolaos B and Theophanis Katsiloulis. 2000. "Calculation of iodine value
from measurements of fatty acid methyl esters of some oils: comparison with the relevant American oil chemists society method." Journal of the American Oil Chemists' Society 77 (12): 1235-1238.
Labeckas, Gvidonas and Stasys Slavinskas. 2006. "The effect of rapeseed oil methyl
ester on direct injection diesel engine performance and exhaust emissions." Energy Conversion and Management 47 (13): 1954-1967.
Lacey, Paul I and Sidney J Lestz. 1992. Effect of Low-Lubricity Fuels on Diesel
Injection Pumps-Part I: Field Performance: SAE Technical Paper.
Lam, Man Kee, Keat Teong Lee and Abdul Rahman Mohamed. 2010. "Homogeneous,
heterogeneous and enzymatic catalysis for transesterification of high free fatty acid oil (waste cooking oil) to biodiesel: a review." Biotechnology advances 28 (4): 500-518.
Lance, DL, CL Goodfellow, J Williams, W Bunting, I Sakata, K Yoshida, S Taniguchi
and K Kitano. 2009. The impact of diesel and biodiesel fuel composition on a Euro V HSDI engine with advanced DPNR emissions control: SAE Technical Paper.
Lang, X, AK Dalai, NN Bakhshi, MJ Reaney and PB Hertz. 2001. "Preparation and
characterization of bio-diesels from various bio-oils." Bioresource technology 80 (1): 53-62.
Lapuerta, M., O. Armas and J. Rodríguez-Fernández. 2008a. "Effect of biodiesel fuels
on diesel engine emissions." Progress in Energy and Combustion Science 34 (2): 198-223.
Lapuerta, Magín, Octavio Armas and Rosario Ballesteros. 2002. Diesel particulate
emissions from biofuels derived from Spanish vegetable oils: SAE Technical Paper.
Lapuerta, Magin, Octavio Armas and Jose Rodriguez-Fernandez. 2008. "Effect of
biodiesel fuels on diesel engine emissions." Progress in energy and combustion science 34 (2): 198-223.
248 6 Bibliography
Lapuerta, Magín, Octavio Armas and José Rodríguez-Fernández. 2008b. "Effect of the
Degree of Unsaturation of Biodiesel Fuels on NOx and Particulate Emissions." SAE Int. J. Fuels Lubr. 1 (1): 1150-1158.
Lapuerta, Magín, José M Herreros, Lisbeth L Lyons, Reyes García-Contreras and
Yolanda Briceño. 2008. "Effect of the alcohol type used in the production of waste cooking oil biodiesel on diesel performance and emissions." Fuel 87 (15): 3161-3169.
Lebedevas, Sergejus and Andrius Vaicekauskas. 2006. "Research into the application of
biodiesel in the transport sector of Lithuania." Transport 21 (2): 80-87.
Lebedevas, Sergejus, Andrius Vaicekauskas, Galina Lebedeva, Violeta Makareviciene,
Prutenis Janulis and Kiril Kazancev. 2006. "Use of waste fats of animal and vegetable origin for the production of biodiesel fuel: quality, motor properties, and emissions of harmful components." Energy & fuels 20 (5): 2274-2280.
Lee, Inmok, Lawrence A Johnson and Earl G Hammond. 1995. "Use of branched-chain
esters to reduce the crystallization temperature of biodiesel." Journal of the American Oil Chemists’ Society 72 (10): 1155-1160.
Lee, Ki-Teak, Thomas A Foglia and Kyu-Seob Chang. 2002. "Production of alkyl ester
as biodiesel from fractionated lard and restaurant grease." Journal of the American Oil Chemists' Society 79 (2): 191-195.
Lee, Seang-wock, Daisuke Tanaka, Jin Kusaka and Yasuhiro Daisho. 2002. "Effects of
diesel fuel characteristics on spray and combustion in a diesel engine." JSAE review 23 (4): 407-414.
Leung, D. Y. C. and Y. Guo. 2006a. "Transesterification of neat and used frying oil:
Optimization for biodiesel production." Fuel Processing Technology 87 (10): 883-890.
Leung, DYC and Y Guo. 2006b. "Transesterification of neat and used frying oil:
optimization for biodiesel production." Fuel Processing Technology 87 (10): 883-890.
Lim, Steven and Lee Keat Teong. 2010. "Recent trends, opportunities and challenges of
biodiesel in Malaysia: An overview." Renewable and Sustainable Energy Reviews 14 (3): 938-954.
Lin, Bai-Fu, Jyun-Han Huang and Dao-Yi Huang. 2009. "Experimental study of the
effects of vegetable oil methyl ester on DI diesel engine performance characteristics and pollutant emissions." Fuel 88 (9): 1779-1785.
6 Bibliography 249
Lin, Cherng-Yuan and Rong-Ji Li. 2009a. "Engine performance and emission
characteristics of marine fish-oil biodiesel produced from the discarded parts of marine fish." Fuel Processing Technology 90 (7): 883-888.
Lin, Cherng-Yuan and Rong-Ji Li. 2009b. "Fuel properties of biodiesel produced from
the crude fish oil from the soapstock of marine fish." Fuel Processing Technology 90 (1): 130-136.
Lin, Cherng-Yuan, Hsiu-An Lin and Lang-Bang Hung. 2006. "Fuel structure and
properties of biodiesel produced by the peroxidation process." Fuel 85 (12): 1743-1749.
Lin, Lin, Zhou Cunshan, Saritporn Vittayapadung, Shen Xiangqian and Dong
Mingdong. 2011. "Opportunities and challenges for biodiesel fuel." Applied Energy 88 (4): 1020-1031.
Liu, Guozhu, Li Wang, Haijie Qu, Huiming Shen, Xiangwen Zhang, Shuting Zhang and
Zhentao Mi. 2007. "Artificial neural network approaches on composition–property relationships of jet fuels based on GC–MS." Fuel 86 (16): 2551-2559.
Loh, Soh-Kheang, Soo-Mooi Chew and Yuen-May Choo. 2006. "Oxidative stability and
storage behavior of fatty acid methyl esters derived from used palm oil." Journal of the American Oil Chemists' Society 83 (11): 947-952.
Luque, Rafael, Lorenzo Herrero-Davila, Juan M Campelo, James H Clark, Jose M
Hidalgo, Diego Luna, Jose M Marinas and Antonio A Romero. 2008. "Biofuels: a technological perspective." Energy & Environmental Science 1 (5): 542-564.
Ma, Fangrui and Milford A Hanna. 1999. "Biodiesel production: a review." Bioresource
technology 70 (1): 1-15.
Majewski, WA. 2002. "Diesel exhaust particle size." Dieselnet Technology Guide.
Manjunatha, R, P Badari Narayana and K Hema Chandra Reddy. 2010. "Application of
Artificial Neural Networks for emission modelling of biodiesels for a CI engine under varying operating conditions." modern applied science 4 (3): p77.
Marchetti, J. M., V. U. Miguel and A. F. Errazu. 2008. "Techno-economic study of
different alternatives for biodiesel production." Fuel Processing Technology 89 (8): 740-748.
Marinović, Slavica, Tomislav Bolanča, Šime Ukić, Vinko Rukavina and Ante Jukić.
2012. "Prediction of diesel fuel cold properties using artificial neural networks." Chemistry and technology of fuels and oils 48 (1): 67-74.
250 6 Bibliography
Mata, Teresa M, Antonio A Martins and Nidia S Caetano. 2010. "Microalgae for
biodiesel production and other applications: a review." Renewable and Sustainable Energy Reviews 14 (1): 217-232.
McCormick, R. L., M. S. Graboski, T. L. Alleman, A. M. Herring and K. S. Tyson.
2001. "Impact of biodiesel source material and chemical structure on emissions of criteria pollutants from a heavy-duty engine." Environmental Science and Technology 35 (9): 1742-1747.
Mccormick, Robbert L., Michael S. Graboski, Teresa L. Alleman and Andrew M.
Harrin. 2001. "Impact of biodiesel source material and chemical structure on emissions and criteria pollutants from a heavy duty engine." Environ. Sci. Technol. 35 (9): 1742-1747.
Meher, L. C., D. Vidya Sagar and S. N. Naik. 2006a. "Technical aspects of biodiesel
production by transesterification—a review." Renewable and Sustainable Energy Reviews 10 (3): 248-268.
Meher, LC, D Vidya Sagar and SN Naik. 2006b. "Technical aspects of biodiesel
production by transesterification—a review." Renewable and sustainable energy reviews 10 (3): 248-268.
Metz, B.; Davidson, O.; Bosch, P.; Dave, R.; Meyer, L. . 2007. Fourth Assessment
Report of the Intergovernmental Panel on Climate Change. New York, NY, USA.
Miller, RL, GA Schmidt and DT Shindell. 2006. "Forced annular variations in the 20th
century intergovernmental panel on climate change fourth assessment report models." Journal of Geophysical Research: Atmospheres (1984–2012) 111 (D18).
Minnett, Rupert C. J., Andrew T. Smith, William C. Lennon Jr and Robert Hecht-
Nielsen. 2011a. "Neural network tomography: Network replication from output surface geometry." Neural Networks 24 (5): 484-492.
Minnett, Rupert CJ, Andrew T Smith, William C Lennon Jr and Robert Hecht-Nielsen.
2011b. "Neural network tomography: Network replication from output surface geometry." Neural Networks 24 (5): 484-492.
Mohibbe Azam, M, Amtul Waris and NM Nahar. 2005a. "Prospects and potential of
fatty acid methyl esters of some non-traditional seed oils for use as biodiesel in India." Biomass and Bioenergy 29 (4): 293-302.
6 Bibliography 251
Mohibbe Azam, M., Amtul Waris and N. M. Nahar. 2005b. "Prospects and potential of fatty acid methyl esters of some non-traditional seed oils for use as biodiesel in India." Biomass and Bioenergy 29 (4): 293-302.
Monyem, A, JH Van Gerpen and M Canakcl. 2001. "The effect of timing and oxidation
on emissions from biodiesel–fueled engines." Carbon 44 (1): 35-42.
Monyem, Abdul and Jon H Van Gerpen. 2001. "The effect of biodiesel oxidation on
engine performance and emissions." Biomass and Bioenergy 20 (4): 317-325.
Moraes, Maria Silvana Aranda, Laiza Canielas Krause, Michele Espinosa da Cunha,
Candice Shimitt Faccini, Eliana Weber de Menezes, Renato Cataluña Veses, Maria Regina Alves Rodrigues and Elina Bastos Caramão. 2008. "Tallow biodiesel: properties evaluation and consumption tests in a diesel engine." Energy & Fuels 22 (3): 1949-1954.
Morshed, Mahbub, Kaniz Ferdous, Maksudur R. Khan, M. S. I. Mazumder, M. A. Islam
and Md T. Uddin. 2011. "Rubber seed oil as a potential source for biodiesel production in Bangladesh." Fuel 90 (10): 2981-2986.
Moser, Bryan R. 2008. "Influence of Blending Canola, Palm, Soybean, and Sunflower
Oil Methyl Esters on Fuel Properties of Biodiesel†." Energy & fuels 22 (6): 4301-4306.
Moser, Bryan R. 2011. "Biodiesel production, properties, and feedstocks." In Biofuels,
285-347: Springer.
Moser, Bryan R. 2012. "Preparation of fatty acid methyl esters from hazelnut, high-oleic
peanut and walnut oils and evaluation as biodiesel." Fuel 92 (1): 231-238.
Moser, Bryan R and Steven F Vaughn. 2010. "Evaluation of alkyl esters from Camelina
sativa oil as biodiesel and as blend components in ultra low-sulfur diesel fuel." Bioresource technology 101 (2): 646-653.
Moser, Bryan R. 2014. "Impact of fatty ester composition on low temperature properties
of biodiesel–petroleum diesel blends." Fuel 115 (0): 500-506.
Murillo, S, JL Miguez, J Porteiro, E Granada and JC Moran. 2007. "Performance and
exhaust emissions in the use of biodiesel in outboard diesel engines." Fuel 86 (12): 1765-1771.
Nabi, Md Nurun, Md Mustafizur Rahman and Md Shamim Akhter. 2009. "Biodiesel
from cotton seed oil and its effect on engine performance and exhaust emissions." Applied Thermal Engineering 29 (11): 2265-2270.
252 6 Bibliography
Nakpong, P and S Wootthikanokkhan. 2013. "Optimization of biodiesel production from Jatropha curcas L. oil via alkali-catalyzed methanolysis." Journal of Sustainable Energy & Environment 1 (3): 105-109.
Nakpong, Piyanuch and Sasiwimol Wootthikanokkhan. 2010. "High free fatty acid
coconut oil as a potential feedstock for biodiesel production in Thailand." Renewable Energy 35 (8): 1682-1687.
Nascimento, RSV, VLP Soares, S Albinante and LR Barreto. 2005. "Effect of ester-
additives on the crystallization temperature of methyl hexadecanoate." Journal of thermal analysis and calorimetry 79 (2): 249-254.
Ng, Hoon Kiat and Suyin Gan. 2010. "Combustion performance and exhaust emissions
from the non-pressurised combustion of palm oil biodiesel blends." Applied Thermal Engineering 30 (16): 2476-2484.
Ng, Jo-Han, Hoon Kiat Ng and Suyin Gan. 2012. "Development of emissions predictor
equations for a light-duty diesel engine using biodiesel fuel properties." Fuel 95: 544-552.
No, Soo-Young. 2011. "Inedible vegetable oils and their derivatives for alternative
diesel fuels in CI engines: a review." Renewable and Sustainable Energy Reviews 15 (1): 131-149.
Nogueira Jr, Carlos A, Filipe X Feitosa, Fabiano AN Fernandes, Rílvia S Santiago and
Hosiberto B de Sant’Ana. 2010. "Densities and viscosities of binary mixtures of babassu biodiesel+ cotton seed or soybean biodiesel at different temperatures." Journal of Chemical & Engineering Data 55 (11): 5305-5310.
Obodeh, O and CI Ajuwa. 2009. "Evaluation of artificial neural network performance in
predicting diesel engine NOx emissions." European Journal of Scientific Research 33 (4): 642-653.
Okano, J. B. Friday; Dana. 2006. "Calophyllum inophyllum (kamani)." Species Profiles
for Pacific Island Agroforestry 2 (1).
Okonko, Iheanyi O, Roosevelt A Egwame, Enobong A Fajobi, Aniefiok O Nkang,
Chinedum I Iheakanwa, Tolulope A Ogunnusi and Babatunde I Onajobi. 2009. "Current trends in biofuel production and its use as an alternative energy security." Electronic Journal of Environmental, Agricultural and Food Chemistry 8 (11): 1233-1260.
Öner, Cengiz and Şehmus Altun. 2009. "Biodiesel production from inedible animal
tallow and an experimental investigation of its use as alternative fuel in a direct injection diesel engine." Applied Energy 86 (10): 2114-2120.
6 Bibliography 253
Özcanlı, Mustafa, Ali Keskin and Kadir Aydın. 2011. "Biodiesel production from
terebinth (Pistacia terebinthus) oil and its usage in diesel engine." International Journal of Green Energy 8 (5): 518-528.
Ozsezen, Ahmet Necati, Mustafa Canakci, Ali Turkcan and Cenk Sayin. 2009.
"Performance and combustion characteristics of a DI diesel engine fueled with waste palm oil and canola oil methyl esters." Fuel 88 (4): 629-636.
Pal, Amit, Ashish Verma, SS Kachhwaha and S Maji. 2010. "Biodiesel production
through hydrodynamic cavitation and performance testing." Renewable Energy 35 (3): 619-624.
Park, Ji-Yeon, Deog-Keun Kim, Joon-Pyo Lee, Soon-Chul Park, Young-Joo Kim and
Jin-Suk Lee. 2008. "Blending effects of biodiesels on oxidation stability and low temperature flow properties." Bioresource technology 99 (5): 1196-1203.
Parlak, Adnan, Yasar Islamoglu, Halit Yasar and Aysun Egrisogut. 2006. "Application
of artificial neural network to predict specific fuel consumption and exhaust temperature for a diesel engine." Applied Thermal Engineering 26 (8): 824-828.
Paroissien, Guy. 2012. "ARBORICULTURAL IMPACT REPORT.
Pasadakis, N, S Sourligas and Ch Foteinopoulos. 2006. "Prediction of the distillation
profile and cold properties of diesel fuels using mid-IR spectroscopy and neural networks." Fuel 85 (7): 1131-1137.
Pasadakis, Nikos, Vassilis Gaganis and Charalambos Foteinopoulos. 2006. "Octane
number prediction for gasoline blends." Fuel processing technology 87 (6): 505-509.
Pereira, Roberto G, Cesar D Oliveira, Jorge L Oliveira, Paulo Cesar P Oliveira, Carlos E
Fellows and Oscar E Piamba. 2007. "Exhaust emissions and electric energy generation in a stationary engine using blends of diesel and soybean biodiesel." Renewable Energy 32 (14): 2453-2460.
Pérez, Ángel, Abraham Casas, Carmen María Fernández, María Jesús Ramos and
Lourdes Rodríguez. 2010. "Winterization of peanut biodiesel to improve the cold flow properties." Bioresource technology 101 (19): 7375-7381.
Pham, P. X., T. A. Bodisco, S. Stevanovic, M. D. Rahman, H. Wang, Z. D. Ristovski, R.
J. Brown and A. R. Masri. 2013. "Engine Performance Characteristics for Biodiesels of Different Degrees of Saturation and Carbon Chain Lengths." SAE Int. J. Fuels Lubr. 6 (1): 188-198.
254 6 Bibliography
Phan, Anh N and Tan M Phan. 2008. "Biodiesel production from waste cooking oils." Fuel 87 (17): 3490-3496.
Pinzi, Sara, Paul Rounce, José M. Herreros, Athanasios Tsolakis and M. Pilar Dorado.
2013. "The effect of biodiesel fatty acid composition on combustion and diesel engine exhaust emissions." Fuel 104: 170-182. doi: 10.1016/j.fuel.2012.08.056.
Posten, Clemens and Georg Schaub. 2009. "Microalgae and terrestrial biomass as source
for fuels—a process view." Journal of biotechnology 142 (1): 64-69.
Puhan, Sukumar, R Jegan, K Balasubbramanian and G Nagarajan. 2009. "Effect of
injection pressure on performance, emission and combustion characteristics of high linolenic linseed oil methyl ester in a DI diesel engine." Renewable Energy 34 (5): 1227-1233.
Qi, DH, H Chen, LM Geng and Y Bian. 2010. "Experimental studies on the combustion
characteristics and performance of a direct injection engine fueled with biodiesel/diesel blends." Energy Conversion and Management 51 (12): 2985-2992.
Qi, DH, LM Geng, H Chen, YZH Bian, J Liu and X Ren. 2009. "Combustion and
performance evaluation of a diesel engine fueled with biodiesel produced from soybean crude oil." Renewable Energy 34 (12): 2706-2713.
Quintao, Nara LM, Christiane Meyre-Silva, Gislaine F Silva, Carla S Antonialli, Lilian
W Rocha, Ruth M Lucinda-Silva, Angela Malheiros, Márcia M Souza, Valdir Cechinel Filho and Tania Bresolin. 2011. "Aleurites moluccana (L.) Willd. leaves: mechanical antinociceptive properties of a standardized dried extract and its chemical markers." Evidence-Based Complementary and Alternative Medicine 2011.
Raadnui, Surapol and Anant Meenak. 2003. "Effects of refined palm oil (RPO) fuel on
wear of diesel engine components." Wear 254 (12): 1281-1288.
Radha, KV and G Manikandan. 2011. "Novel production of biofuels from neem oil." In
World Renewable Energy Congress–Sweden, Bioenergy Technology. Linkoping, Sweden, edited, 8-13.
Ragit, SS, SK Mohapatra, K Kundu and Prashant Gill. 2011. "Optimization of neem
methyl ester from transesterification process and fuel characterization as a diesel substitute." Biomass and bioenergy 35 (3): 1138-1144.
Raheman, H and SV Ghadge. 2007. "Performance of compression ignition engine with
mahua Madhuca indica biodiesel." Fuel 86 (16): 2568-2573.
6 Bibliography 255
Raheman, H and AG Phadatare. 2004. "Diesel engine emissions and performance from blends of karanja methyl ester and diesel." Biomass and Bioenergy 27 (4): 393-397.
Rahman, M. M., S. Stevanovic, R. J. Brown and Z. Ristovski. 2013. "Influence of
Different Alternative Fuels on Particle Emission from a Turbocharged Common-Rail Diesel Engine." Procedia Engineering 56 (0): 381-386.
Rajendra, Machavaram, Prakash Chandra Jena and Hifjur Raheman. 2009. "Prediction
of optimized pretreatment process parameters for biodiesel production using ANN and GA." Fuel 88 (5): 868-875.
Rakopoulos, CD, DC Rakopoulos, DT Hountalas, EG Giakoumis and EC Andritsakis.
2008. "Performance and emissions of bus engine using blends of diesel fuel with bio-diesel of sunflower or cottonseed oils derived from Greek feedstock." Fuel 87 (2): 147-157.
Ramadhas, A. S., S. Jayaraj, C. Muraleedharan and K. Padmakumari. 2006a. "Artificial
neural networks used for the prediction of the cetane number of biodiesel." Renewable Energy 31 (15): 2524-2533.
Ramadhas, Arumugam Sakunthalai, Simon Jayaraj and Chandrashekaran
Muraleedharan. 2005. "Biodiesel production from high FFA rubber seed oil." Fuel 84 (4): 335-340.
Ramadhas, AS, S Jayaraj, C Muraleedharan and K Padmakumari. 2006b. "Artificial
neural networks used for the prediction of the cetane number of biodiesel." Renewable Energy 31 (15): 2524-2533.
Ramadhas, AS, C Muraleedharan and S Jayaraj. 2005. "Performance and emission
evaluation of a diesel engine fueled with methyl esters of rubber seed oil." Renewable energy 30 (12): 1789-1800.
Ramalho, EFSM, JR Carvalho Filho, AR Albuquerque, SF de Oliveira, EHS Cavalcanti,
L Stragevitch, IMG Santos and AG Souza. 2012. "Low temperature behavior of poultry fat biodiesel: diesel blends." Fuel 93: 601-605.
Ramírez-Verduzco, Luis Felipe, Javier Esteban Rodríguez-Rodríguez and Alicia del
Rayo Jaramillo-Jacob. 2012. "Predicting cetane number, kinematic viscosity, density and higher heating value of biodiesel from its fatty acid methyl ester composition." Fuel 91 (1): 102-111.
Ramos, María Jesús, Carmen María Fernández, Abraham Casas, Lourdes Rodríguez and
Ángel Pérez. 2009. "Influence of fatty acid composition of raw materials on biodiesel properties." Bioresource Technology 100 (1): 261-268.
256 6 Bibliography
Rashid, Umer and Farooq Anwar. 2008a. "Production of biodiesel through base-
catalyzed transesterification of safflower oil using an optimized protocol." Energy & Fuels 22 (2): 1306-1312.
Rashid, Umer and Farooq Anwar. 2008b. "Production of biodiesel through optimized
alkaline-catalyzed transesterification of rapeseed oil." Fuel 87 (3): 265-273.
Rashid, Umer, Farooq Anwar and Gerhard Knothe. 2009. "Evaluation of biodiesel
obtained from cottonseed oil." Fuel Processing Technology 90 (9): 1157-1163.
Rashid, Umer, Farooq Anwar, Bryan R Moser and Samia Ashraf. 2008. "Production of
sunflower oil methyl esters by optimized alkali-catalyzed methanolysis." Biomass and bioenergy 32 (12): 1202-1205.
Redel-Macías, MD, S Pinzi, MF Ruz, AJ Cubero-Atienza and MP Dorado. 2012.
"Biodiesel from saturated and monounsaturated fatty acid methyl esters and their influence over noise and air pollution." Fuel 97: 751-756.
Refaat, A. A. 2009a. "Correlation between the chemical structure of biodiesel and its
physical properties." International Journal of Environmental Science & Technology 6 (4): 677-694.
Refaat, AA. 2009b. "Correlation between the chemical structure of biodiesel and its
physical properties." International Journal of Environmental Science & Technology 6 (4): 677-694.
Reijnders, Lucas. 2006. "Conditions for the sustainability of biomass based fuel use."
Energy policy 34 (7): 863-876.
Renald, CJ and P Somasundaram. 2012. "Experimental investigation on attenuation of
emission with optimized LPG jet induction in a dual fuel diesel engine and prediction by ANN model." Energy Procedia 14: 1427-1438.
Reyes, JF and MA Sepulveda. 2006. "PM-10 emissions and power of a diesel engine
fueled with crude and refined biodiesel from salmon oil." Fuel 85 (12): 1714-1719.
Ristovski, Z. D., B. Miljevic, N. C. Surawski, L. Morawska, K. M. Fong, F. Goh and I.
A. Yang. 2012. "Respiratory health effects of diesel particulate matter." Respirology 17 (2): 201-212..
Rodrigues Jr, Jorge de A, Fabianne de P Cardoso, Elizabeth R Lachter, Luciana RM
Estevão, Edson Lima and Regina SV Nascimento. 2006. "Correlating chemical
6 Bibliography 257
structure and physical properties of vegetable oil esters." Journal of the American Oil Chemists' Society 83 (4): 353-357.
Roosta, Aliakbar, Payam Setoodeh and Abdolhossein Jahanmiri. 2011. "Artificial neural
network modeling of surface tension for pure organic compounds." Industrial & Engineering Chemistry Research 51 (1): 561-566.
Royon, D, M Daz, G Ellenrieder and S Locatelli. 2007. "Enzymatic production of
biodiesel from cotton seed oil using t-butanol as a solvent." Bioresource Technology 98 (3): 648-653.
S.S. Gilla, A. Tsolakisa, , , J.M. Herrerosb, A.P.E. York. 2011. "Diesel emissions
improvements through the use of biodiesel or oxygenated blending components." Fuel 95, 578-586
Sahoo, PK, LM Das, MKG Babu, P Arora, VP Singh, NR Kumar and TS Varyani. 2009.
"Comparative evaluation of performance and emission characteristics of jatropha, karanja and polanga based biodiesel as fuel in a tractor engine." Fuel 88 (9): 1698-1707.
Sahoo, PK, LM Das, MKG Babu and SN Naik. 2007. "Biodiesel development from high
acid value polanga seed oil and performance evaluation in a CI engine." Fuel 86 (3): 448-454.
Saidur, R, MI Jahirul, M Hasanuzzaman and HH Masjuki. 2008. "Analysis of exhaust
emissions of natural gas engine by using response surface methodology." Journal of Applied Sciences 8 (19): 3328-3339.
Salamanca, M., F. Mondragón, J. R. Agudelo, P. Benjumea and A. Santamaría. 2012.
"Variations in the chemical composition and morphology of soot induced by the unsaturation degree of biodiesel and a biodiesel blend." Combustion and Flame 159 (3): 1100-1108.
Salvi, B. L. and N. L. Panwar. 2012. "Biodiesel resources and production technologies –
A review." Renewable and Sustainable Energy Reviews 16 (6): 3680-3689.
Samios, D, F Pedrotti, A Nicolau, QB Reiznautt, DD Martini and FM Dalcin. 2009a. "A
transesterification double step process—TDSP for biodiesel preparation from fatty acids triglycerides." Fuel Processing Technology 90 (4): 599-605.
Samios, D., F. Pedrotti, A. Nicolau, Q. B. Reiznautt, D. D. Martini and F. M. Dalcin.
2009b. "A Transesterification Double Step Process — TDSP for biodiesel preparation from fatty acids triglycerides." Fuel Processing Technology 90 (4): 599-605.
258 6 Bibliography
Samura, Toshikazu and Hatsuo Hayashi. 2012. "Directional spike propagation in a recurrent network: Dynamical firewall as anisotropic recurrent inhibition." Neural Networks 33: 236-246.
Sanford, Shannon D, James Matthew White, Parag S Shah, Claudia Wee, Marlen A
Valverde and Glen R Meier. 2009. "Feedstock and biodiesel characteristics report." Renewable Energy Group 416.
Sardar, N., M. Ahmad, M. A. Khan, S. Ali, M. Zafar, N. Khalid and S. Sultana. 2011.
"Prospects and potential of non edible neem oil biodiesel based on physico-chemical characterization." Energy Sources, Part A: Recovery, Utilization, and Environmental Effects 33 (15): 1422-1430.
Sarin, Rakesh, Meeta Sharma, S Sinharay and RK Malhotra. 2007. "Jatropha–palm
biodiesel blends: an optimum mix for Asia." Fuel 86 (10): 1365-1371.
Satyanarayana, M and C Muraleedharan. 2010. "Prediction of acid values of vegetable
oils having high free fatty acids using artificial neural networks." Energy Sources, Part A: Recovery, Utilization, and Environmental Effects 32 (16): 1479-1489.
Sayin, Cenk, H Metin Ertunc, Murat Hosoz, Ibrahim Kilicaslan and Mustafa Canakci.
2007. "Performance and exhaust emissions of a gasoline engine using artificial neural network." Applied Thermal Engineering 27 (1): 46-54.
Schenk, Peer M, Skye R Thomas-Hall, Evan Stephens, Ute C Marx, Jan H Mussgnug,
Clemens Posten, Olaf Kruse and Ben Hankamer. 2008. "Second generation biofuels: high-efficiency microalgae for biodiesel production." Bioenergy Research 1 (1): 20-43.
Schmidt, Kevin and Jon Van Gerpen. 1996. The effect of biodiesel fuel composition on
diesel combustion and emissions: SAE Technical Paper.
Schwab, AW, MO Bagby and B Freedman. 1987. "Preparation and properties of diesel
fuels from vegetable oils." Fuel 66 (10): 1372-1378.
Scott, Stuart A, Matthew P Davey, John S Dennis, Irmtraud Horst, Christopher J Howe,
David J Lea-Smith and Alison G Smith. 2010. "Biodiesel from algae: challenges and prospects." Current Opinion in Biotechnology 21 (3): 277-286.
Senatore, Adolfo, Massimo Cardone, Vittorio Rocco and Maria Vittoria Prati. 2000. A
comparative analysis of combustion process in DI diesel engine fueled with biodiesel and diesel fuel: SAE Technical Paper.
6 Bibliography 259
Senthil Kumar, M, Anthony Kerihuel, Jérôme Bellettre and Mohand Tazerout. 2005. "Experimental investigations on the use of preheated animal fat as fuel in a compression ignition engine." Renewable Energy 30 (9): 1443-1456.
Serdari, A, K Fragioudakis, C Teas, F Sakellaropoulos, F Zannikos, S Stournas and E
Lois. 1998. "Adding biodiesel corn oil and sunflower oil to diesel fuel: the impact on the performance of conventional road vehicles." Journal of the Institute of Energy 71 (488): 126-136.
Shah, Shailesh N, Osvaldo K Iha, Flávio CSC Alves, Brajendra K Sharma, Sevim Z
Erhan and Paulo AZ Suarez. 2013. "Potential Application of Turnip Oil (Raphanus sativus L.) for Biodiesel Production: Physical–Chemical Properties of Neat Oil, Biofuels and their Blends with Ultra-Low Sulphur Diesel (ULSD)." BioEnergy Research 6 (2): 841-850.
Shanmugam, P, V Sivakumar, A Murugesan and M Ilangkumaran. 2011. "Performance
and exhaust emissions of a diesel engine using hybrid fuel with an artificial neural network." Energy Sources, Part A: Recovery, Utilization, and Environmental Effects 33 (15): 1440-1450.
Sharma, Y. C., B. Singh and S. N. Upadhyay. 2008. "Advancements in development and
characterization of biodiesel: A review." Fuel 87 (12): 2355-2373.
Sharon, H, R Jayaprakash, A Sundaresan and K Karuppasamy. 2012. "Biodiesel
production and prediction of engine performance using SIMULINK model of trained neural network." Fuel 99: 197-203.
Shi, X., K. He, J. Zhang, Y. Ma, Y. Ge and J. Tan. 2010. "A comparative study of
particle size distribution from two oxygenated fuels and diesel fuel." Frontiers of Environmental Science and Engineering in China 4 (1): 30-34.
Sims, Ralph EH, Warren Mabee, Jack N Saddler and Michael Taylor. 2010. "An
overview of second generation biofuel technologies." Bioresource technology 101 (6): 1570-1580.
Singh, RK and Saroj K Padhi. 2009. "Characterization of Jatropha oil for the preparation
of biodiesel." Nat. product Radiance 8 (2): 127-132.
Singh, SP and Dipti Singh. 2010. "Biodiesel production through the use of different
sources and characterization of oils and their esters as the substitute of diesel: a review." Renewable and Sustainable Energy Reviews 14 (1): 200-216.
Sinha, Shailendra, Avinash Kumar Agarwal and Sanjeev Garg. 2008. "Biodiesel
development from rice bran oil: Transesterification process optimization and fuel characterization." Energy Conversion and Management 49 (5): 1248-1257.
260 6 Bibliography
Sivalakshmi, S and T Balusamy. 2012. "Influence of ethanol addition on a diesel engine
fuelled with neem oil methyl ester." International Journal of Green Energy 9 (3): 218-228.
Song, JT and CH Zhang. 2008. "An experimental study on the performance and exhaust
emissions of a diesel engine fuelled with soybean oil methyl ester." Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering 222 (12): 2487-2496.
Song, Juhun, Vince Zello, André L Boehman and Francis J Waller. 2004. "Comparison
of the impact of intake oxygen enrichment and fuel oxygenation on diesel combustion and emissions." Energy & Fuels 18 (5): 1282-1290.
Soriano Jr, Nestor U and Akash Narani. 2012. "Evaluation of biodiesel derived from
Camelina sativa oil." Journal of the American Oil Chemists' Society 89 (5): 917-923.
Srinivasa Pai, P and BR Shrinivasa Rao. 2011. "Artificial Neural Network based
prediction of performance and emission characteristics of a variable compression ratio CI engine using WCO as a biodiesel at different injection timings." Applied Energy 88 (7): 2344-2354.
Stern, Nicholas. 2008. "The economics of climate change." The American Economic
Review: 1-37.
SUN, Yu-qiu, Bo-shui CHEN, Guang ZENG, Jiu WANG, Jian-hua FANG, Wei XIONG
and Yu-li SUN. 2008. "Research of Low Temperature Properties of Peanut Oil and Peanut Biodiesel [J]." Internal Combustion Engines 3: 018.
Surawski, N. C., B. Miljevic, G. A. Ayoko, S. Elbagir, S. Stevanovic, K. E. Fairfull-
Smith, S. E. Bottle and Z. D. Ristovski. 2011a. "Physicochemical characterization of particulate emissions from a compression ignition engine: The influence of biodiesel feedstock." Environmental Science and Technology 45 (24): 10337-10343.
Surawski, N. C., B. Miljevic, G. A. Ayoko, S. Elbagir, S. Stevanovic, K. E. Fairfull-
Smith, S. E. Bottle and Z. D. Ristovski. 2011b. "Physicochemical characterization of particulate emissions from a compression ignition engine: the influence of biodiesel feedstock." Environmental Science & Technology 45 (24): 10337-43.
Surawski, N. C., B. Miljevic, G. A. Ayoko, B. A. Roberts, S. Elbagir, K. E. Fairfull-
Smith, S. E. Bottle and Z. D. Ristovski. 2011. "Physicochemical characterization of particulate emissions from a compression ignition engine employing two
6 Bibliography 261
injection technologies and three fuels." Environmental science & technology 45 (13): 5498-505.
Surawski, Nicholas C, Branka Miljevic, Godwin A Ayoko, Sohair Elbagir, Svetlana
Stevanovic, Kathryn E Fairfull-Smith, SE Bottle and ZD Ristovski. 2011c. "Physicochemical characterization of particulate emissions from a compression ignition engine: the influence of biodiesel feedstock." Environmental science & technology 45 (24): 10337-10343.
Surawski, Nicholas C, Branka Miljevic, Boyd A Roberts, Robin L Modini, Rong Situ,
Richard J Brown, Steven E Bottle and Zoran D Ristovski. 2009. "Particle emissions, volatility, and toxicity from an ethanol fumigated compression ignition engine." Environmental science & technology 44 (1): 229-235.
Suryanarayanan, Saikishan, Vijay Manikandan Janakiraman, G Lakshmi Narayana Rao
and S Sampath. 2008. Comparative study of the performance and emission characteristics of biodiesels from different vegetable oils with diesel: SAE Technical Paper.
Szybist, James P, Stephen R Kirby and André L Boehman. 2005. "NO x emissions of
alternative diesel fuels: a comparative analysis of biodiesel and FT diesel." Energy & fuels 19 (4): 1484-1492.
Szybist, James P, Juhun Song, Mahabubul Alam and André L Boehman. 2007.
"Biodiesel combustion, emissions and emission control." Fuel Processing Technology 88 (7): 679-691.
Tamura, Shin'ichi and Masahiko Tateishi. 1997. "Capabilities of a four-layered
feedforward neural network: four layers versus three." Neural Networks, IEEE Transactions on 8 (2): 251-255.
Tan, Raymond R, Alvin B Culaba and Michael RI Purvis. 2004. "Carbon balance
implications of coconut biodiesel utilization in the Philippine automotive transport sector." Biomass and Bioenergy 26 (6): 579-585.
Tang, Haiying, Steven O Salley and KY Simon Ng. 2008. "Fuel properties and
precipitate formation at low temperature in soy-, cottonseed-, and poultry fat-based biodiesel blends." Fuel 87 (13): 3006-3017.
Taravus, Sevilay, Hakan Temur and Ahmet Yartasi. 2009. "Alkali-catalyzed biodiesel
production from mixtures of sunflower oil and beef tallow." Energy & Fuels 23 (8): 4112-4115.
Tasdemir, Sakir, Ismail Saritas, Murat Ciniviz and Novruz Allahverdi. 2011. "Artificial
neural network and fuzzy expert system comparison for prediction of
262 6 Bibliography
performance and emission parameters on a gasoline engine." Expert Systems with Applications 38 (11): 13912-13923.
Tat, ME and JH Van Gerpen. 2003a. "Biodiesel blend detection with a fuel composition
sensor." Applied Engineering in Agriculture 19 (2): 125-131.
Tat, ME and JH Van Gerpen. 2003b. "Measurement of biodiesel speed of sound and its
impact on injection timing." National Renewable Energy Laboratory, NREL/SR-510-31462.
Terry, Barbara, Robert L McCormick and Mani Natarajan. 2006. Impact of biodiesel
blends on fuel system component durability: SAE Technical Paper.
Todaka, Masatoshi, Wasana Kowhakul, Hiroshi Masamoto, Mikiji Shigematsu and Siaw
Onwona-Agyeman. 2013. "Thermal decomposition of biodiesel fuels produced from rapeseed, jatropha, and coffee oils with different alcohols." Journal of thermal analysis and calorimetry 113 (3): 1355-1361.
Usta, N. 2005. "An experimental study on performance and exhaust emissions of a
diesel engine fuelled with tobacco seed oil methyl ester." Energy Conversion and Management 46 (15): 2373-2386.
Utlu, Zafer and Mevlüt Süreyya Koçak. 2008. "The effect of biodiesel fuel obtained
from waste frying oil on direct injection diesel engine performance and exhaust emissions." Renewable Energy 33 (8): 1936-1941.
Van Gerpen, Jon H, Seref Soylu and ME Tat. 1999. "Evaluation of the lubricity of
soybean oil-based additives in diesel fuel." In Proceedings, Annual International Meeting of the ASAE, edited.
Varuvel, Edwin Geo, Nadia Mrad, Mohand Tazerout and Fethi Aloui. 2012.
"Experimental analysis of biofuel as an alternative fuel for diesel engines." Applied Energy 94 (0): 224-231.
Vedaraman, N, Sukumar Puhan, G Nagarajan and KC Velappan. 2011. "Preparation of
palm oil biodiesel and effect of various additives on NOx emission reduction in B20: An experimental study." International Journal of Green Energy 8 (3): 383-397.
Veny, Harumi, Saeid Baroutian, Mohamed Kheireddine Aroua, Masitah Hasan, Abdul
Aziz Raman and Nik Meriam Nik Sulaiman. 2009. "Density of Jatropha curcas seed oil and its methyl esters: measurement and estimations." International Journal of Thermophysics 30 (2): 529-541.
6 Bibliography 263
Vicente, Gemma, A Coteron, M Martinez and J Aracil. 1998. "Application of the factorial design of experiments and response surface methodology to optimize biodiesel production." Industrial crops and products 8 (1): 29-35.
Wadumesthrige, Kapila, Mahbuba Ara, Steven O Salley and KY Simon Ng. 2009.
"Investigation of lubricity characteristics of biodiesel in petroleum and synthetic fuel." Energy & Fuels 23 (4): 2229-2234.
Wang, Jianxin, Fujia Wu, Jianhua Xiao and Shijin Shuai. 2009. "Oxygenated blend
design and its effects on reducing diesel particulate emissions." Fuel 88 (10): 2037-2045.
WANG, Li-bing, Hai-yan YU, Xiao-hui HE and Rui-ying LIU. 2012. "Influence of fatty
acid composition of woody biodiesel plants on the fuel properties." Journal of Fuel Chemistry and Technology 40 (4): 397-404.
Wang, WG, DW Lyons, NN Clark, M Gautam and PM Norton. 2000. "Emissions from
nine heavy trucks fueled by diesel and biodiesel blend without engine modification." Environmental Science & Technology 34 (6): 933-939.
Weitzman, Martin L. 2007. "A Review of The Stern Review on the Economics of
Climate Change." Journal of Economic Literature 45 (3): 703-724.
Winayanuwattikun, Pakorn, Chutima Kaewpiboon, Kingkaew Piriyakananon, Supalak
Tantong, Weerasak Thakernkarnkit, Warawut Chulalaksananukul and Tikamporn Yongvanich. 2008. "Potential plant oil feedstock for lipase-catalyzed biodiesel production in Thailand." Biomass and Bioenergy 32 (12): 1279-1286.
Worapun, Ittipon, Kulachate Pianthong and Prachasanti Thaiyasuit. 2012. "Optimization
of biodiesel production from crude palm oil using ultrasonic irradiation assistance and response surface methodology." Journal of Chemical Technology and Biotechnology 87 (2): 189-197.
Wu, Chuanjie, Jinli Zhang, Wei Li, Yiping Wang and Hui Cao. 2006. "Artificial neural
network model to predict cold filter plugging point of blended diesel fuels." Fuel processing technology 87 (7): 585-590.
Wu, F., Wang, J., Chen, W., and Shuai, S. 2008. "Effects of Different Biodiesels and
their Blends with Oxygenated Additives on Emissions from a Diesel Engine." SAE Technical Paper 008-01-1812.
Wu, Fujia, Jianxin Wang, Wenmiao Chen and Shijin Shuai. 2009. "A study on emission
performance of a diesel engine fueled with five typical methyl ester biodiesels." Atmospheric Environment 43 (7): 1481-1485.
264 6 Bibliography
Wu, Xuan and Dennis YC Leung. 2011. "Optimization of biodiesel production from camelina oil using orthogonal experiment." Applied Energy 88 (11): 3615-3624.
Wyatt, Victor T, Melissa A Hess, Robert O Dunn, Thomas A Foglia, Michael J Haas
and William N Marmer. 2005. "Fuel properties and nitrogen oxide emission levels of biodiesel produced from animal fats." Journal of the American Oil Chemists' Society 82 (8): 585-591.
Xue, J., T. E. Grift and A. C. Hansen. 2011. "Effect of biodiesel on engine performances
and emissions." Renewable and Sustainable Energy Reviews 15 (2): 1098-1116.
Yamane, Koji, Atsushi Ueta and Yuzuru Shimamoto. 2001. "Influence of physical and
chemical properties of biodiesel fuels on injection, combustin and exhaust emission characteristics in a direct injection compression ignition engine." International Journal of Engine Research 2 (4): 249-261.
Yang, Hong, Zbigniew Ring, Yevgenia Briker, Norma McLean, Wally Friesen and
Craig Fairbridge. 2002. "Neural network prediction of cetane number and density of diesel fuel from its chemical composition determined by LC and GC–MS." Fuel 81 (1): 65-74.
Yap, Wai Kean and Vishy Karri. 2012. "Emissions predictive modelling by
investigating various neural network models." Expert Systems with Applications 39 (3): 2421-2426.
Young, Li-Hao, Yi-Jyun Liou, Man-Ting Cheng, Jau-Huai Lu, Hsi-Hsien Yang, Ying I
Tsai, Lin-Chi Wang, Chung-Bang Chen and Jim-Shoung Lai. 2012. "Effects of biodiesel, engine load and diesel particulate filter on nonvolatile particle number size distributions in heavy-duty diesel engine exhaust." Journal of hazardous materials 199: 282-289.
Yuan, W., A. C. Hansen and Q. Zhang. 2009. "Predicting the temperature dependent
viscosity of biodiesel fuels." Fuel 88 (6): 1120-1126.
Yusaf, Talal F, DR Buttsworth, Khalid H Saleh and BF Yousif. 2010. "CNG-diesel
engine performance and exhaust emission analysis with the aid of artificial neural network." Applied Energy 87 (5): 1661-1669.
Yusaf, TF, BF Yousif and MM Elawad. 2011. "Crude palm oil fuel for diesel-engines:
Experimental and ANN simulation approaches." Energy 36 (8): 4871-4878.
Zendehboudi, Sohrab, Mohammad Ali Ahmadi, Lesley James and Ioannis Chatzis.
2012. "Prediction of condensate-to-gas ratio for retrograde gas condensate reservoirs using artificial neural network with particle swarm optimization." Energy & Fuels 26 (6): 3432-3447.
6 Bibliography 265
Zhang, Junhua and Lifeng Jiang. 2008a. "Acid-catalyzed esterification of Zanthoxylum
bungeanum seed oil with high free fatty acids for biodiesel production." Bioresource technology 99 (18): 8995-8998.
Zhang, Junhua and Lifeng Jiang. 2008b. "Acid-catalyzed esterification of Zanthoxylum
bungeanum seed oil with high free fatty acids for biodiesel production." Bioresource Technology 99 (18): 8995-8998.
Zhang, Yen, MA Dube, DDl McLean and M Kates. 2003. "Biodiesel production from
waste cooking oil: 1. Process design and technological assessment." Bioresource technology 89 (1): 1-16.
Zheng, Ming, Mwila C Mulenga, Graham T Reader, Meiping Wang, David SK Ting and
Jimi Tjong. 2008. "Biodiesel engine performance and emissions in low temperature combustion." Fuel 87 (6): 714-722.
Zhu, Ruijin, C. S. Cheung and Zuohua Huang. 2011. "Particulate Emission
Characteristics of a Compression Ignition Engine Fueled with Diesel–DMC Blends." Aerosol Science and Technology 45 (2): 137-147. doi: 10.1080/02786826.2010.526655.
266 Appendices
Appendices
APPENDIX A: MatLab code for ANN models training
%Md Jahirul Islam; BERF, QUT, Brisbane %date: August 18, 2014 %** Preparation of training data. clc, clear all; close all; load('C:\Users\n7325819\Desktop\Matlab Codes_Kajol\Data_raw.mat'); Data=Data_raw; % get the data set as you like from initial total data size % percentage of data set for training, testing & validation training_set=0.6 ; validating_set=0.1; testing_set=0.3; %(%) % call the data_Preparation.m file for split the data for training, testing & validation; And prepare the data %for NN traing DataVec = size(Data,2); No_of_Samples = length(Data); %SampleIndex = [1:No_of_Samples]'; % use this if random sample is not allowed SampleIndex = randperm(No_of_Samples)'; % Use this if random sampling is allowed. data_Preparation % call the data preparation m file cd([pwd,'\Results']); save Training_data % save the whole workspace % Prepare the data here for training, testing and validation. %Call the matrix 'Data' with size n*m. First m-1 columns are inputs and last column is the target.) %=================================================================================================== function [TrainSet ValdSet TestSet] = datasplit(A, ptr, pvd, pts) rows = max(size(A)); nVald = round(pvd * rows); % number of samples for validation nTest = round(pts * rows); % number of samples for test nTrain = rows - nVald - nTest; TrainSet = A(1:nTrain,:); ValdSet = A(nTrain+1:nTrain+nVald,:); TestSet = A(nTrain+nVald+1:end,:); %Data=data_PI_dis; %========================================================================== [Index_TrainSet Index_ValdSet Index_TestSet] = datasplit(SampleIndex,training_set,validating_set,testing_set); % Split Data into Test, Validation, and Testing for j = 1:size(Index_TrainSet,1) % Index for Training Set k = Index_TrainSet(j);
Appendices 267
TrainSet(1:DataVec,j) = Data(k,1:DataVec); end for j = 1:size(Index_ValdSet,1) % Index for Validation Set k = Index_ValdSet(j); ValdSet(1:DataVec,j) = Data(k,1:DataVec); end for j = 1:size(Index_TestSet,1) % Index for Testing Set k = Index_TestSet(j); TestSet(1:DataVec,j) = Data(k,1:DataVec); end %=============================================================================================== Data=Data_raw; % get the data set as you like from initial total data size (10,998) %Data=data_NN_PI_ND; % percentage of data set for training, testing & validation training_set=0.7 ; validating_set=0.1; testing_set=0.2; %(%) % call the data_Preparation.m file for split the data for training, testing & validation; And prepare the data %for NN traing DataVec = size(Data,2); No_of_Samples = length(Data); %SampleIndex = [1:No_of_Samples]'; % use this if random sample is not %allowed SampleIndex = randperm(No_of_Samples)'; % Use this if random sampling is %allowed. data_Preparation %=============================================================================================== I = [1:DataVec-1]; % Inputs O = [DataVec]; % Output %=============================================================================================== % Training Set pTr = TrainSet(I,:); tTr = TrainSet(O,:); [pTr,pTrMin,pTrMax,tTr,tTrMin,tTrMax] = premnmx(pTr,tTr); % Preprocessing data with converting the range from -1 to 1. % Validation data pVd = ValdSet(I,:); tVd = ValdSet(O,:); pVd = tramnmx(pVd,pTrMin,pTrMax); tVd = tramnmx(tVd,tTrMin,tTrMax); % Test data pTs = TestSet(I,:); tTs = TestSet(O,:); pTs = tramnmx(pTs,pTrMin,pTrMax); tTs = tramnmx(tTs,tTrMin,tTrMax); MinMax = minmax([pTr pVd pTs]); %===================================================================================================
268 Appendices
Data=Data_raw; % get the data set as you like from initial total data size (10,998) %Data=data_NN_PI_ND; % percentage of data set for training, testing & validation training_set=0.6 ; validating_set=0.1; testing_set=0.3; %(%) % call the data_Preparation.m file for split the data for training, testing & validation; And prepare the data %for NN traing DataVec = size(Data,2); No_of_Samples = length(Data); %SampleIndex = [1:No_of_Samples]'; % use this if random sample is not %allowed SampleIndex = randperm(No_of_Samples)'; % Use this if random sampling is %allowed. data_Preparation % Training the NN to generate *** model % define the training, testing and validating data set %========================================================================= %clear all load('C:\Users\n7325819\Desktop\Matlab Codes_Kajol\Data_raw.mat'); val.P =pVd; val.T=tVd; % Define the validation data test.P = pTs; test.T=tTs; % Define the test data % define the network structure % pTr and tTr are for training data set %========================================================================= Ntotal=10; % total NN models Neur=[6*ones(1,Ntotal/10), 8*ones(1,Ntotal/10), 10*ones(1,Ntotal/10), 12*ones(1,Ntotal/10), 14*ones(1,Ntotal/10), 16*ones(1,Ntotal/10),18*ones(1,Ntotal/10),22*ones(1,Ntotal/10),24*ones(1,Ntotal/10),26*ones(1,Ntotal/10)]; % neuron size in hidden layer MAPE=[]; RMSE=[]; SSE=[]; for netopt=1:Ntotal clear Net1 an MAPE1 RMSE1 SSE1 neur=Neur(netopt) mout=1; % the number of moel output Net1=newff(minmax(pTr),[neur mout],{ 'tansig' 'purelin'}, 'trainbr'); % tansig= tan sigmoid transfer function for hidden neuron input % purelin=linear output function for hidden neuron % Set training parameters % Net.trainParam.epochs=500; % Maximum number of epochs Net1.trainParam.show=100; % Period of showing calculation progress Net1.trainParam.lr=0.1; % Algorithm learning rate Net1.trainParam.goal=.01; % Optimisation goal Net1.trainParam.min_grad=1e-10; % Minimum gradient Net1.trainParam.mem_reduc=1; % Memory reduction parameter Net1.trainParam.max_fail=1000; time0 = cputime; % Use the 'train' command to start the training process. The trained % network will be saved in the structure Net. pTr and tTr are input and
Appendices 269
% targets for the training data set. val and test are the validation and testing data sets respectively. %Net = train(Net,pTr,tTr); [Net1,tr]=train(Net1,[pTr,pVd,pTs],[tTr,tVd,tTs],[],[], val,test); %========================================================================= % Simulate the network with the testing (normalized) data. an = sim(Net1,pTs); % Un-normalize the network prediction data % convert the data as real values a = postmnmx(an,tTrMin,tTrMax); Output=postmnmx(tTs,tTrMin,tTrMax); close all figure(3); plot(a,Output);ylabel('Actual'); xlabel('Predicted'); figure(4); plot(Output); hold; plot(a,'m'); legend('Actual','Predicted'); MAPE1 = mean(abs((tTs-an)./tTs))*100 RMSE1=sqrt((sum((tTs-an).^2))/size(tTs,2)); SSE1=sum((tTs-an).^2); MAPE=[MAPE,MAPE1]; RMSE=[RMSE,RMSE1]; SSE=[SSE,SSE1]; NNtrain(netopt)=struct('NNmodel',Net1, 'trecord', tr, 'MAPE',MAPE1,'RMSE1',RMSE,'SSE1',SSE,'actual',Output,'Predicted',a); end %========================================================================== cd([pwd,'\Results']); save NNtrain %==========================================================================
270 Appendices
APPENDIX B: The eigenvalue for each of the PCs
CN KV Density HHV OS CFPP FP IV
PC 1 13.8644 13.3400 10.0326 12.0796 11.5437 12.0405 11.1067 13.3676
PC 2 4.8300 5.5200 6.7413 5.3383 6.7804 6.0536 5.7753 6.0168
PC 3 1.8400 2.7050 1.7298 2.2664 2.1142 2.0142 2.5236 1.5142
PC 4 0.8820 1.0232 1.8750 1.7064 1.5755 1.3542 1.7066 1.1954
PC 5 0.6046 0.2555 1.8756 1.1426 0.7002 0.7764 0.5986 0.3072
PC 6 0.4692 0.1122 0.4364 0.2334 0.1409 0.4972 0.3698 0.1960
PC 7 0.0951 0.0311 0.1827 0.1396 0.0843 0.2662 0.2962 0.1509
PC 8 0.0962 0.0226 0.0738 0.0564 0.0341 0.0267 0.3302 0.1750
PC 9 0.0759 0.0149 0.0157 0.0120 0.0072 0.0057 0.0701 0.0372
PC 10 0.0710 0.0110 0.0115 0.0088 0.0053 0.0042 0.0516 0.0273
PC 11 0.0643 0.0102 0.0107 0.0082 0.0050 0.0039 0.0480 0.0255
PC 12 0.0516 0.0071 0.0075 0.0057 0.0035 0.0027 0.0335 0.0178
PC 13 0.0365 0.0045 0.0047 0.0036 0.0022 0.0023 0.0287 0.0152
PC 14 0.0132 0.0003 0.0003 0.0003 0.0002 0.0012 0.0147 0.0078
PC 15 0.0059 0.0025 0.0026 0.0020 0.0012 0.0009 0.0117 0.0062
PC 16 0.0011 0.0011 0.0022 0.0017 0.0010 0.0008 0.0098 0.0052
PC 17 0.0005 0.0005 0.0016 0.0012 0.0007 0.0006 0.0078 0.0041
PC 18 0.0003 0.0003 0.0015 0.0012 0.0007 0.0005 0.0075 0.0040
PC 19 0.0002 0.0002 0.0011 0.0009 0.0005 0.0004 0.0055 0.0029
PC 20 0.0002 0.0002 0.0008 0.0006 0.0003 0.0003 0.0040 0.0021
PC 21 0.0001 0.0001 0.0005 0.0001 0.0000 0.0000 0.0000 0.0000
PC 22 0.0000 0.0000 0.0002 0.0000 0.0000 0.0000 0.0000 0.0000
PC 23 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000