Khan K. Biodiesel Kinetics & Catalyst 2 (Master)

RESEARCH INTO BIODIESEL CATALYST SCREENING & DEVELOPMENT
By Adam KHAN
A thesis submitted to the Department of Chemical Engineering In partial fulfilment of the requirements For an Individual Inquiry Topic at The University of Queensland Brisbane, Queensland, Australia 18 October 2002
Adam Khan
Individual Inquiry B
ACKNOWLEDGEMENTS
I would like to thank my supervisor, Assoc Prof Victor Rudolph for his constant support and much needed direction, also his perseverance with my some of my wild ideas and running hypotheses, some of which where a little pre-emptive. From this and previous work I am beginning to appreciate how a structured, directed and enthusiastic approach to projects such as this has many benefits, including success. Although, in saying that there are many walls encountered in research and without perseverance and lateral thinking some of them would never be overcome. I would also like to thank Dr Yinghe He for his help throughout this project, although my Chinese lessons seem to have come to an end. I would also like to thank Ms Allison Hanley for her assistance and input on the project and the many offers of help and her laboratory skills which saved me much needed time on several occasions. Finally I would like to thank Fleur Khan and my daughters Whitney and Isobel for their eternal support and understanding of my goals and aspirations. It is from them that I have drawn all of my determination and perseverance as there have been many occasions where I have questioned my direction and purpose. Without them I would not have been able to complete much of what I have done and become who I am.
-2-
Adam Khan
ABSTRACT
This report follows on from previous research that attempted to cover the topic of biodiesel production with a broad stroke, highlighting some key areas of interest, of which catalyst development is still dominate. Catalyst development is still the Achilles heel of continuous biodiesel production. This combined with the use of crude feedstocks is one of the most important areas of research for the development of an economically viable and sustainable energy source. The use of crude feedstocks achieves two objectives, firstly it utilises a waste source that would otherwise have to be disposed of and secondly it reduces the pressure on the environment with the use of renewable resources and lower emissions of toxic pollutants. However, crude feedstocks have the disadvantage of having high concentrations of free fatty acids and or water which current catalyst will not tolerate. It was found that the water and free fatty acids are major contributors to delivering insignificant yields. Measuring these yields and the extent of reaction was resolved in previous research. Although gas chromatography is to date the most accurate method to assess the extent of reaction, it is time consuming and requires extensive sample preparation. It was due to these factors that an alternative method was sought to grossly quantify the extent of reaction. A simple mass balance was used to for this purpose which was conducted thorough the separation and weighing of the individual phases. From this actual mass and a theoretical mass, the method could be cross checked for validity. Several heterogenous catalysts were tested and have proven to be unsuitable with the exception of one, synthesised zirconium. This catalyst and an acidified version will undergo further testing as at this stage they have shown promise, but have not been extensively tested. The use of sodium hydroxide gave some interesting results including the fact that it is only marginally soluble in the reaction mixture. This has lead to some interesting ideas including the use of Sodium hydroxide on a support as a solid catalyst. Although it is only marginally soluble it still seems to catalyse the reaction. Emulsion formation proved to be a hindrance to this research and identifying the underlying causses and preventing them would lead to more accurate results. Preliminary tests have been conducted and some hypotheses have been established on some potential factors, but these are not definitive.
-3-
Adam Khan
TABLE OF CONTENTS
1.0 1.1 1.2 2.0 2.1 2.2 2.3 2.4 3.0 4.0 4.1 5.0 6.0 6.1 INTRODUCTION........................................................................................... 6 OBJECTIVES ................................................................................................... 6 IMPLIMENTATION........................................................................................ 6 LITERATURE REVIEW .............................................................................. 7 GENERAL........................................................................................................ 7 Washing ............................................................................................................ 7 WATER and FREE FATTY ACIDS................................................................ 8 HETEROGENOUS CATALYSTS .................................................................. 8 LITERATURE REVIEW CONCLUSION .................................................. 8 SCOPE ............................................................................................................. 9 PROPOSED OUTCOMES ............................................................................... 9 INTRODUCTION........................................................................................... 9 EXPERIMENTAL........................................................................................ 10 REACTION METHODOLOGY .................................................................... 10 Equipment ............................................................................................... 10 Reactants ................................................................................................. 12 Conditions ............................................................................................... 12 Mass Balance Method............................................................................. 13 6.1.1 6.1.2 6.1.3 6.2 6.3 6.4 6.5 6.2.1
ANALYSIS METHODOLOGY..................................................................... 13 ANALYSIS ERROR ...................................................................................... 14 REACTIONS .................................................................................................. 14 High Temperature Conversion................................................................ 15 High Temperature Conversion................................................................ 16 Traditional Catalysts ............................................................................... 17 Potential Heterogenous Catalyst ............................................................. 17 Emulsion Formation................................................................................ 17 Gravimetric Method................................................................................ 17 RESULTS ....................................................................................................... 16
6.4.2 6.5.1 6.5.2 6.5.3 6.5.4 6.5.5 7.0 7.1
DISCUSSION ................................................................................................ 17 Catalysts.......................................................................................................... 18 Sodium hydroxide................................................................................... 18 Zirconium................................................................................................ 18 Zirconia Sand .......................................................................................... 18 Titanium.................................................................................................. 18 7.1.1 7.1.2 7.1.3 7.1.4
7.2
Gravimetric Method........................................................................................ 18
-4-
Adam Khan 7.3 7.4 8.0 9.0 10.0
Emulsion formation ........................................................................................ 19 High Temperature reactions............................................................................ 19 CONCLUSIONS ........................................................................................... 20 FUTURE WORK AND RESEARCH ......................................................... 20 REFERENCES.............................................................................................. 22
APPENDIX A ................................................................................................................ 25 APPENDIX B ................................................................................................................ 28 APPENDIX C ................................................................................................................ 41
TABLE OF FIGURES
Figure 6.1. Parr reactor and PID controller.................................................................. 10 Figure 6.2. Batch reactor with thermocouple and heating mantle.............................. 11 Figure 6.3 No conversion Bio0044 .................................................................................. 15 Figure 6.4 Bio0040, Bio0041 & Bio0041 with water ..................................................... 15 Figure 6.5 Formation of emulsion .................................................................................. 16 Figure 6.6 Progressive series of emulsion breaking ...................................................... 16
TABLE OF TABLES
Table 6.1. List of chemicals, supplier, quantity and grade .......................................... 12 Table 6.2. List of catalysts, supplier, quantity and grade ............................................ 12 Table 6.3. Typical reaction data sheet............................................................................ 13 Table 6.4. Mass balance procedure Bio0054.................................................................. 14 Table 11.1 Mass balance Bio0046 ................................................................................... 25 Table 11.2 Mass balance Bio0047 ................................................................................... 25 Table 11.3 Mass balance Bio0048 ................................................................................... 25 Table 11.4 Mass balance Bio0049 ................................................................................... 26 Table 11.5 Mass balance Bio0051 ................................................................................... 26 Table 11.6 Mass balance Bio0053 ................................................................................... 27
-5-
Adam Khan
1.0 INTRODUCTION
1.1 OBJECTIVES
This project follows on from previous work into biodiesel and is directed at towards the kinetics and catalyst development. It was found that catalyst screening trials was required to identify suitable heterogenous candidates for the transesterification reaction. Although, a potential catalyst was previously identified from the literature for this reaction, it was found that the catalyst properties i.e. activity deviated substantially from the results reported in the literature, which could not be replicated. Some potential catalysts such as titanium oxides and various forms of zirconia have been identified for performance testing. The use of crude feedstocks is important in the overall economics for the production of biodiesel, so catalysts will be tested with tallow and used oils. However, for simplicity and to minimise variables, food grade oils will be used to generate kinetic data and initially determine catalyst suitability. The development of a laboratory technique to estimate conversion will be assessed which includes the separation of phases, a mass balance and a final washing of the ethyl esters with reverse osmosis water. Catalysing the reaction is by far the most crucial constraint in the development of a continuous process for the production of biodiesel and as such the basis of this work has focused on identifying and testing for a suitable catalyst.
1.2 IMPLIMENTATION
The development of a laboratory method for estimating conversion preceded reactions in an attempt to decrease the turnaround time. Although, gas chromatography is currently the most accurate method to determine product concentrations, it is also quite time consuming and requires several hours sample preparation for each run. Including this will be an attempt to confliction information on some problems associated with washing biodiesel. As an additional activity the plug flow tubular reactor was assembled an pre-commissioning and pressure testing carried out.
-6-
Adam Khan
2.0 LITERATURE REVIEW

2.1 GENERAL
A more comprehensive survey of the biodiesel literature is given in Appendix C. This review focuses specifically on issues of significant practical importance poorly identified and generally not discussed to any substantial degree in the literature. The use of homogeneous alkali and to some extent acid catalysts, in the transesterification reaction of animal fats and oils, is a well-established practice. Alkali catalysts are generally used due to their relative simplicity, cost and the high levels of conversion achieved at low temperatures. This has thus lead to extensive research being carried out on their properties, outlining their advantages and disadvantages. The main alkali catalysts used are sodium hydroxide, sodium ethoxide (methoxide) potassium hydroxide and potassium ethoxide (methoxide). However, it was found that sodium hydroxide is only sparingly soluble in ethanol and was also found to promote undesirable gel and emulsion formation during reaction (Peterson, 1995). Methanol is usually used in the production of biodiesel as it is more cost effective, so there has been substantial research conducted on the methanol, sodium hydroxide combination. Recognising methanol as relatively toxic, its use has been more restricted when eventually environmental considerations are given greater weight. Irrelevant of the actual catalyst used, all homogeneous catalysts require neutralisation at the completion of reaction. This then raises issues of catalyst consumption and separation of the salt produced and the product phases. It is known that the glycerol, which is produced during reaction, tends to carry most of the catalyst with it when it is allowed to settle out (Peterson, 1995). This then limits the amount of catalyst remaining in the product phase but does not eliminate the problem.
2.2 Washing
The purpose of washing is to purify the biodiesel by removing catalyst, residual glycerol and unreacted or partially reacted oils. Peterson found that if the reaction mixture was mixed for around 20 minutes with pure glycerol and then 15% by weight of water was added with a further 20 minutes mixing, washing became a lot simpler and required less total washes. This method also tended to remove all of the mono and diglycerides from the product phase. However, it was found that vigorous stirring, before all of the contaminates had been removed with glycerol, caused emulsions to form that where difficult to break and when any soaps where still present the formation of gels or a high viscosity liquid was likely (Freedman, 1984). Fanguri also found that adding water in the presence of glycerol formed and emulsion resulting in the loss in yield of both glycerol and esters. Incomplete reactions and or reverse reactions where thought to be the cause of formation of solids or a gel like phase, due to the emulsion formation characteristics of mono-, di- or triglycerides (Praven, 1996). In the transesterification reaction the forward reaction is only favoured when an excess of alcohol is used, however it was found that ratios greater than 6:1, which is generally considered to be the optimum ratio, inhibited the separation of glycerol (Schuchardt, 1997). When a large quantity of alcohol is present it has been found that the glycerol phase favours the alcohol phase and will not settle out (Schuchardt, 1997). Although this phenomenon is not readily reported, this must be a problem with esterified reactions or transesterification with an acid catalyst as it has been found that ratios below 30:1 alcohol to oil give poor conversions (Van Gerpen, 1999). Research is not as extensive for acid catalyses (esterification or acidolysis) as it is for alkali catalysis. This may be due to the reactions having to be carried out at high temperatures and subsequently high pressures to increase the reaction rate. Mineral acids such as hydrogen -7-
Adam Khan
chloride or sulphuric acid are generally used as they are hydrophilic and considering that water is a major problem in this reaction, this is an advantageous trait.
2.3 WATER and FREE FATTY ACIDS

One common topic that occurs throughout most literature is the affect that water and free fatty acids (FFA) have on both esterification and transesterification reactions. The fatty acids are problem because they form soap with the metal ion from the alkali catalysts, which hinder separation, decrease yield and lead to emulsion formation. Water has a two fold affect it combined with soap form the emulsions and secondly the oils can undergo hydrolysis with the water forming fatty acids. Ma found that water was a more critical variable than was FFA, but more importantly when both present, they acted synergistically to substantially decrease the yield of esters (Ma, 1998). Water is detrimental to the reaction from an equilibrium standpoint as it is a product in the reaction and will favour the reverse reaction as the concentration increases. Refined oils are thus generally favoured due to their low concentrations of FFA and water. The source of water is not limited to that found in the feedstock, as some water is produced with the reaction of metal hydroxides with alcohol i.e. sodium ethoxide from sodium hydroxide and ethanol (Schuchardt, 1997). If there is sufficient alkoxide produced then hydrolysis occurs producing FFA and with a metal hydroxide present saponification occurs, which not only decreases the yield of esters but also consumes catalyst possibly leading to incomplete reaction.
2.4 HETEROGENOUS CATALYSTS

The limitations of alkali and acid catalysts are well documented, however there are few reports regarding catalysts. Catalyst development remains of primary importance in biodiesel research and unless more tolerant catalysts can be developed, feedstock quality will remain the main impediment. There have been brief mentions of solid catalysts throughout literature and in some situations reasonably detailed experimentation. Calcium carbonate has been said to be a satisfactory catalyst (Supps et al, 1996) but it is reported that alkali catalyst should be low in carbonates, such as potassium carbonate as carbonates do not serve as satisfactory catalysts (Peterson, 1996). Guanidines supported on an organic polymer showed some promise but have been found to be subject to leaching, irreversible protonation and rapid deactivation (Schuchardt, 1997). Organic tins and titanium are applied in the polymer industry for transesterification but the combination of toxicity and separation issues make them unsuitable for biodiesel production.
3.0 LITERATURE REVIEW CONCLUSION

All of the most significant issues with biodiesel production have a common thread the inclusion of water. Water leads to the production of FFA and the subsequent production of soap. Thus soap increases the complexity of the separation plant very significantly, besides its obvious detrimental effect as a by-product competing with biodiesel. Water does not need to be present in the feed if alkali catalysts are used as it evolves with the production of alkoxides albeit in small amounts. This does not occur with the use of acid catalysts, but the esterification of FFA produces water which can cause the subsequent production of more FFA through hydrolysis. The production of FFA in most industrial applications seems to be lower due to the use of refined oils. If however crude feedstocks are used this soon becomes a significant problem. Of the few catalysts that have been mentioned throughout literature, the transition metals may represent the most promising in particular titanium and zirconium. Both of these catalysts have been used in industry for similar reactions for some time and have proven to be effective. As has been discussed, catalyst development is integral to the development of a viable method of biodiesel production and thus it is here that this research will directed its focus. The most obvious start in catalyst screening will be with different forms of titanium and zirconium.
-8-
Adam Khan
4.0 SCOPE
4.1 PROPOSED OUTCOMES
Although gas chromatography was chosen to be the most suitable method of analysis for this project due to the complexity of the product mixture, it proved to be time consuming as it required extensive sample preparation. In addition the lack of a dedicated facility caused difficulties associated with access, timing and calibration. NIR was identified as being a potentially suitable method for online qualitative analysis, but due to equipment requirements it is not yet available. Therefore it was proposed that a suitable method of analysis be develop that could be carried out in the laboratory that was simple, reliable and quick. It was proposed that this method would take the form of a mass balance analysis requiring a series of separation steps. This would represent an intermediate step between preliminary analysis and detailed analysis via gas chromatography. This would be combined with catalyst screening with four potential catalysts being targeted. They have been selected on the basis of their composition not necessarily their structure. These are; 1. 2. 3. 4. Rutile Synthesised micro porous Zirconia Synthesised micro porous Zirconia Sulphate (Super Acid) Zircon Sand
There were also several runs conducted using Agro Lime (CaCO3) replicating previous trials to try and establish why the catalytic activity reported in the literature for limes could not be reproduced.
5.0 INTRODUCTION
This thesis is part two of a two part project that has been designed to identify some important aspects of biodiesel production, research and development. Part one Research into Biodiesel Kinetics & Catalyst Development attempted to identify the kinetics of the transesterification reaction with a suitable heterogenous catalyst, namely calcium carbonate. This part also covered the development of a suitable analysis technique namely gas chromatography and an in depth literature overview of biodiesel. Unfortunately the use of calcium carbonate as a transesterification catalyst could not be confirmed, despite several attempts to exactly replicate the reaction conditions as provided in the literature. However the analysis method which was developed based on GC proved to be effective. Since the ultimate objective of the project is to establish a continuous process, the physical attributes of the catalyst need to be such that it can be used within a high temperature high pressure plug flow reactor. Therefore catalyst screening was the primary goal of this work being targeted for suitability in the plug flow tubular reactor PFTR. The PFTR was designed and developed concurrently with the catalyst research. When the PFTR construction was completed it then required pre-commissioning and further development for optimal and safe operation. The second part of the research was conducted over a semester long period which proved insufficient to complete many of the activities which would have proved a more complete result. Nevertheless having a well structured and detail plan of attack has permitted important aspect of the work to be addressed.
-9-
Adam Khan
6.0 EXPERIMENTAL
6.1 REACTION METHODOLOGY 6.1.1 Equipment 6.1.1.1 Pressure Reactor (Batch)
The reactor was a Parr 2100 series 300ml pressure reactor. It is rated to 300bar, but is installed with a 1981.45 psi rupture disk and for operation up to 100C without cooling and 500C with cooling. It has a programmable PID controller that can be seen in Figure 6.1.
Figure 6.1. Parr reactor and PID controller
The procedure taken to use the reactor was as follows; 1. 2. 3. 4. The initial condition of the reactor was empty and clean. The reactants were weighed out in the required amounts added to a beaker with a magnetic stirrer and mixed for several minutes. The reactor was then charged with the contents of the beaker, the manifold placed on and the bolts tightened firmly. All outlets were checked to be closed; the reactor was then placed in the heating mantle, the stirrer turned on and the controller started.
6.1.1.2
Non-Pressure Reactor (Modified Beaker)
- 10 -
Adam Khan
The second reactor that was used was a modified beaker fitted with a thermocouple, on a heating mantle with a magnetic stirrer, as can be seen in Figure 6.2. The beaker was sealed with a rubber stopper to minimise the loss of ethanol in the vapour phase. The reaction temperature was kept below the boiling point of ethanol to maintain a constant pressure within the beaker. Since the internal beaker pressure was slightly less than atmospheric, the tendency for leakage was inward. This caused the stopper to be drawn inwards, providing an effective seal, demonstrated by a small quantity of liquid as a bead around the outside edge.
Figure 6.2. Batch reactor with thermocouple and heating mantle.
The procedure taken to use the bench top reactor was as follows; 1. 2. 3. 4. The initial condition of the reactor was empty and clean. The reactants were weighed out in the required amounts and added to the beaker with a magnetic stirrer. The rubber stopper was place on the top of the beaker and pressed down until neatly seated with the rim. The thermocouple was left out until the beaker reached 65C, then the thermo couple was inserted and the temperature was reduced to 60C. What this accomplished was to form a small vacuum within the beaker which kept the rubber stopper firmly in place and minimised the loss of ethanol vapour. The reaction temperature was then maintained for the duration of the reaction. Once the reaction time was complete, the thermocouple was removed and the heating mantle was turned off. It is important
5. 6.
- 11 -
Adam Khan
Individual Inquiry B that the thermocouple is removed before the temperature is decreased or the rubber stopper will be sucked into the beaker.
6.1.2 Reactants 6.1.2.1 Chemicals

The reactants used are listed in Table 6.1. Table 6.1. List of chemicals, supplier, quantity and grade
Chemical Grade Calcium Carbonate AR 99% Potassium Hydroxide AR 85% Sodium Hydroxide AR 85% Anhydrous Ethanol AR 99% Sulphuric acid AR 98.2% Canola Oil Edible Beef Tallow Prime Triglycerides 100% FFA 44% Save All Triglycerides 56% Source Univar Chem Supply Pty Ltd Chem Supply Pty Ltd University Chemical Store source unknown Coles Pty Ltd Beef City Quantity 500g 500g 500g 1.5L 500ml 4L 60kg 30kg 30kg
6.1.2.2
Catalysts
The catalysts used that have not been mentioned in Table 6.1 can be seen in Table 6.2. Table 6.2. List of catalysts, supplier, quantity and grade
Catalyst TiO2 R3412A Zircon Sand Z3094 ZrO ZrSO4 zr-NH-T110 DML Lime Source Consolidated Rutile Ltd Consolidated Rutile Ltd Nanomac Nanomac David Mitchell P. L Quantity 500g 500g ~5g ~5L 20kg Grade 95% TiO2 5% unknown 60% ZrO2 ~20%ZrO ~20%Unknown 99.9% 99.9% 35% CaCO3 65% unknown Form Granular Granular <10-6m <10-6m Course Powder
6.1.3 Conditions
The experiments were carried out with the two different reactor vessels. Firstly there were nine experiments carried out in the pressure reactor, these were primarily conducted to run some tests with DML Lime, Rutile and Zircon sand. Then there was six experiments carried out in the beaker reactor with an alkali catalyst and one with an acid catalyst. There was a simple strategy for conducting the experiments. Each reaction in was repeated in triplicate in an attempt to asses the mass balance method of analysis. A typical reaction data sheet can be seen in Table 6.3 where the total mass and the reaction conditions change between the two different reaction vessels.
- 12 -
Adam Khan Table 6.3. Typical reaction data sheet

Date Batch # 07/08/02 Bio0048 Mass g Mol Reactants: Canola Oil 300.00 0.35 EtOH 97.82 2.12 KOH 3.98 0.04 Reaction Conditions: TEMPERATURE 60C PRESSURE 1 atm TIME 180 min Comments: 6:1 Mole ratio of ethanol to oil ,~3% catalyst by weight
Reactor Vessel Beaker Run Molecular Weight 847.40 46.05 98.07
6.2 ANALYSIS METHODOLOGY 6.2.1 Mass Balance Method

The mass balance method comprises of several stages of isolating phases, weighing and washing the final product to eliminate any un-reacted components. The underlying assumption of the process is that the reaction gone to completion and one hundred percent of the oil has been converted to biodiesel. This assumption can be justified when the reaction is left for an extended period of time far exceeding that required for normal reaction conditions. Typical reaction times were 3 hours. When considering that ninety five percent of the reaction is complete in the first 30min it is reasonable to assume that the reaction has gone to completion.
6.2.1.1
Reaction Procedure
Once the reaction was complete the rubber stopper and magnetic stirrer were removed and the reactants weighed. The beaker was then periodically weighed until the loss of mass was equal to the approximate mass of excess ethanol that was calculated theoretically. The contents of the beaker where then place in a separating funnel and allowed to settle overnight. The glycerol was then separated and weighed. The remaining contents of the separating funnel were also weighed and then returned to the funnel. Some of the later batches were added to beaker with pure glycerol allowed to mix for 20min and then returned to the funnel where the glycerol was removed. This procedure was only conducted on some of the later batches based on earlier experience that there were some complications with washing. This will be discussed in detail later. The procedure outline can be found in Table 6.4.
- 13 -
Adam Khan Table 6.4. Mass balance procedure Bio0054

a b c d e f g h i Component Beaker Reactants After Rxn After Evap Empty Beaker Glycerol Flask Glycerol Biodiesel* Biodiesel# Mass (g) 169.1 396.5 566.0 523.1 171.8 116.4 146.7 319.4 315.7 Adjustment Actual
Theoretical Difference 0 0 0.4 3.62 2.68 -2.305 3.11 -0.62 -2.3
c-b-a d-b-a e-a g-f
42.4 171.75 30.3 319.42 315.7 394.2
46.05 169.07 32.605 316.31 316.31 396.5
1 2 3 4 5
Total Mass Lost In Overall Process 1. Mass of Excess Ethanol Evaporated 2. Mass of remaining components in empty beaker usually glycerol 3. Mass of actual glycerol compared with amount theoretically predicted 4. Mass of biodiesel before washing 5. Mass of biodiesel after washing
It was found that adding the biodiesel phase back into beaker with pure glycerol, mixing for 20min and then adding 15% by mass of water, enhanced the washing step and removed the impurities that seemed to be responsible for the formation of an emulsion. Emulsion formation was a major restriction in this process. They tended to be formed from reactions that were catalysed with alkalis and formed when the mixture was washed and agitated with water. The emulsions are hard to break once formed so it is best to limit their formation. To minimise there formation the later batches were not washed and the mass of the catalyst taken into consideration when weighing out the esters.
6.3 ANALYSIS ERROR

Since the gravimetric method is a novel feature of this work, was entirely developed here and noting that it requires a number of separation steps, it is appropriate to consider the errors that may occur. The main areas where error occurred were in mass losses between steps, such as glycerol sticking to the beaker when emptied into the separation funnel. Some of these where minimised as much as possible, but where taken into consideration when calculating the mass balance. The actual process was found to account for 99.42% of the overall mass, this seems very acceptable. The overall error is minimised by using the highest possible mass of reactants (obviously in their correct molar ratios). Overall error was relatively easy to asses, whereas the error associated to the differences based on theoretical masses and actual masses includes additional complexities. There are two unknowns in this process and they are the molecular mass of the oil and biodiesel, which are required for the balances but represent assumed values. There are several ways to calculate these values, however all of them are taken as averages and range between 847g/mol 963g/mol. Fro simplicity one value was chosen for oil based on an assumed molecular composition and used for the entire project. Since the molecular mass of biodiesel is directly dependant on the value of oil this was also therefore set for the duration of the project.
6.4 REACTIONS
The reactions were carried out in two reactors. Some were carried out in the Parr pressure reactor and the others were in a beaker. All of the reactions can be found in tables in Appendices A and B. Appendix A are the reactions carried out in the beaker that were used for the gravimetric
- 14 -
Adam Khan
method. There were some difficulties with soap and emulsion formation, with soap production leading to emulsification.
6.4.1.1
Pressure Reactor
A table of reactions can be found in Appendix B, which outlines the reaction conditions carried out in the pressure reactor, reactants and a one-line description on the results. A close examination of the samples in Figure 6.3 reveals a layer of ethanol across the top of each oil layer. These samples were taken from the pressure reactor with the Khan-Guinness method. The amount of ethanol varies across the range of samples, which is independent of the mass of oil. This can most readily be seen from the last samples on the far right of the diagram, in which the volume of ethanol is approximately 50% of the total volume. The final sample the larger vial is the residual of the reactor.
Figure 6.3 No conversion Bio0044
6.4.2 High Temperature Conversion

An anomaly found with some of the reactions carried out at high temperatures with and without catalysts can be seen in Figure 6.4. This is unusual in that the fats (solids at room temperature) have been converted to liquids at room temperature. What is also unusual is the absence of a glycerol phase.
Figure 6.4 Bio0040, Bio0041 & Bio0041 with water
- 15 -
Adam Khan
6.4.2.1
Beaker
The results for the gravimetric method can be found in Appendix A, these are Bio0046 to Bio0053. Some of the initial gravimetric methods involved washing of the final ester phase to remove any residual glycerol or unreacted/ partially reacted oil. What can be seen from Figure 6.5 (Bio0046) is the impact of this washing with the addition of rigorous agitation. The ensuing emulsion seemed to be unbreakable, but on the addition of table salt which is used in the soap industry, the emulsion dissipated to some extent. What can be seen in Figure 6.6 is the progressive addition of table salt starting from left to right, with the left most sample having no salt. The emulsion formation is a significant problem and will be discussed further in the results section.
Figure 6.5 Formation of emulsion Figure 6.6 Progressive series of emulsion breaking
6.5 RESULTS
Bio0044 represents a case where no reaction occurred. It nevertheless provides useful results in that all the samples taken, if representative of the reaction mixture, should have the same ethanol content. It is apparent from Figure 6.3 that this not the case, indicating that the sampling system does not provide a correctly and consistently representative sample of the pressure reactor. Switching the mixer off briefly while taking the sample may alleviate this problem. This process will have to be done over a short period of time to minimise any phase separation. There may be other options but at this stage further investigation is required.
6.5.1 High Temperature Conversion

It was initially thought that calcium carbonate was capable of catalysing the transesterification reaction. This was based on literature reports and also some early experiments which showed total conversion of a solid fat to a liquid. The only obvious difference between these experiments and previous ones was higher temperatures. What was also observed with these later samples was the absence of the glycerol layer. A possible explanation is that all of these experiments where carried out with around 12:1 ethanol to oil to compensate for the ethanol in the vapour phase. Under these conditions of excess ethanol, there are reports in the literature that the glycerol remains in solution in the ethanol. It was then thought that evaporating all of the ethanol from the sample should free up the glycerol. Therefore some of the samples were evaporated and some washed. It was found washing caused emulsion formation and in very few samples evaporation resulted in a small layer of glycerol forming. This was not consistent across all - 16 -
Adam Khan
samples, although there were some differences in the detail of the procedures followed and the work did not extend to determining the underlying reasons for the different behaviour.
6.5.2 Traditional Catalysts

One of the beaker reactions (Bio0048) was carried out with sodium hydroxide as the catalyst. What was interesting about this reaction was that it was still catalysed despite the fact that sodium hydroxide was obviously only marginally soluble in the reaction mixture. This poses several questions about how this reaction is catalysed industrially when ethanol is used as the alcohol. It was found that potassium hydroxide was readily soluble in the ethanol oil mixture and gave reasonable qualitative results.
6.5.3 Potential Heterogenous Catalyst

It was found that the Zirconium catalyst showed significant promise, however since only a small amount of catalyst was available, only one trial could be conducted. An acidified version of the Zirconium catalyst has been supplied for testing but at this stage GC analysis is not available so it will not be tested until quantitative analysis can be done. Zirconium catalyst is to date the only heterogenous catalyst that has showed any promise in catalysing the transesterification reaction.
6.5.4 Emulsion Formation

The formation of emulsions is a significant problem. It was found that all the reactions that were catalysed with alkali catalysts were more likely to form an emulsion if shaken with water. A typical example of the emulsions formed can be seen in figure 6.5. These emulsions were very hard to break and would not settle under their own accord. It was found that the acid catalysed reactions were not susceptible to emulsion formation although this is a preliminary finding due to the limited number of runs being conducted with an acid catalyst. The formation of soap may be a key factor as with acid catalysed reactions there is no soap formation.
6.5.5 Gravimetric Method

There was an interesting result with the mass balance method in that the amount of glycerol theoretically predicted was consistently half of what was actually found. This was with the reactions that were catalysed with an alkali catalyst. The reaction that was conducted with an acid catalyst gave almost the amount of glycerol that was theoretically predicted. There is still an unaccountable amount of mass missing from the ethyl esters and at this stage it is not clear if this can be found in the glycerol phase or there is an error in the procedure. All of the values for the mass balance method can be found in Appendix A. Pure glycerol is clear and immiscible with ethyl esters. The glycerol that is found in the biodiesel reactions is amber in colour and not quite as viscus as that found with pure glycerol. This will be discussed in detail in the discussion section. The glycerol that was found from the acid catalysed reaction was found to be a lot less viscus than that found from the alkali reactions. The quality of this will have to be analysed quantitatively to determine its purity. It is fairly conclusive that CaCO3 is not a suitable catalyst for this reaction, since it has shown on numerous occasions little to no activity.
7.0 DISCUSSION
Based on the overall objective to develop a heterogenous catalyst for a continuous reactor, the following speculations may present opportunities for exploration;
- 17 -
Adam Khan
7.1 Catalysts 7.1.1 Sodium hydroxide

Sodium hydroxide is relatively insoluble in ethanol, but nevertheless appears, even at very low concentrations, to be a strongly catalytic agent. One possibility is that the sodium hydroxide is reacted to the ethoxide form. This rases the possibility to use sodium hydroxide as a sacrificial solid catalyst in the continuous reactor. A simple test in which the catalyst was shaken with ethanol showed that it remained in the solid phase. This suggests that a reactor packed with sodium hydroxide pellets may be subject to blocking over time, as the catalyst recrystallises. One possibility to over come this would be to impregnate an inert porous carrier with sodium hydroxide solution, evaporate of the remaining water, and use this as the plug flow reactor packing.
7.1.2 Zirconium
If Zirconium proves to be as active as sodium hydroxide, it would be a more suitable catalyst, as it is not sacrificial. This would reduce the frequency of catalyst replacement and eliminate the need for post treatment. Post treatment in the form of catalyst neutralisation is required when alkali catalysts are used as the pH of the product is generally quite high. Currently the zirconium is supplied as a powder which is sub micron. This in its current form could not be used in a continuous reactor as it would induce an infinite pressure drop across the bed. A suitable support system will be required or particle size enlargement i.e. sintering to obtain a larger particle size. The trade off with this type of practice is the loss of surface area however in its current form, the catalyst is not suitable.
7.1.3 Zirconia Sand

As the preliminary results for the zirconium were good, it was thought finding a readily available suitable form of zirconium, could short cut the catalyst design to implementation stage. The most readily available form found was zirconia sand. The catalysts was tested repeatedly for activity but proved to be inactive and thus unsuitable for this purpose. Initial speculation about the form of the zirconia being unsuitable was confirmed but it the trial highlighted the importance of molecular structure.
7.1.4 Titanium
Titanium has previously been identified from literature as being a potential catalyst for this reaction. However the current application in a similar reaction is in a homogenous form. Although this is reported to have a high level of activity, it is not suitable for our application as a homogenous catalyst. Therefore as with zirconium, a heterogenous form was sought that could be used in a packed bed. Rutile proved to be readily available and sufficiently large enough to be employed in a packed bed. However, after several runs with it in the batch reactor it failed to show any signs of activity. Catalyses is complex my nature and understanding how catalysts work is by no means a simple task. Future work may show that the other crystal form of titanium, anatase may be more active, however rutile has proven to be unsuitable to catalyse this reaction.
7.2 Gravimetric Method

The gravimetric method gave some interesting results. However, there are still several issues to be resolved with this method. These include; 1. 2. An amount of glycerol consistently almost double what was theoretically predicted, The actual mass of ethyl esters is less than the theoretical amount, by approximately 50g. - 18 -
Adam Khan 3.
The composition of the glycerol that was produced with the acid catalyst seems to be a lot less viscous than that produced by the alkali catalysts, implying that significant impurities exist. The purity of this glycerol will be tested and compared to pure glycerol and glycerol produced with the alkali catalysts. This will be done with gas chromatography.
It still remains to be established the extent of water produced from the reaction of sodium hydroxide and ethanol. It would seem that the maximum amount possible would not account for the large amount of extra glycerol, typically 32g per 300g of oil. It is proposed that any discrepancies that have been found with gravimetric method can be clarified with GC analysis.
7.3 Emulsion formation

The occurrence of emulsion formation is hindrance to the production ethyl esters. It not only signifies the production of soap which decreases the yield of ethyl esters, but complicates the separation of the by-product phase, glycerol. A key factor in emulsion formation is the presence free fatty acids, combined with sodium hydroxide or similar mono basic alkali form soap. Although only vigorous agitation with water causes the more severe emulsions, gentle washing still produces a white milky gel at the interface between glycerol and ethyl esters. This gel makes measurement in the gravimetric method difficult and subject to a large error factor. This problem is not evident with the use of acid catalysts, which would esterify any fatty acids present or produced in the reaction. It is therefore necessary to conduct further investigations into the degree of fatty acid production in the base catalysed reaction and find ways to minimise their production. There is literature evidence suggesting that back mixing of glycerol before washing will minimise the formation of an emulsion by removing the impurities that cause it. This was briefly tested but requires further investigation. Monitoring the pH levels of the reaction and the final product would give an accurate indication of the magnitude of fatty acid production. This would be easily implemented and the equipment required would be minimal.
7.4 High Temperature reactions

The reactions that were conducted at high temperatures with fats, gave some unexpected results. Fats are solids at room temperature. However after reaction with ethanol at temperatures exceeding 250C it was found that the resulting product was a continuous phase liquid at room temperature. These reactions were found to be independent of catalyst. It was found that the oil did not undergo the same transformation. The difference between these two reactants is the possible presence of fatty acids. The fats could possible contain up to 15% by mass of fatty acids. This factor is driving the theory that the fats undergo hydrolysis, the water being supplied by the reaction of fatty acids and ethanol. Hydrolysis will occur without a catalyst at temperatures above 210C. However, some preliminary calculations suggest that there is an insufficient amount of water available to propagate the reaction. For every fatty acid there is one water produced (per mol) and every triglyceride molecule requires three water to create three fatty acids and glycerol. There is generally no glycerol phase with this reaction, where there should be if fatty acids are being produced. There is a possibility that the glycerol is dissolved in the ethanol phase which is in excess, but tests conducted to remove the ethanol failed to produce the required amount of glycerol .
- 19 -
Adam Khan
8.0 CONCLUSIONS
The conclusion is best summarised in point form and is as follows; 1. No effective solid catalysts have been reported in literature or patents. There is several in development but all have failed to perform. 2. Calcium carbonate is not effective, and literature reports could not be replicated. 3. Miroporous tetragonal zirconia shows good promise as a potential catalyst. 4. The sulphated zirconia (superacid) may also be a good candidate, but as yet has been untested. 5. Zircon sand showed no catalytic activity, as did rutile. 6. Titania based zeolite may be a promising material, and should be investigated further. 7. Alkali based catalysts e.g. supported sodium hydroxide may offer opportunity for use with refined oils, but suffer from soap formation when any fatty acids are present. 8. The gravimetric method has several finer details to be addressed, but could be a simple method to analysis the extent of reaction. 9. Emulsion formation presents serious practical problems of separation, but is not widely addressed in literature.
9.0 FUTURE WORK AND RESEARCH

Future work and research will include the systematic testing of sulphated zirconia, which will involve the acquisition of kinetic, data if it proves to be a suitable catalyst, for application in the continuous reactor. This will also include the development of a suitable support, as the particle size is currently unsuitable for a packed bed. Potential candidates could include but are not exclusive to aluminium, zeolite and polymers. As with acid catalysts it is proposed that zirconium catalysts will not induce emulsions in the product. The emulsion formation is an interesting occurrence but should only be resolved on completion of catalyst selection. It seems that the emulsions are strongly correlated with alkali catalysts which are significant only if an alkali catalyst is selected for this project. However, the gravimetric method has been severely hampered by emulsion formation as the catalysts used were alkalis. This may explain the losses in the gravimetric method. A short term solution to this problem may lie in the back mixing of pure glycerol to reduce the concentration of condiments to reduce or even eliminate emulsion formation. This was solution was briefly assessed but will require more testing over various sets of conditions to confirm its suitability. It is thought that the main contaminate responsible for the emulsion formation is fatty acids which leads to soap production. The transformation that fats undergo at high temperatures, in the presence of alcohol is still partially unknown. It is thought that the reaction is hydrolysis which is producing fatty acids which in the presence of alkali catalysts such as sodium hydroxide forms soap. This will need to be investigated further as it may restrict the maximum operating temperature of the process to reduce fatty acid production. However, this can only be confirmed on quantitative analyses using GC. The method of taking samples from the pressure reactor may need to be modified to ensure that a each sample is a good representation of the reaction mixture. The Khan-Guinness sampler is the - 20 -
Adam Khan
current method of taking samples and has been adequate up to date, but for future samples the method will need to be modified. This would be approached by conducting a systematic sampling regime checking the volume of ethanol across a range of sampling sizes to determine if the proportionality of ethanol to oil is constant. This would be conducted at temperatures just above the boiling point of ethanol and without a catalyst to reduce the reaction rate. Work is still required on the continuous reactor as full commissioning is not yet complete. Preliminary commissioning has already been started with modifications being made to facilitate temperature and pressure testing. There have been some additional modifications made for occupational health and safety reasons which were only minor. Intal pressure and temperature tests were found to be in line with the design specifications and operational testing with reactants can now proceed.
- 21 -
Adam Khan
10.0 REFERENCES
1. Provisional specification for biodiesel fuel (B100) blend stock for distillate fuels. 2. Alcantara, R. et al. Catalytic production of biodiesel from soy-bean oil, used frying oil and tallow. Biomass and Bioenergy 18, 515-527 (2000). 3. Ali, Y., Hanna, M. A. & Borg, J. E. Optimization of diesel, methyl tallowate and ethanol blend for reducing emissions from diesel engine. Bioresource Technology 52, 237-243 (1995). 4. Allen, C. A. W., Watts, K. C., Ackman, R. G. & Pegg, M. J. Predicting the viscosity of biodiesel fuels from their fatty acid ester composition. Fuel 78, 1319-1326 (1998). 5. Boocock, D. G. B., Konar, S. K., Mao, V. & Sidi, H. Fast one-phase oil-rich processes for the preparation of vegetable oil methyl esters. Biomass and Bioenergy 11, 43-50 (1995). 6. Boocock, D. G. B., Konar, S. K., Mao, V., Lee, C. & Buligan, S. Fast formation of highpurity methyl esters from vegetable oils. Journal of American Oil Chemical Society 75 (1998). 7. Boocock, D. G. B. (2001). 8. Boocock, D. G. B., Konar, S. K., Blakansky, G. V. & Zhou, W. in 51st Canadian Chemical Engineering Conference (2001). 9. Canakci, M. & Gerpen, V. J. Biodiesel production via acid catalysis. American Society of Agricultural Engineers 42, 1203-1210. 10. Corma, A., Miquel, S., Iborra, S. & Velty, A. 11. Cvengrosova, Z., Cvengros, J. & Hronec, M. Rapeseed oil ethyl esters as alternative fuels and their quality control. Petroleum and coal 39, 36-40 (1997). 12. Foglia, T. A., Jones, K. C., Haas, M. J. & Scott, K. M. Technologies supporting the adoption of biodiesel as an alternative fuel. The cotton gin and oil mill press (2000). 13. Freedman, B., Pryde, E. H. & Mounts, T. L. Variables affecting the yields of fatty esters from transesterified vegetable oils. Journal of american Oil Chemical Society (1984). 14. Freedman, B., Kwolek, W. F. & Pryde, E. H. Quantitation in the analysis of transesterified soybean oil by capillary gas chromatography. Journal of American Oil Chemical Society 63 (1986). 15. Graboski, M. S. & McCormick, R. L. Combustion of fat and vegetable oil derived fuels in diesel engines. Prog. Energy Combustion Sci. 24, 125-164 (1998). 16. Holcapek, M., Jandera, P., Fischer, J. & Prokes, B. Analytical monitoring of the production of biodiesel by high performance liquid chromatography with various detection methods. Journal of chromatography A, 858, 13-31 (1999). 17. Knothe, G. Rapid monitoring of transesterification and assessing biodiesel fuel quality by near-infrared spectroscopy using a fiber-optic probe. Journal of American Oil Chemical Society 76, 795-800 (1999). 18. Knothe. Monitoring a progressing transesterification reaction by fiber-optic color near infrared spectorscopy with correlation to 1h nuclear magnetic resonance spectroscopy. Journal of american Oil Chemical Society 77 (2000). 19. Korbitz, W. in World fuel ethanol congress (Beijing, China, 2001). 20. Kusdiana, D. & Saka, S. Kinetics of transesterification in rapeseed oil to biodiesel fuel as treated in supercritical methanol. Fuel 80, 693-698 (2000). 21. Ma, F., Clements, L. D. & Hanna, M. A. The effects of catalyst, free fatty acids, and water on transesterification of beef tallow. American Society of Agricultural Engineers 41, 12611264 (1998). 22. Ma, F. & Hanna, M. A. Biodiesel production: a review. Bioresource Technology 70, 1-15 (1999).
- 22 -
Adam Khan
23. Ma, F., Clements, L. D. & Hanna, M. A. The effect of mixing on transesterification of beef tallow. Bioresource Technology 69, 289-293 (1999). 24. Mittelbach, M. Diesel fuel derived from vegetable oils, V[1]: gas chromatographic determination of free glycerol in transesterified vegetable oils. Chromatographia 37 (1993). 25. Mittelbach, M. Diesel fuel derived from vegetable oils, VI: specifications and quality control of biodiesel. Bioresource Technology 56, 7-11 (1996). 26. Mittelbach, M., Roth, G. & Bergmann, A. Simultaneous gas chromatographic determination of methanol and free glycerol in biodiesel. Chromatographia 42 (1996). 27. Muniyappa, P. R., Brammer, S. C. & Noureddini, H. Improved conversion of plant oils and animal fats into biodiesel and co-product. Bioresource Technology 56, 19-24 (1996). 28. Peterson, C. L., Reece, D. L., Thompson, J. C., Beck, S. M. & Chase, C. Ethyl ester of rapeseed used as a biodiesel fuel - a case study. Biomass and Bioenergy 10, 331-336 (1995). 29. Peterson, C. L. et al. 1-8 (University of Idaho, College of Agriculture, Idaho, 1996). 30. Plank, C. & Lorbeer, E. Simultaneous determination of glycerol, and mono-, di- and triglycerides in vegetable oil methyl esters by capillary gas chromatography. Journal of chromatography A, 697, 461-468 (1995). 31. Resources, D. E. (2001). 32. Shimada, Y., Watanabe, Y., Sugihara, A. & Tominaga, Y. 33. Sonntag, N. O. V. Oleochemicals as renewable molecules; Prospects for the 21st century. Journal of american Oil Chemical Society 59, 795A (1982). 34. Suppes, G. J. et al. Calcium carbonate catalyzed alcoholysis of fats and oils., (2001). 35. Trathnigg, B. & Mittelbach, M. Analysis of triglyceride methanolysis mixtures using isocratic hplc with density detection. Journal of liquid chromatography 13, 95-105 (1990). 36. Van Gerpen, J. H. et al. 1-28 (Iowa State University, Iowa, 1996). 37. Ali, Y. & Hanna, M. A. (1994a). Physical properties of tallow ester and diesel fuel blends. Bioresourses Technology, 47, 131-4 38. Bondioli P, Gasparoli A, Lanzani A, Fedeli E, Veronese S, Sala M. Storagability of biodiesel. Journal of American Chemicals Society 1995: 72:699-702. 39. Freedman B, Pryde EH, Mounts TL. Variables affecting the yields of fatty esters from transesterified vegetable oils. Journal of American Oil Chemicals Society. 1984; 61:;163843. 40. Kemp W. In: Practical organic chemistry. London: McGraw-Hill, 1967. p 99. 41. Allen CAW, Watts KC. A batch type transesterification unit for biodiesel fuels, Technical paper 96-404, Canadian society of Agricultural Engineers, 1996. 42. Allen CAW, Watts KC, ACKMAN rg. Properties of methyl esters of interesterified triacylglycerols Proceeding of the Third Liquid Fuel Conference. Liquid fuels and industrial products from renewqable resources. St. Joseph, MI: ASAE, 1996 pp. 73-82. 43. Boocock DGB, Konar SK, Mao V, Lee C, Buligan S. JAOCS 1998;75:1167. 44. Deslandes N, Bellenger V, Jaffiol F, Verdu J. Applied Polymer Science 1998; 69:2663 45. Freedman, B., Butterfield, R.O., Pryde, E.H., 1986. Transesterification kinetics of soybean oil. JAOCS 63, 1375-1380. 46. Freedman, B., Pryde, E.H., Mounts, T.L., 1984. Variables affecting the yields of fatty esters from transesterified vegetable oils. JAOCS 61, 1638-1643. 47. Sonntag, N. O. V. (1982). Glycerolysis of fats and methyl esters status, review, and critique. JAOCS, 59 (10), 795A-802A. 48. Peterson, G. L. (1986). Vegetable oil as a diesel fuel: staus and research priorities. Trans. ASAE, 29 (5), 1413-22. 49. Hanna, M. A., Ali, Y., Cuppett, S. L., Zhang, D., 1996. Crystallization characterisics of methyl tallowate and its blend with ethanol and diesel fuel. JAOCS 73, 759-763. 50. Ma, F., Clements, L.D., Hanna, M.A., 1998. Biodiesel fuel from animal fat. Ancillary stusies on transesterification of beef tallow. Ind. Eng. Chem. Res. 37, 3768-3771. - 23 -
Adam Khan
51. X. Zhang, C. L. Peterson, D. Reece, G. Moller and R. Haws, Biodegradability of biodiesel in the aquatic environment. ASAE Paper No. 956742. ASAE, St. Joseph, MI 49085-9659 (1995). 52. Reference: list from Biodiesel production: a review. Fangrui Ma, Milford A. Hanna. 53. Smelser, N.J. (1987). Fats and Fatty Oils. In Ullmanns Encyclopaedia of Industrial Chemistry. Fifth ed: Weinhein. Vol. A10, pp. 173-243.
- 24 -
Adam Khan
APPENDIX A
Table 11.1 Mass balance Bio0046
Component a Beaker + Stirrer b Total c After Evap Mass (g) 176.75 578.73 Actual Theoretical Difference 0 0 0.4
282.35 Loss of product due to emulsion formation

Component a b c d e f g h Beaker Reactants After Rxn After Evap Empty Beaker Glycerol Flask Glycerol Biodiesel* Mass (g) 169.1 398.61 561.96 509 58.71 46.05 Actual Theoretical Difference 0 0 -5.75 -12.66 1 2 3 4 -10.24 5
306.07 i Biodiesel# 306.07 316.31 Total Mass Lost In Overall Process 1. Mass of Excess Ethanol Evaporated 2. Mass of remaining components in empty beaker usually glycerol 3. Mass of actual glycerol compared with amount theoretically predicted 4. Mass of biodiesel before washing 5. Mass of biodiesel after washing

Component a b c d e f g h Beaker Reactants After Rxn After Evap Empty Beaker Glycerol Flask Glycerol Biodiesel* Mass (g) 169.1 396.17 377.34 Actual Theoretical Difference 0 0 0.4 1 2 53.15 53.15 32.605 20.545 -36.08 3 4 5
280.23 i Biodiesel# 280.23 316.31 Total Mass Lost In Overall Process 1. Mass of Excess Ethanol Evaporated 2. Mass of remaining components in empty beaker usually glycerol 3. Mass of actual glycerol compared with amount theoretically predicted 4. Mass of biodiesel before washing 5. Mass of biodiesel after washing
- 25 -
Adam Khan Table 11.4 Mass balance Bio0049

Component a b c d e f g h Beaker Reactants After Rxn After Evap Empty Beaker Glycerol Flask Glycerol Biodiesel* Mass (g) 185.23 396.04 390.65 186.97 64.77 266.75 390.65 186.97 64.77 266.75 396.04 185.23 32.605 316.31 Actual Theoretical Difference 0 0
-5.39 1.74 32.165 -49.56 1 2 3 4 5
i Biodiesel# Total Mass Lost In Overall Process 1. Mass of Excess Ethanol Evaporated 2. Mass of remaining components in empty beaker usually glycerol 3. Mass of actual glycerol compared with amount theoretically predicted 4. Mass of biodiesel before washing 5. Mass of biodiesel after washing

Component a b c d e f g h Beaker Reactants After Rxn After Evap Empty Beaker Glycerol Flask Glycerol Biodiesel* Mass (g) 169.16 396.08 560.08 517.56 168.8 68.4 258.06 560.08 47.68 396.08 46.05 Actual Theoretical Difference 0 0 -5.16 -1.63 1 2 3 4 5
68.4 258.06
32.605 316.31
35.795 -58.25
i Biodiesel# Total Mass Lost In Overall Process 1. Mass of Excess Ethanol Evaporated 2. Mass of remaining components in empty beaker usually glycerol 3. Mass of actual glycerol compared with amount theoretically predicted 4. Mass of biodiesel before washing 5. Mass of biodiesel after washing
- 26 -
Adam Khan Table 11.6 Mass balance Bio0053

Component a b c d e f g h Beaker Reactants After Rxn After Evap Empty Beaker Glycerol Flask Glycerol Biodiesel* Mass (g) 169.06 396.08 562.99 521.82 173.54 96.94 84.64 319.4 393.93 -41.17 173.54 84.64 260.92 396.08 46.05 169.06 32.605 316.31 Actual Theoretical Difference
-2.15 -4.88 4.48 52.035 -55.39
1 2 3 4 5
315.7 i Biodiesel# Total Mass Lost In Overall Process 1. Mass of Excess Ethanol Evaporated 2. Mass of remaining components in empty beaker usually glycerol 3. Mass of actual glycerol compared with amount theoretically predicted 4. Mass of biodiesel before washing 5. Mass of biodiesel after washing
- 27 -
Adam Khan
APPENDIX B LITERATURE REVIEW

SCOPE
This literature survey discusses the benefits of using biodiesel as a suitable diesel substitute, whilst briefly touching on the history behind alternative compression engine fuels. It discusses some of the more current manufacturing techniques for the production of biodiesel as well as some new technology that is being researched. This is primarily focusing on developments in catalyst application, problems associated with different catalysts and the benefits of various catalyst systems. The area of downstream processing is briefly discussed, however separation is outside the focus of this work. Some critical analysis is discussed on the pros and cons of different systems and details of the analysis methods available and why and why not they are implemented in this work. Finally there is a discussion section and recommendation for further research section that highlights the areas in this research that are in need of further development.
INTRODUCTION
With the exception of hydroelectricity and nuclear energy, the majority of the worlds energy needs are supplied through petrochemical sources, coal and natural gas. All of these sources are finite and at current usage rates will be consumed by the end of the next century (Aksoy, 1990). The depletion of world petroleum reserves and increased environmental concerns has stimulated recent interest in alternative sources for petroleum-based fuels. Biodiesel has arisen as a potential candidate for a diesel substitute due to the similarities it has with petroleum-based diesel.
BACKGROUND Fats and Oils

The use of vegetable oils as alternative fuels has been around since 1900 when the inventor of the diesel engine Rudolph Diesel first tested them, in his compression engine (Foglia, Jones, Haas, & Scott, 2000). To date there have been many problems found with using vegetable oils directly in diesel engines (especially in direct injection engines). These include (Ma, 1999):
1.
2. 3. 4. 5.
Coking and trumpet formation on the injectors to such an extent that fuel atomisation does not occur properly or is even prevented as a result of plugged orifices, Carbon deposits, Oil ring sticking, Thickening and gelling of the lubricating oil as a result of contamination by vegetable oils, and Lubricating problems.
Other disadvantages to the use of vegetable oils and especially animal fats, are the high viscosity (about 11 to 17 times higher than diesel fuel), lower volatilities content which causes the formation of deposits in engines due to incomplete combustion and incorrect vaporisation characteristics. Table 1 shows the percentages of unsaturated fatty acids in some typical oils, the degree of saturation is dictated by the number of double bonds in the fatty acids e.g. 18:01
- 28 -
Adam Khan
denotes a carbon length of 18 with on double bond. At high temperatures there can be some problems with polymerisation of unsaturated fatty acids, this is where cross linking starts to occur between other molecules, causing very large agglomerations to be formed and consequently guming occurs (Pryde, 1983). This problem does not occur with fats as they have a very low concentration of unsaturated fatty acids as can be seen in table 2. Ethanol can be used as an additive to improve these properties, but pure vegetable oils are rarely used for a straight diesel fuel substitute. Ethanol as an additive or a reactant is often preferred in place of other simple alcohols, such as methanol because it is less toxic and because it is easily produced from renewable sources such as biomass or grains (Foglia et al., 2000). Fats, due to their high melting point and viscosity, can not be used directly in diesel engines or mixed with diesel fuels. The degree of saturation of the fatty acids attached to the glycerol backbone determines the boiling point of the triglyceride. As such, compared with oils, little research has been done on their potential as diesel fuel. Some typical composition of fats and oils can be found in tables 1& 2 (Ma, 1999). Fats and oils are primarily composed of triglycerides, esters of glycerol (mono- & diglycerides) and fatty acids (carboxylic acids). The term monoglyceride or diglyceride refers to the number of fatty acids that are attached to the glycerol backbone i.e. a diglyceride would have one hydroxyl group and two fatty acid groups attached to the glycerol backbone as in figure 1.
O CH2 CH C OH R R
COO
CH2 COO Diglyceride
Fig. 1. Diglyceride with two fatty acid groups
Table 1. Typical chemical properties of vegetable oil (Groering et al., 1982*)

Fatty acid composition, % by weight Palmitic Stearic Vegetable oil 16:00 18:00 Corn 11.67 1.85 Cottonseed 28.33 0.89 Crambe 2.07 0.7 Peanut 11.38 2.39 Rapeseed 3.49 0.85 Soybean 11.75 3.15 Sunflower 6.08 3.26 Oleic 18:01 25.16 13.27 18.86 48.28 64.4 23.26 16.93 Linoleic 18:02 60.6 57.51 9 31.95 22.3 55.53 73.73 Linolenic 18:03 0.48 0 6.85 0.93 8.23 6.31 0 arachidic 20:00 0.24 0 2.09 1.32 0 0 0
Acid*Value 0.11 0.07 0.36 0.2 1.14 0.2 0.15
- 29 -
Adam Khan
Fatty acid composition, % by weight Lauric Myristic Palmitic Stearic 12:00 0.1 0.1 0.1 0.1 0.1 46.5 14:00 0.1 0.7 1 1.4 2.8 19.2 16:00 10.2 20.1 42.8 23.6 23.3 9.8 18:00 3.7 2.6 4.5 14.2 19.4 3
Table 2. Typical fatty acid composition - of common oil source (Kincs, 1985)
Lipid Soybean Cottonseed Palm Lard Tallow Coconut Oleic 18:01 22.8 19.2 40.5 44.2 42.4 6.9 Linoleic 18:02 53.7 55.2 10.1 10.7 2.9 2.2 Linolenic 18:03 8.6 0.6 0.2 0.4 0.9 0
Biodiesel
Subsequent research found that triglycerides could be converted into simple alkyl fatty acid esters (Biodiesel) which has similar properties to diesel see table 3 & 4. The first documented commercial production of rapeseed oil methyl esters is reported to be in 1988 (Korbitz 2001). To date biodiesel production can be found in over 28 countries of which Germany and France are the worlds largest [Kusdiana, 2002]. Table 3. Comparison of typical properties of diesel, canola oil and biodiesel
Density (kg/L) Gross calorific value (MJ/L) Viscosity ( mm 2 / s @ 37.8C) C:H:O ( ratio ) Sulphur (%) Diesel 0.835 38.3 3.86 3.59 0.15 Canola 0.922 36.9 37 3.26 0.0012 Biodiesel 0.88 33.3 4.7 2.38 <0.01
Source: Adapted from Table 6.1 of BTCE (1994) and from www.afdc.doe.gov . The C:H:O ratio for biodiesel is taken from http://www.biodiesel.org/fleets/summary.shtml#attributes
Table 4. Properties of ethyl ester and diesel control fuel*

REE (Rapeseed Ethyl Ester). Diesel Gross heat of combustion (MJ/kg) 40.5 45.2 Flash Point C 124 82 Cloud Point C -2 -14 Pour Point C -10 -21 Viscosity cs @ 40C 6.17 2.98 Cetane Number 59.7 49.2 *Based on analysis by Phoenix Chemical Lab, Inc. and the Agricultural Engineering Analytical Laboratory, University of Idaho. REE
Biodiesel has many benefits such as it is biodegradable, non-toxic, has a low emission profile (including potential carcinogens) and is a renewable resource (Ma, 1999). European biodiesel is typically made from rapeseed oil and methanol, whereas in the US it is predominately made from soybean oil. This reflects the agricultural practices of the two regions.
Production Costs
The greatest hurdle in commercialisation of biodiesel is the cost of production. Currently, raw material costs and the cost of production are keeping the retail price of biodiesel too high for it to be an option for many users. The current method of production is with large batch reactors. Raw materials generally consist of methanol and high-quality vegetable oils. Only approximately 55%
- 30 -
Adam Khan
of the US biodiesel industry can use any fat or oil feedstock, the remainder is limited to refined vegetable oils, the least expensive of which is soy bean oil. Although the production of biodiesel is now conducted on a large scale, there are still many problems with using crude feed stock (Ma, 1999). Until these problems are resolved with measures such as a continuous process and the use of crude oils waste fats, such as used cooking oils and abattoir fats, the cost of production will remain relatively high. The recovery of high quality glycerol, a by-product which is required for many other processes, would also contribute to substantially reducing production costs. The European Union has set an objective to secure for motor biofuels a market share of 5% of total motor fuel consumption by 2005 (a significant amount of this will be biodiesel). The US Department of Energy estimates that up to 50% of the total diesel fuel consumption could be replaced with biodiesel. US production of biodiesel, from both fats and oils is currently approximately 7.12 billion litres ( Alcantara 2000).
PROCESS OVERVIEW
There are several generally accepted ways to make biodiesel some more common than others, e.g. blending and transesterification, and several others that are more recent developments .e.g. reaction with supercritical methanol. An overview of these processes is as follows;
1. 2. 3. 4.
Direct use and Blending, which is the use of pure vegetable oils or the blending with diesel fuel in various ratios, Microemulsions with simple alcohols, Thermal Cracking (pyrolysis) to alkanes, alkenes, alkadienes etc Transesterification (alcoholysis) which consists of several sub categories; i. esterification , ii. saponification, iii. hydrolysis (reaction with water) and iv. aminolysis (reaction with amines) Other forms of catalysis
i. biocatalysts ii. reaction with supercritical methanol iii. catalyst free
5.
Direct use and blending

The direct use of vegetable oils in diesel engines is problematic and has many inherent failings. It has only been researched extensively for the past couple of decades, but has been experimented with for almost a hundred years. Although some diesel engines can run pure vegetable oils engines, turbocharged direct injection engines such as trucks are prone to many problems (BTCE, 1994). Energy consumption, with the use of pure vegetable oils, was found to be similar to that of diesel fuel (Hemmerlein et al. 1991). For short term use ratios of 1:10 to 2:10 oil to diesel have been found to be successful (Ma, 1999). The difficulties may be grouped into three key areas:
Viscosity
As can be seen from table 3 the properties of canola oil and diesel are very similar, except a significant difference in viscosity, with canola oil having twelve times the viscosity of diesel. Even after heating to around 80C it is still six times as viscous as diesel. This leads to problems with flow of oils from the fuel tank to the engine, blockages in filters and subsequent engine power losses. Even if preheating is used to lower the viscosity, difficulties may still be encountered with starting due to the temperatures required for oils to give off ignitable vapours.
- 31 -
Adam Khan
Further, engines can suffer coking and gumming which leads to sticking of piston rings due to multi-bonded compounds undergoing pyrolyses. Polyunsaturated fatty acids also undergo oxidation in storage causing gum formation and at high temperatures where complex oxidative and thermal polymerisation can occur ( Ma 1999).
Microemulsions
Microemulsions are defined as a colloidal equilibrium dispersions of optically isotropic fluid microstructures, with dimensions generally in the 1-150 nm range. These are formed spontaneously from two normally immiscible liquids and one or more ionic or non-ionic amphophiles (Schwab et al., 1987). A microemulsion is designed to tackle the problem of the high viscosity of pure vegetable oils by reducing the viscosity of oils with solvents such as simple alcohols. The performances of ionic and non-ionic microemulsions where found to be similar to diesel fuel, over short term testing. They also achieved good spray characteristics, with explosive vaporisation which improved the combustion characteristics (Ma, 1999). In longer term testing no significant deterioration in performance was observed, however significant injector needle sticking, carbon deposits, incomplete combustion and increasing viscosity of lubricating oils where reported ( Ma, 1999).
Thermal Cracking
Pyrolysis is the conversion of one substance into another by means of applying heat i.e. heating in the absence of air or oxygen with temperatures ranging from 450C 850C (Sonntag, 1979b). In some situations this is with the aid of a catalyst leading to the cleavage of chemical bonds to yield smaller molecules (Weisz et al., 1979). Unlike direct blending, fats can be pyrolyised successfully to produce many smaller chain compounds. The pyrolysis of fats has been investigated for over a hundred years, especially in countries where there is a shortage of petroleum deposits. Typical catalyst that can be employed in pyrolysis are SiO2 and Al 2 O3 . The ratios of light to heavy compounds are temperature and time dependent. Typical breakdown of compounds found from pyrolyisis of safflower and soybean oil, are listed in table 5.
Table 5. Compositional data of pyrolysis of oils ( Ma 1999)
Percent by weight *HO Safflower Alkanes 40.9 Alkenes 22 Alkadienes 13 Aromatics 2.2 Unresolved unsaturates 10.1 Carboxylic acids 16.1 Unidentified 12.7 *HO High Oleic Safflower oil Soybean 29.9 24.9 10.9 1.9 5.1 9.6 12.6
The equipment for pyrolysis or thermal cracking is expensive for modest throughputs. Although, the products are chemically similar to pyrochemically based diesel, oxygen removal from the process decreases the products benefits of being an oxygenated fuel. This decreases its environmental benefits and generally produces more fuel similar in properties of gasoline than diesel, with the addition of some low value materials.
Transesterification
Transesterification is the reaction of a lipid with an alcohol to form esters and a by-product, glycerol. It is in principle the action of one alcohol displacing another from an ester, the term
- 32 -
Adam Khan
alcoholysis (cleavage by an alcohol).The reaction, as shown in figure 2 is reversible and thus an excess of alcohol is usually used to force the equilibrium to the product side. The stoichiometry for the reaction is 3:1 alcohol to lipids; however in practice this is usually increased to 6:1 to increase product yield. A catalyst is usually used to speed up the reaction and may be basic, acid or enzymatic in nature (Ma, 1999). The alkalis that are generally used include NaOH, KOH, carbonates and corresponding sodium and potassium alkoxides such as sodium methoxide, ethoxide, propoxide and butoxide. Sodium hydroxide is the most common alkali catalyst that is used, due to economical reasons and availability. Alkali-catalysed reactions are used more often commercially than acid catalysts, as the reactions are faster. Only simple alcohols can be used in transesterification such as, methanol, ethanol, propanol, butanol and amyl alcohol. Methanol is most often used for commercial and process reasons related to its physical and chemical nature (shortest chain alcohol and is polar). However ethanol is becoming more popular as it is a renewable resource and does not raise the same toxicity concerns as methanol (Ma, 1999). The type of catalyst, the reaction conditions and the concentration of impurities in a transesterification reaction determine the path that the reaction follows. For alkali catalysed transesterification, water and FFA are not favourable to the reaction, so anhydrous triglycerides and alcohol are necessary to minimise the production of soap. Soap production decreases the amount of esters and renders the separation of glycerol and esters difficult. In current commercial processes using crude feed stock, excess alkali is added to remove all the FFAs.
CH2 CH COO COO R1 R2 + 3R'OH R3 Simple Alcohol Catalyst R1 R2 R3 COO COO COO Esters R' CH2
OH OH OH
R' + CH R' CH2
CH2 COO Triglyceride
Glycerol
Fig. 2. Transesterification of triglycerides with alcohol
Saponification
The production of soap sometimes called alkaline hydrolysis, converts triacylglycerols to glycerol and a mixture of salts of long-chain carboxylic acids. As can be seen from figures 3 & 4, the reaction can be carried out with an ester (i.e. triglycerides) or with carboxylic acids (i.e. free fatty acids). However, the production of fatty acids is an intermediate step when triglycerides are directly used for saponification. The commercial production of soap is usually conducted in two phases. The first phase is the conversion of lipids into FFAs by boiling with aqueous sodium hydroxide until hydrolysis is complete and then adding sodium chloride to precipitate the soap (Solomon, 1996).
O R C OH + NaOH Heat R O C Salt O -Na + + H 2O Water
Free Fatty Acid Metalic alkoxide
Fig. 3. Saponification from free fatty acid
- 33 -
Adam Khan
or
O R C Ester OR' + NaOH Water R O C Salt O-Na+ + R'OH
Metalic alkoxide
Simple Alcohol
Fig. 4. Saponification from ester
Esterification
The formation of esters occurs through a condensation reaction known as esterification. This requires two reactants, carboxylic acids (fatty acids) and alcohols. (Solomon, 1996). Esterification reactions are acid catalysed and proceed slowly in the absence of strong acids such as sulphuric acid, phosphoric acid, organic sulfonic acids and hydrochloric acid. The equation for and esterification reaction can be seen in figure 5.
O R C OH + R'OH H+ R O C Esters OR' + H2O Water
Free Fatty Acid Simple Alcohol
Fig. 5. Esterification
Hydrolysis
The hydrolysis of lipids forms a heterogenous reaction system made up of two liquid phases. The disperse aqueous phase consists of water and glycerol; the homogenous lipid phase consists of fatty acids and glycerides. The hydrolysis of glycerides takes place in the lipid phase in several stages via partial glycerides (diglycerides and monoglycerides) (Ullmanns, 1987). Acid catalysts are very effective at accelerating the hydrolysis reaction. However, at high temperatures substantial material corrosion occurs. Diabasic metal oxides have a higher activity than more strongly alkaline monobasic metal oxides. Zinc oxide in its soap form has been suggested to be the most active catalyst for hydrolysis reactions (Ullmanns, 1987). Reaction without a catalyst is not economical below 210C, thus requiring the implication of high temperature, pressure techniques. Modern continuous plants operate at pressures between 0.6-1.2 MPa at 210-260C without a catalyst. This increased pressure allows the mutual solubility of the two phases to increases to a point where the formation of continuous phase occurs.
CH2 CH COO COO R1 R2 R3 + 3H2O Water Heat R1 R2 R3 COO COO COO Esters H CH2
OH OH OH
H + CH H CH2
Glycerol
Fig. 6. Hydrolysis of triglycerides
- 34 -
Adam Khan
Aminolysis
Esters undergo nucleophilic substitution at their acyl carbon atoms when they are treated with primary or secondary amines. These reactions are slow but are synthetically useful (Solomon, 1996).
O R4 CH2 CH COO COO R1 R2 R3 Amine +3 H R4 N R5 Heat R1 C N R5 O R4 CH2 OH OH OH
R2 C N R5 + CH O R4 CH2 R3 C N R5 Amides
Glycerol
Fig.7. Aminolysis of triglycerides
Biocatalysts
Biocatalysts are usually lipases; however conditions need to be well controlled to maintain the activity of the catalyst. Hydrolytic enzymes are generally used as biocatalysts as they are ready available and are easily handled. They are stable, do not require co-enzymes and will often tolerate organic solvents. Their potential for regioselective and especially for enantioseletive synthesis makes them valuable tools [Schuchardt, 1998]. Recent patents and articles have shown that reaction yields and times are still unfavourable compared to base-catalysed transesterification for commercial application.
Catalyst Free
Transesterification will occur without the aid of a catalyst, however at temperatures below 300C the rate is very low. It has been said that there are, from a broad perspective, two methods to producing biodiesel and that is with and without a catalyst.
Supercritical Methanol
The study of the transesterification of rapeseed oil with supercritical methanol was found to be very effect and gave a conversion of >95% within 4min. A reaction temperature of 350C, pressure of 30MPa and a ratio of 42:1 of methanol to rapeseed oil for 240s were found to be the best reaction conditions. The rate was substantially high from 300 to 500C but at temperatures above 400C it was found that thermal degradation takes place. Supercritical treatment of lipids with a suitable solvent such as methanol relies on the relationship between temperature, pressure and the thermophysical properties such as dielectric constant, viscosity, specific weight and polarity (Kusdiana, 2000). A comparison of supercritical methanol production and alcoholysis can be seen in table 6.
COMPARATIVE ANALYSIS OF PROCESSES

There are several problems associated with the production of biodiesel with alkali homogenous catalysts. The batch process is time consuming, as purification of the product for catalyst and saponified products is necessary. There is also the problem of immiscible phases of the lipid and alcohol, which requires vigorous stirring to enable good contact of reactants for the reaction to occur (Kusdiana, 2000).
- 35 -
Adam Khan
In the base-catalysed transesterification method use of crude lipids that contain large quantities of FFA leads to catalyst destruction. It is suggested that the concentration of FFA be as low as possible, not exceeding 0.5%.
Table 6.Comparison between production of biodiesel
Common Method Reaction Time 1-6h 0.1MPa, 30-65C Reaction Condition Catalyst acid or alkali Free Fatty Acids Saponified products Yield 97% (normal) methanol, catalyst and saponified Removal for products Purification Process Detailed *SC MeOH method - Supercritical Methanol *SC MeOH method 0.067h 35MPa, 350C none methyl esters 98.50% methanol Simple
PROCESS OVERVIEW OF SEPARATION

Although not within the scope of this report, separation is an integral part of the production of biodiesel and therefore some basic concepts will here within be covered. The refining of the products from the production of biodiesel can be technically difficult and can substantially increase the cost of production. The purity of biodiesel must be high and generally conform to international standards such as the European Union (EU) standards for alternative fuels. Under the EU standards for alternative diesel fuels free fatty acids, alcohol, free and bound glycerin and water content must be kept to a minimum and the fuel must be at least greater than 96.5% pure (Karaosmanoglu, 1996). The typical product mixture of a transesterification reaction contains fatty acid esters (biodiesel), monoglycerides, diglycerides, glycerol, alcohol and catalyst, in varying concentrations. The primary goal is the removal of the esters from the mixture, maintaining low costs and ensuring a high purity product. Glycerol in its pure form is seen to be a secondary product of the reaction as it can be sold to various industries. To keep the cost of production competitive, the removal and resale of glycerol is essential. The remaining mixture contains by-products and alcohol that should have minimal contaminants if the conversion is high, except for the alcohol which would be distilled off.
Alcohol Alcohol Recovery Lipids ~ ~ Reactor Settler Washing ~ Glycerol Catalyst Mineral Acid Neutralisation Distillation Settler Evaporation ~ Purification Evaporation Biodiesel
Fatty Acids
Fig.8. Typical flow chart of the process of esterification of lipids to biodiesel
- 36 -
Adam Khan
If the reaction has reached a high level of conversion the product mixture will form two liquid phases, with a solid phase if a solid catalyst has been used. The bottom phase of the mixture would consist of glycerol and the top phase alcohol and esters. In a reaction that did not reach full conversion the unreacted lipids and bound glycerol would solidify in the bottom layer, as can be seen in figure 12. For molar ratios greater than 5.67 of MeOH it was reported that there are some difficulties in separating glycerin from methanol (Kusdiana, 2000).
Critical Analysis of the Literature

There where only two papers found that made mention of heterogenous catalysts. Of these there were 3 potential catalysts of which only one showed any promise. There is a distinct lack of information or research conducted on heterogenous catalysts. There are several references to co-solvents, however none make mention of biodiesel itself. The references to THF reported that it was suitable as a co solvent claiming it to be a low toxic compound, when the MSDS from Chemwatch showed it to have a high level of toxicity. There was little mention of the thermal degradation of products such as glycerol and triglycerides, particularly as glycerol degrades to a highly toxic compound, acrolein. Generally analysis techniques focused on glycerol and unconverted triglycerides, there was no mention of any residual components such as alcohol. This seems to be due to the currently accepted practice of extensily washing the product before analysis. Testing and quality control seems to focus on downstream fuel that is ready for retail sales, not necessarily testing for optimisation purposes. There is conflicting information on the impact of contaminants such as FFA and water. Some papers have said that water has a negative effect on the reaction and others have implied that acid catalysed reactions are not effected.
Heterogeneously Catalysed Process

The use of hydrogenized guanidines on organic polymers as a catalyst for the transesterification of vegetable oils has been tested by Schuchard et al (1997). Principle tests conducted in a continuous reactor were promising and resulted in a patent (P93246, 1984). However there still appear to be some problems with this technique. For example, catalyst activity seems to deteriorate after an hour of operation (Schuchardt, 1998). This resulted in incomplete reaction and subsequent difficulties in phase separation. Irreversible protonation of the catalyst or catalyst leaching are a possible cause, however no definite test to confirm this had been conducted (Schuchardt, 1998). Experiments conducted by Suppes et al (2000) indicate that CaCO3 may be suitable for use as a heterogeneous alcoholysis catalyst. Preliminary batch experiments used beef tallow or soybean oil as the triglyceride source and diethylene glycol (DEG) as the alcohol. Subsequent attempts to verify these results where conducted using a similar batch apparatus but as yet have failed to replicate those results obtained by Suppes et al. There is only one difference in the experimental procedure used to replicate Suppess work which was the source of the CaCO3 , which was not sourced from Lancaster (Windham, NH). Trials conducted using CaO and Ca (OH ) 2 as the catalyst also appear to have failed to achieve significant conversion. Although, this has not been confirmed with GC analyses, the physical appearance of the samples very much favours a high content of lipids thus subsequent low conversion. CaO is documented as being an effective heterogeneous alcoholysis catalyst and is in common use for monoglycerides(which would be solids at room temperature) production at temperatures ranging from 200oC to 220oC with reaction times of 1 to 4 hours (Suppes, 2001).
The apparent failure CaCO3 as an alcoholysis catalyst may have been due to thermal degradation considering the high temperatures that the experiments were conducted under (>2200C). Other factors that require investigation include the role of glycerolises, polymerisation and hydrolysis
- 37 -
Adam Khan
and the effect of high concentrations of free fatty acids. Certain carbonate systems are also known to promote glycerolisis however the overall effect this would have under these reaction conditions is not yet known. Glycerolisis is the reaction of glycerol with triglycerides to form monoglycerides. Table 7. List of potential catalysts
Ti (OR) 4 (C12 H 27 Sn) 2 O Na 2 CO3 K 2 CO3
- Titanium IV alkoxides - bis(tri-n-butyltin)oxide - Organometallic tin complexes - (Solubility of some of these?)
ZnCO3
MgCO3 CaCO3 ZnO CH 3 COOCa
CH 3 COOBa
- supported on alumina - acetates of calcium and barium (Salts or esters of ethanoic acid)
Lancaster (Windham, NH) Iowa Limestone Co. (Des Moines, IA)
The calcium carbonate catalyst could function as a simple Bronsted base catalyst. The double bond migration, observed in the Na2CO3 and K2CO3 catalysed reactions, indicated that at these reaction temperatures the carbonate anion is an extremely strong base, however this has not been proven.
PRODUCT ANALYSIS
Biodiesel can be significantly contaminated with both free and bound glycerol, triglycerides and alcohol due to incomplete transesterification and or insufficient purification. As noted in literature, the presence of these minor contaminates can be detrimental to both engines and the environment through pollution. There are currently limits on the levels of these compounds in biodiesel that are set out in guidelines such as American Society for Testing and Materials (ASTM) which can be found in table 8. Over the past ten years there have been various investigations into methods to investigate the analysis of biodiesel, its impurities and by-products. The list includes Gas Chromatography (GC) distillation, solid phase separation, thin film liquid chromatography (TLC), High Precision Liquid Chromatography (HPLC), Refractometry, Near Infrared Spectroscopy with a Fibre-Optic probe (NIR) and Thin layer chromatography/flame-ionisation detection (TLC/FID) with an Iactroscan instrument [Plank, 1995, Freedman et al 1984]. The analysis technique needs to be accurate, reliable, reproducible, relatively quick and simple, and require equipment that is readily available. Formally HPLC, but more commonly now GC analysis is by far the most accepted methods for the analysis of biodiesel fuels (Knothe, 1998). The reasons behind this are simplicity, (once the system has been set up) and the high level of accuracy with which the results can be obtained (Plank, 1995).
- 38 -
Adam Khan Table 8. Detailed requirements for biodiesel fuel*

Property Flash Point (closed cup) Water sediment Kinematic Viscosity Sulphated ash Sulphur Copper strip corrosion Cetane number Cloud Point Carbon residue Acid number Free glycerol Total glycerine *As out lined in ASTM PS-121 99 Test Method D93 D2709 D445 D874 D2622 D130 D613 D2500 D4530 D664 F F Limits 100 min 0.05 max 1.9-1.6 0.02 max 0.05 max No. 3 max 40 min RTC 0.05 max 0.8 max 0.02 0.24 Units C %Vol
mm 2 / s
% mass % mass
C % mass mg KOH/g % mass % mass
High Precision Liquid Chromatography

Mittelbach & Trathnigg detailed a method that is similar to GC in that the samples require derivatisation. This was using HPLC coupled with isocratic separation, such as differential refractometry or density detection. There where some problems associated with separation according to molecular mass and hydroxyl content, but where over come by using a combination of a GPC-column set and a polar column in series [Trathnigg, 1990]. Free fatty acids and glycerol where separated from the mixture before the analysis was carried out and it was said that the analysis time was approximately 20 minutes.
Gas Chromatography
The method proposed by Plank and Lorbeer using capillary gas chromatography, can provide qualitative and quantitative information about the concentrations of contaminants in biodiesel. The method is appropriate for measuring minor and major components in a sample, gives a high reliability of results, has simple instrumentation, requires a small amount of sample preparation and has a short analysis time (Plank, 1995).
Near Infrared Spectroscopy

Near Infrared Spectroscopy (NIR) is becoming a preferred method for quality control of biodiesel fuel. Due to its operational ease, rapidity of measurement and non-destructiveness it can be implemented as an online assessment of conversion in the transesterification reaction. The NIR method is less sensitive than GC for quantifying minor components, but can be used in combination with GC analysis for a higher level of analysis. However, that being said it is accurate to the point that it can be used in qualifying online conversion rates and major component concentrations. Standards testing for biodiesel is time consuming and costly, but can be superseded by NIR which would significantly reduce both the cost and the time of testing (Knothe, 1998). The determination of glycerol after water extraction was described by Bondioli et al (Plank, 1995). This would involve the extraction of glycerol from biodiesel into an aqueous phase which has been shown to be highly effective. Glycerol has a high affinity for water, very low for methyl esters, moderate for monoglycerides and small for diglycerides. Solubility of glycerol in methyl esters is low in the order of 0.15% and in the presence of a large amount of water is negligible (Van Gerpen, 1996). The glycerol free mixture would then be analysed via GC or HPLC. Another method is the Conradson carbon residue, sulphate ash and phosphorous content. This involves the combustion of a sample and subsequently weighed to reveal the mass of
- 39 -
Adam Khan
carbon, although quit crude it is accepted as a quick estimate of the conversion in transesterification (Hodl, 1994). LITERATURE REVIEW CONCLUSION To date there has been extensive research into the area of the derivitisation of triglycerides for utilization in diesel engines. The development of a continuous reactor would greatly improve the economic viability of biodiesel as an alternative fuel source to petrochemical derived products. As a result of recent changes in eating patterns, there is an increasing surplus of animal fats that cannot be consumed by the soap industry. Companies are increasingly interested in this as an energy source to be combusted in furnaces for grid energy production. Large-scale production (i.e. continuous reactor) would be required for this to be viable. The development of such a facility would require extensive research into the kinetics of tallow derived biodiesel production as opposed to the well-documented kinetics of vegetable oil derived biodiesel.
- 40 -
Adam Khan
APPENDIX C
Batch # Prime Bio0037 100 Bio0038 31.64 Bio0039 100 Bio0040 100 Reactants g Tallow SaveAll Oil DEG EtOH 95.26 15.42 137.42 Catalyst KOH H2SO4 ZrO2 Rutile Zirconia Sand 2.5 Reaction Conditions CaCO3 Temperature Pressure bar Time min 6.41 210 33 180 210 38 75 6.42 240 1 180 6.412* 253 240 236 236 236 60 60 60 60 237 260 60 60 70 55 61 61 1 1 1 1 32 58 1 1 180 180 180 180 50 180 180 180 180 180 60 180 420 Conversion Glycerol Ethyl Esters Poor Good Quantitative only Poor Unknown# Unknown# No Visible Reaction No Visible Reaction Soap formation Emulsion formation Table 11.2 for results Table 11.3 for results Table 11.4 for results No conversion Table 11.5 for results No conversion Table 11.7 for results Table 6.4 for results
112.3
Bio0041 100 137.42 6.42* Bio0042 Equipment failure Bio0043 100 137.42 3 Bio0044 100 137.42 None Bio0045 100 65.6 2.05 Bio0046 100 31 1.14 Bio0047 300 93.03 2.03 Bio0048 300 93.03 3.14 Bio0049 300 93.03 3.01 Bio0050 100 46.52 None Bio0051 300 93.03 3.05 Bio0052 100 46.5 3 Bio0053 300 93.03 3.05 Bio0054 300 93.03 3.56 * CaCO3 from DML Lime ~35% supplied from David Mitchell (NSW) Pty Ltd # Higher temperatures gave 100% liquid at room temperature, when original state solid
- 41 -

Khan K. Biodiesel Kinetics &amp; Catalyst 2 (Master)

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Khan K. Biodiesel Kinetics &amp; Catalyst 2 (Master)

Uploaded by

Copyright:

Available Formats

RESEARCH INTO BIODIESEL CATALYST SCREENING & DEVELOPMENT

6.4.2 6.5.1 6.5.2 6.5.3 6.5.4 6.5.5 7.0 7.1

Adam Khan 7.3 7.4 8.0 9.0 10.0

2.0 LITERATURE REVIEW

2.3 WATER and FREE FATTY ACIDS

2.4 HETEROGENOUS CATALYSTS

3.0 LITERATURE REVIEW CONCLUSION

Figure 6.1. Parr reactor and PID controller

Non-Pressure Reactor (Modified Beaker)

Figure 6.2. Batch reactor with thermocouple and heating mantle.

6.1.2 Reactants 6.1.2.1 Chemicals

Adam Khan Table 6.3. Typical reaction data sheet

Reactor Vessel Beaker Run Molecular Weight 847.40 46.05 98.07

6.2 ANALYSIS METHODOLOGY 6.2.1 Mass Balance Method

Adam Khan Table 6.4. Mass balance procedure Bio0054

c-b-a d-b-a e-a g-f

42.4 171.75 30.3 319.42 315.7 394.2

46.05 169.07 32.605 316.31 316.31 396.5

6.3 ANALYSIS ERROR

Figure 6.3 No conversion Bio0044

6.4.2 High Temperature Conversion

Figure 6.4 Bio0040, Bio0041 & Bio0041 with water

6.5.1 High Temperature Conversion

6.5.2 Traditional Catalysts

6.5.3 Potential Heterogenous Catalyst

6.5.4 Emulsion Formation

6.5.5 Gravimetric Method

7.1 Catalysts 7.1.1 Sodium hydroxide

7.1.3 Zirconia Sand

7.2 Gravimetric Method

7.3 Emulsion formation

7.4 High Temperature reactions

9.0 FUTURE WORK AND RESEARCH

282.35 Loss of product due to emulsion formation

Table 11.2 Mass balance Bio0047

Table 11.3 Mass balance Bio0048

Adam Khan Table 11.4 Mass balance Bio0049

-5.39 1.74 32.165 -49.56 1 2 3 4 5

Table 11.5 Mass balance Bio0051

Adam Khan Table 11.6 Mass balance Bio0053

-2.15 -4.88 4.48 52.035 -55.39

APPENDIX B LITERATURE REVIEW

BACKGROUND Fats and Oils

CH2 COO Diglyceride

Fig. 1. Diglyceride with two fatty acid groups

Table 1. Typical chemical properties of vegetable oil (Groering et al., 1982*)

Acid*Value 0.11 0.07 0.36 0.2 1.14 0.2 0.15

Table 4. Properties of ethyl ester and diesel control fuel*

Direct use and blending

R' + CH R' CH2

CH2 COO Triglyceride

Fig. 2. Transesterification of triglycerides with alcohol

Free Fatty Acid Metalic alkoxide

Fig. 3. Saponification from free fatty acid

Fig. 4. Saponification from ester

Free Fatty Acid Simple Alcohol

CH2 COO Triglyceride

Fig. 6. Hydrolysis of triglycerides

CH2 COO Triglyceride

Fig.7. Aminolysis of triglycerides

COMPARATIVE ANALYSIS OF PROCESSES

PROCESS OVERVIEW OF SEPARATION

Fig.8. Typical flow chart of the process of esterification of lipids to biodiesel

Khan K. Biodiesel Kinetics & Catalyst 2 (Master)

Khan K. Biodiesel Kinetics & Catalyst 2 (Master)