Analisis Comparativo de Usos de Glicerol

Design and Analysis of Technological Schemes for Glycerol Conversion to Added Value Products
John Alexander Posada Duque
Universidad Nacional de Colombia Facultad de Ingeniera y Arquitectura, Departamento de Ingeniera Elctrica, Electrnica y Computacin Manizales, Colombia 2011
Diseo y Evaluacin de los Esquemas Tecnolgicas para la Conversin de Glicerol en Produtos de Valor Agregado


Universidad Nacional de Colombia Sede Manizales Facultad de Ingeniera y Arquitectura, Departamento de Ingeniera Elctrica, Electrnica y Computacin Manizales, Colombia
Design and Analysis of Technological Schemes for Glycerol Conversion to Added Value Products
Thesis submitted in partial fulfillment of the requirements for the degree of: Doctor of Philosophy in Engineering
Advisor: Ph.D., M.Sc, Chemical Engineer Carlos Ariel Cardona Alzate
Research line: Chemical and Biotechnological Process Engineering Research group: Chemical, Catalytic and Biotechnological Process
Universidad Nacional de Colombia Facultad de Ingeniera y Arquitectura, Departamento de Ingeniera Elctrica, Electrnica y Computacin Manizales, Colombia 2011
______________________ A mi madre y mi hermana por su apoyo, a Patricia por su valiosa presencia.
Acknowledgment
I would like to express thanks to God and to my mother for the most essential reason, the life. I also would like to express sincere thanks to my advisor, Dr. Carlos Ariel Cardona Alzate, who has been for more than six years the most important influence not only in my career but also in my life. Kind thanks to Dr. Ramon Gonzalez for receiving me at his laboratory during the internship. Special thanks to my research fellows, Luis Rincon, Julian Quintero, and Javier Naranjo, for their friendship and unconditional help. And thanks to Patricia Arevalo and to my friends.
Thanks to the National University of Colombia for the financial support, to the Research Office of National University of Colombia branch Manizales for the financial support in the internship in United States, to the Research and Extension Projects Office of National University of Colombia branch Manizales for the financial support in air tickets to United States, to the Department of Electricity, Electronic and Computational Engineering of National University of Colombia branch Manizales for the financial support for attending to congresses.
Resumen y Abstract
Resumen
Uno de los principales problemas relacionados con la creciente industria del biodiesel es la sobreproduccin de glicerol, que se obtiene en una relacin en peso de 1/10 (glicerol/ biodiesel). Lo que ha llevado a que su precio de venta caiga en un orden de magnitud. As, la gran cantidad de glicerol co-producido puede ser usada como una materia prima renovable y de bajo costo para producir compuestos qumicos y combustibles. Aqu se analiza la conversin qumica y bioqumica de glicerol hacia productos de valor agregado basado en criterios tecno-econmicos. Entonces se consideraron nueve productos finales (va qumica: gas de sntesis, acrolena y 1,2-propanodiol; va bioqumica: etanol, 1,3-propanodiol, cido D-lctico, cido succnico, cido propinico y poly-3-hidroxibutirato). Adems, un total de 27 esquemas tecnolgicos fueron diseados, simulados y evaluados econmicamente utilizando Aspen Plus y Aspen Icarus Process Evaluator. Como una conclusin, una plataforma de biorefinerias basada en glicerol fue obtenida para la produccin rentable de combustibles fsiles y bioclsticos.
Palabras clave: Conversin de glicerol, diseo de procesos, simulacin de procesos, evaluacin de procesos, biorefinerias basadas en glicerol.
VI
Glycerol Conversion to Added Value Products
Abstract
An important concern related to the growing biodiesel industry is the over-production of raw glycerol as by-product, which is obtained in a weight ratio of 1/10 (glycerol/biodiesel). This fact had led to a 10-field drop of its sale price. Thus, the large amount of byproduced glycerol can be used as low-cost and renewable feedstock in order to produce chemicals and fuels. Here, the chemical and bio-chemical conversion of glycerol to added-value products was analyzed based on techno-economic criteria. In this way, nine final products (for chemical conversion: syn-gas, acrolein, and 1,2-propanediol; while for fermentative conversion: ethanol, 1,3-propanediol, D-lactic acid, succinic acid, propionic acid, and poly-3-hydroxybutyrate) were considered. And a total 27 technological schemes were designed, simulated, and economically assessed, using Aspen Plus and Aspen Icarus Process Evaluator. As a conclusion a glycerol-based platform for biorefineries was obtained for the profitable production of fuels, chemicals, and bio-plastics.
Keywords:
Glycerol
conversion;
process
design;
process
simulation;
process
assessment; glycerol-based biorefineries.
Contents
VII
Table of Contents
ACKNOWLEDGEMENTS RESEMEN ABSTRACT TABLE OF CONTENTS LIST OF FIGURES LIST OF TABLES 1. Introduction 1.1. Application field and Motivation 1.2. Thesis objectives 1.3. Thesis structure References Pg. IV V VI VII X XIII 1 2 5 5 7
2. Chapter 2: The glycerols world 2.1. Overview 2.2. Biodiesel industry 2.3. Glycerol market and its oversupply problem 2.4. Glycerol as raw material References
9 9 10 12 13 14
3. Chapter 3: Methodology for processes design and analysis 3.1. Processes design 3.2. Processes simulation 3.3. Processes assessment References
17 17 20 22 24
4. Chapter 4: Separation and purification of glycerol 4.1. Commercial qualities of glycerol 4.2. Effect of the feedstock for biodiesel production on glycerol composition 4.3. Conventional purification process 4.4. Alternative purification process
25 25 27 28 29
VIII
Glycerol Conversion to Added Value Products 4.5. Simulation of the glycerol purification process 4.6. Economical assessment for glycerol purification processes 4.7. Conclusions References 31 33 34 35 36 36 39 41 41 43
5. Chapter 5: Chemical conversion of glycerol 5.1. Oxidation 5.2. Reduction 5.3. Etherification 5.4. Pirolysis and gasification References
6. Chapter 6: Biochemical conversion of glycerol 6.1. 1,3-Propanediol 6.2. Ethanol 6.3. Poly-3-hydroxybutirate 6.4. D-Lactic acid 6.5. Succinic acid 6.6. Propionic acid References
47 47 48 51 53 54 58 59
7. Chapter 7: Study cases for chemical conversion of glycerol 7.1. Generalities 7.2. Acrolein production 7.3. Hydrogen production 7.4. 1,2-propanediol production 7.5. Economic assessment 7.6. Conclusions References
69 69 70 73 76 78 81 81
8. Chapter8: Study cases for biochemical conversion of glycerol 8.1. 1,3-Propanediol production 8.2. Ethanol production 8.3. PHB production 8.4. D-Lactic acid production 8.5. Succinic acid production 8.6. Propionic acid production 8.7. Economic assessment 8.8. Conclusions References
85 85 104 115 124 138 148 153 167 169
Contents 9. Chapter 9: Experimental setup for glycerol fermentation to PHB 9.1. Generalities 9.2. Materials and methods 9.3. Analytical Methods 9.4. Results and discussion References 183 183 186 187 189 192
IX
10. Chapter 10: Conclusions
194
11. Chapter 11: List of Publications
199
Figures
List of Figures
Page Figure 1.1. World biodiesel production and capacity Figure 1.2. Global biodiesel production by feedstock Figure 3.1. Hierarchical decomposition according to the "onion diagram" Figure 3.2. The process design method based on the so called breadth-first Figure 4.1. Flowsheet of conventional schemes for glycerol purification Figure 4.2. Flowsheet of the Ambersep BD50 process Figure 4.3. Simplified flowsheet for raw glycerol purification Figure 5.1. Possible products for glycerol oxidation Figure 6.1. Schematic representation of glycerol degradation process on the part of Escherichia coli, on non fermentative process. Figure 6.2. Main metabolic pathways for fermentative degradation of glycerol by Escherichia coli. Figure 6.3. Pathways involved in the microaerobic utilization of glycerol in E. coli Figure 6.4. Products that can be synthesized from succinic acid Figure 6.5. Pathways involved in the micro aerobic utilization of glycerol Figure 7.1. Simplified flowsheet for acrolein production by glycerol dehydration Figure 7.2. Simplified flowsheet for hydrogen production by gasification Figure 7.3. Simplified flowsheet for 1,2-propanediol production by hydrogenolysis Figure 8.1. Hysteresis loops and multiple steady states Figure 8.2.a) 1,3-propanediol volumetric productivity, (the column in the right side gives the scale). b) Region of multiplicity of steady states, optimal productivity for each dilution rate, global optimal productivity, and wash-out line. 3 4 19 20 29 30 31 38 49
50
54 55 57 71 74 77 89 90
Figures Figure 8.3. 1,3-Propanediol productivity and concentration in the second fermentation stage Figure 8.4. Volumetric productivity in the second fermentation stage using the optimal dilution rate obtained by the model 1 for the first fermentation stage, (the column in the right side gives the scale). Figure 8.5. Product of productivities of both fermentation stages using the optimal dilution rate obtained by the model 1 for the first fermentation stage, (the column in the right side gives the scale). Figure 8.6. Acetylation reaction of 1,3-propanediol with iso-butyl aldehyde to 2-iso-propyl-1,3-dioxane Figure 8.7. Simplified flowsheet for 1,3-propanediol production from raw glycerol Figure 8.8. Residue map curves for the reactive system Figure 8.9. Direct separation with fed 0.377645/0.6223552iP13DO/Water Figure 8.10. P/W ratio, Direct Separation (XF: 0.377645/0.622355-2iP13DO/water)
XI 91
92
93
95
98 99 100 101
Figure 8.11. iso-Volatility curve (Wateriso-Butyraldehyde2-iso-Propil-1,3-Dioxane) 101 Figure 8.12. Simplified flowsheet of fuel ethanol production from glycerol at 88 wt % and 98 wt %. Figure 8.13. Stages for ethanol production from sugar cane, corn, and crude glycerol Figure 8.14. Simplified flowsheet for ethanol production from: (A) sugar Cane and (B) Corn Figure 8.15. Flowsheet for the integrated process of combined biodiesel and bioethanol production Figure 8.16. Flowsheets for PHB production from glycerol (88 or 98 wt %) Figure 8.17. Scheme for the simulation procedure to synthesize PHB from crude glycerol Figure 8.18. Complex formed during the reactive extraction process of D-lactic acid Figure 8.19. Simplified flowsheet for D-lactic acid production from raw glycerol Figure 8.20. Reaction complexes of succinic acid, formic acid and acetic acid with TOA Figure 8.21. Simplified flowsheet for succinic acid production from raw glycerol 104
107
108
112
120 124
133
134 144
144
XII
Glycerol Conversion to Added Value Products 150 189 190 190 191
Figure 8.22. Simplified flowsheet for propionic acid production from raw glycerol Figure 9.1. Accumulation profile I of Bacillus megaterium Figure 9.2. Accumulation profile II of Bacillus megaterium Figure 9.3. Accumulation profile III of Bacillus megaterium Figure 9.4. Accumulation profile IV of Bacillus megaterium
Tables
XIII
List of Tables
Page Table 3.1. Used costs and prices for the economic assessment Table 4.1. Quality specifications for the main qualities of glycerol Table 4.2. Fatty acid profile of vegetable and used oils Table 4.3. Composition of the glycerol layer obtained by decantation during the biodiesel production from different feedstocks Table 4.4. Simulation results for raw glycerol purification process Table 4.5. Purification costs (PC) of raw glycerol (US$/L) Table 7.1. Simulation results for dehydration process from glycerol Table 7.2. Simulation results for gasification process from glycerol Table 7.3. Simulation results for hydrogenolysis process from glycerol Table 7.4. Production costs for glycerol conversion to added-value Table 7.5. Percentage of Production costs for glycerol conversion to added-value Table 8.1. Results summary for each optimization model Table 8.2. Fermentation results for the three considered scenarios Table 8.3. stoichiometric reactions for each scenario and each fermentation stage Table 8.4. Singular Points ** - Acetilation System of 1,3-PD* with 2iP13DO* Table 8.5. Summary of the main simulation results for 1,3-propanediol production from glycerol Table 8.6. Data representing the behavior of the downstream process Table 8.7. Simulation results for fuel ethanol production from glycerol Table 8.8. Main input data and operation conditions used in the simulation process 23 25 27 28
32 33 72 74 78 79 79 94 96 97 99 102
104 106 113
XIV
Table 8.9. Main process streams for ethanol production from lignocellulosic biomass 114 Table 8.10. Process conditions for glycerol fermentation Table 8.11. PHB Extraction Methods Table 8.12. Process conditions for PHB recovery: Downstream Process I Table 8.13. Process conditions for PHB recovery: Downstream Process II Table 8.14. Process conditions for PHB recovery: Process III 116 118 121 121 122
Table 8.15. Downstream processes for lactic acid recovery from a fermentation broth 131 Table 8.16. Base information for the glycerol fermentation to D-lactic acid Table 8.17. Stoichiometry for glycerol fermentation to D-Lactic Acid by Engineered E. coli Table 8.18. Summary of the main simulation results for D-lactic acid production process Table 8.19. Data representing the behavior of the downstream process for D-lactic acid production Table 8.20. Base information for the glycerol fermentation to succinic acid Table 8.21. Stoichiometry for glycerol fermentation to succinic acid by Engineered E. coli 132 134
135
137 141 141
Table 8.22. Removal efficiency (%) of the carboxylic acids from the fermentation 142 broth. Table 8.23. Summary of the main simulation results for succinic acid production process Table 8.24. Data representing the behavior of the downstream process for succinic acid production Table 8.25. Stoichiometry of the fermentation process for each scenario Table 8.26. Summary of the main simulation results for prpionic acid production process. 145
147
149 151
Table 8.27. Data representing the behavior of the downstream process for propionic 153 acid production Table 8.28. Economic results for raw glycerol conversion to 1,3-propanediol: Cost (USD$/kg) and Share (%) 154
Table 8.29. Bioconversion costs (BCCs) for fuel ethanol production form raw glycerol 166
Tables Table 8.30. Global production costs (GPCs) for fuel ethanol production from raw glycerol Table 8.31. Discriminated costs for integrated biodiesel and raw-ethanol production from oil palm Table 8.32. Total PHB production costs from crude glycerol through raw glycerol (88 wt %) and pure glycerol (98 wt %). Table 8.33. Main producers of PHA in the world Table 8.34. Economic results for raw glycerol conversion to D-lactic acid: Cost (USD$/kg) and Share (%) Table 8.35. Economic results for raw glycerol conversion to succinic acid: Cost (USD$/kg) and Share (%) Table 8.36. Economic results for raw glycerol conversion to propionic acid: Cost (USD$/kg) and Share (%).
XV 157
158
159
160 162
164
166
Table 9.1. Some microorganisms PHB producer from different agroindustrial wastes 185 Table 9.2. Comparison between experimental results for glucose and glycerol fermentation to PHB 191
1. Introduction
Fossil sources have diminished significantly because they have been used as the main raw material for the current economy and life style. For instance, large scale products such as transportation fuels and daily use components are obtained from petrochemical industry. Furthermore, the increasing demand of fossil sources from developing economies (like China and India) and speculations about oil reserves availability have caused high crude oil prices. In fact, experts predict the end of cheap oil in 2040 at the latest [1], which are currently and again above USD$ 100 per barrel. This economic issue added to the environmental conscience, which is focus on the problems derived from pollution and accumulation of greenhouse gases, have taken to develop alternative technologies in order to produce sustainable fuels and chemicals using renewable resources. In this way, biodiesel and bioethanol are the most important technological platforms for liquid fuels production.
Although biofuels such as biodiesel and bioethanol represent a renewable, convenient, and environmental friendly alternative for fossil fuels substitution, they also cause concerns in relation to their economic viability. Implementation of biorefineries as an additional process to the biofuels production is an interesting alternative to both overcome the limited profitability of these technologies and use the generated sub-products. Therefore, the concept of biorefinery could be especially advantageous if the conversion of by-products or wastes to added-value products is considered [2].
Glycerol as the main by-product on biodiesel production is obtained at high concentration in a weight ratio of 1/10 (glycerol/biodiesel). Moreover, the growing market of biodiesel has generated a glycerol oversupply, where its production increased 400% in a two years period and consequently the glycerol commercial price fell down near to 10 fold during the same period of time [3]. As a result of the low prices of glycerol, traditional producers such as Dow Chemical, and Procter and Gamble Chemicals, stopped its production [4].
Since glycerol sales have represented an important profitability for biodiesel industry, it is reasonable that low prices of glycerol could impact the economy of biodiesel producers negatively. For that reason, the correct exploitation of glycerol as raw material should be focused on its transformation to added-value products. Thus, the use of glycerol is a highpriority topic for managers and researchers related to biofuels production. In this sense, the establishment of glycerols biorefineries able to co-generate added-value products is an excellent opportunity not only to raise the profitability but also to produce other chemicals from a biobased raw material.
In order to analyze the glycerol conversion possibilities, this highly functional molecule has been identified as a potential raw material for organic synthesis of many intermediates and chemical products. Chemically glycerol can be transformed by many ways such as oxidation, hydrogenolysis, etherification, pyrolysis, and gasification. Thus, different kinds of products such as acrolein, 1,2-propanediol, polyglycerols, syn gas, among many others compounds can be chemically obtained. On the other hand, because of glycerol is a structural component of many lipids, it can also be biochemically transformed to added value compounds. Some products of glycerol fermentation are: 1,3propanediol, ethanol, propionic acid, citric acid, lactic acid, poly-3-hydroxybutirate, and biosurfactants. Then, due to the wide variety of potential products from glycerol, its biorefineries are an excellent commercial opportunity. In this way, it is necessary to determine the most appropriate alternative for glycerol transformation. In this study, the process design and the assessment of different technological schemes for glycerol transformation to added-value components is systematically performed considering technologic and economic indicators.
1.1 Application field and motivation

Biodiesel production sector is a dynamic industry with a rapid global market growth. For instance, over the past decade the biodiesel production was governmentally driven aiming to the development of large scale industries. Thus, Europe took the lead with more than 1.6 mill Tons of biodiesel produced in 2002 (at capacities of approx. 2.1 mill Tons), while in the USA approx. 40.000 T were produced [5]. Furthermore, in 2008 the global biodiesel production reached more than 11.1 mill Tons (see Figure 1.1), representing around 1 % of all diesel consumption of the USA and between 2-3 % of the total transportation
1. Introduction
consumption in Europe [6]. Even though, Europe represents 80% of global biodiesel production and consumption, the U.S. is increasing its production at a faster rate than Europe, while Brazil is expected to surpass the U.S. and European biodiesel production by the year 2015 [6].
35 30 25 20 15 10 5 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 Production Capacity
Figure 1.1. World biodiesel production and capacity.
Currently, new economic and environmental concerns are leading to create governmental incentives targeting to a combination of: reduction of petroleum imports and increase of production and consumption of renewable fuels. In this way, Europe, Brazil, China, and India each have targets to replace 5% to 20% of total diesel with biodiesel [6]. In addition, if governments promote the development of second generation biofuels (and their production using alternative and non-food feedstocks) throughout investment and politics, the prospects for biodiesel market will be early reached. Figure 1.2 shows the expected biodiesel production from different feedstocks where the share in total biodiesel production from edible vegetable oil could decrease from almost 90% to about 75% by 2019. This expected change is due to the development of biodiesel production from jatropha mainly in India and to the increasing use of animal fats to produce biodiesel in the USA. Also, biomass based biodiesel could represent almost 6.5% of total biodiesel production by 2019.
Figure 1.2. Global biodiesel production by feedstock
Under the above described situation, high quantities of raw glycerol are continuously produced since it is obtained in ratio of 9 wt % respect to the produced biodiesel. Then, in order to avoid both economic and environmental drawbacks related to the use and disposal of glycerol, new applications for glycerol must be proposed. Even though, the most traditional applications of glycerol have been related to its use as additive in: food, tobacco, pharmaceuticals and medicine, and for the synthesis of trinitroglycerine, alkidic resins, and polyurethanes, one of the most attractive alternatives for glycerol utilization is as feedstock for producing added-value compounds such as: bioplastics, platform chemicals, and fuels. Thus, because of both the low prices and high availability of glycerol, this compound could be a great opportunity to make money through biorefineries built adjacent to the biodiesel production plant.
From a chemical view of point, glycerol is a highly versatile molecule with two primary hydroxyl groups and a secondary hydroxyl group which offers different reaction possibilities. Meanwhile from a biochemical view of point, the glycerol molecule is abundant in nature in the form of triglycerides (a chemical combination of glycerol and fatty acids) which are the major constituents of nearly all vegetable oils and animal fats. Thus, the high functionality and occurrence in nature of glycerol allow it to be transformed by a chemical route or a fermentative way, as it was above indicated.
1. Introduction
In this way, the most important possibilities for glycerol transformation to added-value compounds are here reviewed and methodologically assessed by mean of processes engineering tools such as: process design, process simulation, and economic evaluation. And finally all the analyzed possibilities are systematically compared.
1.2 Thesis objectives

This thesis aims to design and assess technological schemes for the conversion of raw glycerol obtained during the biodiesel production to added-value products, in order to identify the best alternatives from a technical and economic view of point. Thus, this research required: (i) to identify and select the most promissory possibilities for glycerol transformation, (ii) to simulate and assess the chosen technological schemes and scenarios for the several identified potential products from glycerol, and (iii) to compare these technological schemes based on economic criteria.
1.3 Thesis structure

This thesis presents the results of different studies that have been already published or are under review for their publication. The thesis is accordingly divided into the following chapters:
Chapter 2: The glycerols world
This chapter introduces to the reader with the current status of glycerol as the by-product on biodiesel production and discusses the glycerol problem related to its oversupply. Additionally, the main uses of glycerol as additive and its market are also presented. Finally, the glycerol conversion possibilities are described.
Chapter 3: Methodology to design and analyze processes based on simulation tools.
This chapter details the used methodology for the process design, processes simulation, and process assessment in both cases, i.e., chemical and fermentative conversion of glycerol.
Chapter 4: Separation and purification of glycerol
This chapter presents the requirements for the most important commercial qualities of glycerol, as well as the influence of the feedstock used for biodiesel production on the glycerol layer. Additionally, conventional and non-conventional processes for raw glycerol purification are discussed. Finally, the purification costs of raw glycerol up to different commercial qualities are obtained based on simulation and economic assessment tools.
Chapter 5: Chemical conversion of glycerol
This chapter reviews the alternatives for chemical conversion of glycerol by different reaction ways such as: oxidation, reduction (hydrogenolysis), etherification, pirolysis, and gasification. Conversion levels, yields, selectivities, and productivities are also presented.
Chapter 6: Biochemical conversion of glycerol
This chapter reviews the alternatives for fermentative conversion of glycerol by different strains. Fermentation products such as: 1,3-propanediol, ethanol, lactic acid, succinic acid, propionic acid, poly-3-hydroxybutyrate, and biosurfactants are discussed.
Additionally, conversion levels, yields, selectivities, and productivities are presented.
Chapter 7: Cases of study for chemical conversion of glycerol
This chapter presents the flowsheets, simulation results, and economic assessments for the chemical conversion of glycerol to: acrolein, 1,2-propanediol, and hydrogen.
Chapter 8: Cases of study for biochemical conversion of glycerol
This chapter presents the flowsheets, simulation results, and economic assessments for the fermentative conversion of glycerol to: 1,3-propanediol, ethanol, D-lactic acid, succinic acid, propionic acid, and poly-3-hydroxybutyrate.
Chapter 9: Experimental setup
This chapter shows the experimental setup performed for poly-3-hydroxybutyrate production from glycerol using two strains: cupriavidus necator and bacillus megaterium.
1. Introduction
Chapter 10: Conclusions
This chapter contains the general conclusions of the thesis and also presents the contributions made during this work. Finally, some recommendations for future works are given.
Chapter 11: List of publications and submitted papers
This chapter shows the published results throughout scientific meeting, papers, book chapters, invited book chapters, and books. Also, a list containing the submitted papers was included.
References
[1] Posada J.A., Orrego C.E., Cardona C.A. 2009. Biodiesel production: Biotechnological approach. International Review of Chemical Engineering (I.Re.Che.), 1(6):571-580. [2] Yazdani S.S. and Gonzalez R., 2007. Anaerobic fermentation of glycerol: A path to economic viability for the biofuels industry. Current Opinion in Biotechnology, 18:213219 [3] Posada J.A., Cardona C.A., 2010. Anlisis de la refinacin de glicerina obtenida como co-producto en la produccin de biodiesel (Validation of glycerin refining obtained as a by-product of biodiesel production). Ingeniera y Universidad 14:2-27. [4] Posada J.A., Cardona C.A., Cetina D.M., Orrego C.E., 2009. Bioglicerol como materia prima para la obtencin de productos de valor agregado (Bioglycerol as raw material to obtain added value products). En: Cardona C.A. (ed). Avances investigativos en la produccin de Biocombustibles (Reasearching advances for biofuels production). Manizales: Ed. Universidad Nacional de Colombia sede Manizales. p. 103-127. ISBN: 978-958-44-5261-0. [5] Worldwide review on biodiesel production. Austrian Biofuels Institute,
www.biodiesel.at, 2003. [6] Biodiesel 2020: Global Market Survey, Feedstock Trends and Market Forecasts.
2. The Glycerols World

This chapter describes the relationship between the market of both glycerol and biodiesel, and it also discusses the influence of the growing biodiesel production on the commercial prices of glycerol. Additionally, the potential of raw glycerol for biorefineries developing using it as a main feedstock is presented. Finally, an overview on the possibilities of glycerol transformation by chemical and biochemical routes is given.
2.1 Overview
The glycerol molecule (1,2,3-propanetriol) is a highly reactive tri-alcohol which has two primary and a secondary hydroxyl groups. Some physical characteristics are: water soluble, colorless, odorless, viscous, and hygroscopic; with a specific gravity of 1.261 g mL-1, melting temperature of 18.2 C, and a boiling temperature of 290 C (accompanied by decomposition). Chemically, glycerol is able for reacting with a stable alcohol under most operational conditions, and it is basically non-toxic to human health and environment. The key of its usefulness is the particular combination among its physicochemical properties, compatibility with other substances, and easy handling. Due to these particular properties set, glycerol has found more than 1500 end-uses or large volume applications.
Glycerol is a commodity chemical obtained mainly as by-product in the oleochemical and biodiesel industry; meanwhile glycerin is the commercial name for mixtures containing high quantities of glycerol. This molecule is one of the most versatile substances known due to its unique combination of physical and chemical properties, which allows it to be used in multitude of products and additionally it is often used as: humectant, plasticizer, emollient, thickener, solvent, dispersing medium, lubricant, sweetener, and antifreeze. Glycerol is naturally combined with triglycerides in all animal fats and vegetable oils, representing about 10% of these materials. This component is derived from fats and oils
10
during fatty acids and soap production, or by the transesterification process with alcohols for biodiesel synthesis. Although glycerol can also be produced synthetically by petrochemical processes from epichlorohydrin and using propylene as raw material, such processes are no longer conducted at the industrial level [1].
The glycerols world has a complex behavior since it is by-produced with biodiesel, and in addition its price is related to the no-predictable network of both its supply and demand. This is a typical behavior for a commodity used as additive in many applications and now being used as raw material for the production of platform chemicals, bioplastics, and biofuels. Here the most important topics related to the glycerol industry are elucidated.
2.2 Biodiesel industry

Biodiesel is defined as a clean burning fuel used for diesel engines, manufactured from renewable sources (vegetable oils, animal fats, or used cooking oils) and short chain alcohols (methanol, ethanol, or butanol), to produce a methyl, ethyl or butyl esters fatty acids mixture. A vegetable oil usually contains up to 14 different kinds of fatty acids [2].
Most of biodiesel production processes were developed in the early 40s, during World War II by explosives manufacturers searching for a simpler way to obtain glycerol. Now, biodiesel is commercially produced from agricultural products such as rapeseed, soy bean, and palm oils. Also, other high fatty acid feedstocks such as: used frying oil, grease trap waste oil, and waste tallow or lard, have been used. Several variables as local availability, cost, government support, and fuel performance must be analyzed in order to choose the best feedstock, since biodiesel production costs are highly dependent upon the feedstocks price.
Biodiesel is a fuel with low viscosity and pour point, non-toxic, and biodegradable, which is also cleaner than diesel. Biodiesel is mainly composed by a mixture of fatty acid alkyl esters (FAAE), which can be produced from vegetable oils, wasted cooking oils, and animal fats. Thus it is considered as renewable fuel source. Recently biodiesel has been promoted as a way for enhancing energy independence, promoting rural development, and reducing green house gas emissions.
11
Biodiesel can be produced through the reaction between feedstock oil with either methanol or ethanol. Oils solubility in methanol is lesser than in ethanol, and rate reaction is mass transfer limited and methanol makes higher equilibrium conversion due to higher reactive intermediate methoxide. Most of the biodiesel is produced currently using methanol, which is petrochemically obtained. This dependence on methanol could be considered as non renewable basis. On this way, different efforts to produce biodiesel from ethanol are carried out to generate a renewable process [3-5]. Also ethanol could be a renewable alternative to produce biodiesel because it can be obtained from glycerol which could be also obtained during the same biodiesel process [3-6].
Transesterification process can be carried out by two ways, chemically or biocatalytically catalyzed. Chemical catalysis has other two alternatives, alkali- and acid- catalysis. Industrial production of fuel biodiesel is performed by methanolysis using alkaline catalysts, and high conversion levels in short reaction times are reached. However this way has several drawbacks: free fatty acids (FFAs) and water interfere with the reaction generating fatty acid alkaline salts (soaps). Soaps should be removed by washing water, which also removes glycerol, methanol (MeOH), and catalyst. Also alkaline catalyst has to be removed from the product. Raw glycerol as by-product should be treated as a waste material making the glycerol recovery difficult, and the alkaline wastewater requires treatment. It is also an energetically intensive process [7-8]. On the other hand, in acid catalysis process, sulfuric and sulfonic acid are preferred because these carry out high alkyl esters yields. But elevated reaction temperatures (>100C) and reaction times (ca 50 h) to complete conversion are required.
Commonly, a catalyst is used to improve reaction rate and yield, and an alcohol excess is utilized to shift the equilibrium towards the products side. Among the used catalysts are alkalis (NaOH, KOH, sodium and potassium alcoxides, carbonates, etc), acids (sulfonic acids, HCl, H3PO4, H2SO4, zeolites), enzymes (lypases), and whole cells.
Acid catalysis produce high alkyl esters yield. High reaction temperatures and reaction times to obtain complete conversion are required. Basic catalysis is a quick reaction, with high yields, which take place under moderate conditions in comparison with acid catalysts, but chemical transesterification using an alkali-catalysis has several drawbacks like soap formation by saponification, and difficult recovery of glycerin by emulsion
12
development. In contrast, biocatalysts allow synthesis of specific alkyl esters, easy recovery of glycerol, and transesterification of glycerides with high free fatty acid content [9], and its main disadvantages are the biocatalyst cost and lower reaction rates [10-12].
Lipase enzymes have been used in biodiesel production in free form or immobilized on some different materials such as ceramics, kaolinites, silica, etc. [13]. In general immobilization enhances the stability of lipase due to the ability of the support material to retain just the right quantity of water for the enzyme to remain active. Different reactive mixtures containing water have been analyzed. For example, immobilized Rhizopus delemar and Rhizomucor miehei lipases efficiently catalyze alcoholysis with long-chain fatty alcohols even in the presence of 20% water [14]. However enzymatic methods have not been industrialized because the enzymes have high price and instability [7, 11-12].
Biodiesel production could be fully sustainable if ethanol is produced from glycerol, which is the by-product in biodiesel production. Also, enzymatic transesterification can be carried out using ethanol with low water content or azeotropic ethanol, without affecting considerably the biodiesel production. Genetically modified E. coli [6] and E. aerogenes [4-5] have been reported to ferment crude glycerol or pure glycerol to ethanol. In order to close the renewable biodiesel production in an integrated biotechnological system the follow structure is analyzed in the next section: aqueous-ethanol as raw material, biocatalysts use, and biological transformation of glycerol to ethanol.
2.3 Glycerol market and its oversupply problem

Until 2003 supply of raw glycerol in the market remained relatively stable, when the production of biodiesel started increasing in the USA [15]. Since then, the availability of crude glycerol has been almost doubled, while its demand has remained almost unchanged. Thus, combined effect of supply excess and limited demand of raw glycerol led to low sale prices. Although pure glycerol is an important feedstock in many industrial sectors, raw glycerol must be refined by large scale biodiesel producers using traditional separation processes to remove impurities such as fatty acids, alcohol, and catalyst. Some of these processes are filtration, chemical additions, and fractional vacuum distillation. Generally processes are expensive and economically unfeasible for small and medium scale plants [16].
13
Traditional commercial applications of glycerol are related to its use as an additive or raw material. The industrial sectors who consume glycerol are: pharmaceutical (18%), personal care (toothpaste and cosmetics 16%), polyether/polyols manufacture (14%), food (11%), triacetin (10%), alkyd (8%), snuff (6%), detergents (2%), cellophane (2%), and explosives (2%). The remaining share (11%) is used in the manufacture of lacquers, varnishes, inks, adhesives, plastic synthetics, regenerated cellulose, and other industrial uses [17].
Annually nearly 160000 tons of glycerol is used for technical applications and it is expected an annual growth rate of 2.8%. Raw glycerin supply in the market remained relatively stable until 2003, when biodiesel production started to increase in the U.S. and the E.U. [15]. Then, availability of raw glycerin has almost doubled, and its demand has remained largely unchanged. This excess supply and limited demand has taken to low glycerol prices, but although refined glycerin prices have decreased in the last years; the strongest impact has been suffered by raw glycerin, and thus its sale prices plummet quickly [18].
Since 2006, the glycerol oversupply forced to biodiesel producers to receive sales prices of 2 cents per pound or even lower prices for the raw product. But at mid-2007, reached prices were between 6 and 10 cents per pound [19]. On the other hand, refined glycerin prices have had a similar behavior, with prices as low as 20-30 cents per pound, depending on the quality and purity [18-19]. In this sense the raw glycerin market will remain weak while large amounts of this raw component being available. Therefore glycerol is nowadays a key problem in biodiesel production, and its low sale price could convert this by-product in a residue, then the biodiesel producers must be found alternative uses to avoid the continue falling on the glycerol price.
2.4 Glycerol as raw material

Development of biorefineries based on raw glycerol to produce high-value compounds is necessary in the biodiesel industry to overcome the economic glycerol drawback. The simplest alternative to increase the value of raw glycerol is refining it to technical glycerin, food or pharmaceutical grade, although to synthesize value-added components by chemical o fermentative via are alternatives with higher potential. Chemically glycerol can
14
be transformed to: oxidation products on metallic catalysts as Pt, Pd, Au, using promoters as Bi and Pb; glycols by hydrogenolysis on Ru, Cu and Pt catalysts; polyglycerols by etherification on zeolites and mesoporous materials; and syngas by pyrolysis and gasification.
Also, due to glycerol is abundant in nature and produced by yeasts during osmoregulation to decrease extracellular water activity [20], its wide occurrence allows to different kinds of microorganisms metabolizing glycerol as a sole carbon and energy source, and then this may substitute traditional carbohydrates, such as sucrose, glucose and starch, in some industrial fermentation processes [21-23]. Glycerol can be transform by fermentative via to 1,3-propanediol, dihydroxyacetone, succinic acid, propionic acid, ethanol, citric acid, pigments, polyhydroxyalcanoate, and biosurfactants. [24]. The following sections review the main technological topics related to glycerol transformation by chemical and biochemical via.
References
[1] McCoy M. Glycerine Surplus. Chem. Eng. News. 2006, 84(6):7-8 [2] Tyson, S.K., Biodiesel Handling and Use Guidelines. 2001, U.S. Department of Energy. NREL/TP-580-30004: CO. USA. p. 17. [3] da Silva, G.P., Mack, M., Contiero, J., Glycerol: A promising and abundant carbon source for industrial microbiology, Biotechnol Adv 27(1) (2009), 30-39. [4] Yazdani, S.S., Gonzalez, R., Engineering Escherichia coli for the efficient conversion of glycerol to ethanol and co-products, Metab Eng 10(6) (2008) 340-351. [5] Ito, T., Nakashimada, Y., Senba, K., Matsui, T., Nishio, N., Hydrogen and ethanol production from glycerol-containing wastes discharged after biodiesel manufacturing process, J Biosci Bioeng 100 (2005) 260265. [6] Dharmadi, Y., Murarka, A., Gonzalez, R., Anaerobic fermentation of glycerol by Escherichia coli: a new platform for metabolic engineering, Biotechnol Bioeng 94 (2006) 821829.
15
[7] Fukuda, H., Kondo, A., Noda, H., Biodiesel fuel production by transesterification of oils, J Biosci Bioeng 92(5) (2001) 405-416. [8] Shimada, Y., Watanabe, Y., Sugihara, A., Tominaga, Y., Enzymatic alcoholysis for biodiesel fuel production and application of the reaction to oil processing, J. Mol. Catal. B: Enzym. 17(3-5) (2002) 133-142. [9] Shah, S., Sharma, S., Gupta, M.N., Biodiesel preparation by lipase-catalyzed transesterification of jatropha oil, Energ Fuel 18 (2004) 154-159. [10] Fukuda, H., Kondo, A., Noda, H., Biodiesel fuel production by transesterification of oils, J Biosci Bioeng 92(5) (2001) 405-416. [11] Xu, Y., Du, W., Liu, D., Zeng, J., A novel enzymatic route for biodiesel production from renewable oils in a solvent-free medium, Biotechnol Lett25 (2003) 1239-1241. [12] Shimada, Y., Watanabe, Y., Samukawa, T., Sugihara, A., Noda, H., Fukuda, H., Tominaga, Y., Conversion of vegetable oil to biodiesel using immobilized Candida antarctica lipase, J. Am. Oil Chem. Soc. 76 (1999) 789-793. [13] Yagiz, F., Kazan, D., Akin, N., Biodiesel production from waste oils by using lipase immobilized on hydrotalcite and zeolites, Chem. Eng. J. 134 (2007) 262-267. [14] Shimada, Y., Watanabe, Y., Sugihara, A., Tominaga, Y., Enzymatic alcoholysis for biodiesel fuel production and application of the reaction to oil processing, J. Mol. Catal. B: Enzym. 17(3-5) (2002) 133-142. [15] Ott, L., Bicker, M., Vogel, H., 2006. Catalytic dehydration of glycerol in sub- and supercritical water: a new chemical process for acrolein production. Green Chem. 8, 214220. [16] Posada, J.A. and C.A. Cardona, Anlisis de la refinacin de glicerina obtenida como co-producto en la produccin de biodiesel. Ingeniera y Universidad, 2010. 14(1) [17] Pagliario, M. and M. Rossi., The future of Glycerol: New usages for a versatile raw Material. 2008, Cambridge: RSC Publishing.
16
[18] Rahmat, N., A.Z. Abdullah, and A.R. Mohamed, Recent progress on innovative and potential technologies for glycerol transformation into fuel additives: A critical review. Renewable and Sustainable Energy Reviews. 14(3): p. 987-1000. [19] Ito, T., et al., Hydrogen and Ethanol Production from Glycerol-Containing Wastes Discharged after Biodiesel Manufacturing Process. J. Biosci. Bioeng., 2005. 100(3): p. 260-265 [20] Wang, Z., et al., Glycerol production by microbial fermentation: A review. Biotechnology Advances, 2001. 19(3): p. 201-223. [21] Solomon, B.O., et al., Comparison of the energetic efficiencies of hydrogen and oxychemicals formation in Klebsiella pneumoniae and Clostridium butyricum during anaerobic growth on glycerol. Journal of Biotechnology, 1995. 39(2): p. 107-117. [22] Barbirato, F. and A. Bories, Relationship between the physiology of Enterobacter agglomerans CNCM 1210 grown anaerobically on glycerol and the culture conditions. Research in Microbiology. 148(6): p. 475-484. [23] Menzel, K., A.P. Zeng, and W.D. Deckwer, High concentration and productivity of 1,3-propanediol from continuous fermentation of glycerol by Klebsiella pneumoniae. Enzyme and Microbial Technology, 1997. 20(2): p. 82-86. [24] da Silva, G.P., M. Mack, and J. Contiero, Glycerol: A promising and abundant carbon source for industrial microbiology. Biotechnology Advances. 27(1): p. 30-39.
3. Methodology for Processes Design and Analysis

This chapter describes a methodological procedure in order to design and assess technological schemes for the conversion of raw glycerol to added-value products. This methodology uses a strategy based on knowledge which employs both heuristic rules and researchers experience. Also it is equally applied to chemical or biochemical processes, as well as conventional technologies or integrated process. Also, directions are given to perform both steps: the process simulation and the process assessment.
3.1 Processes design

This thesis aims to design and assess technological schemes for the conversion of raw glycerol obtained during the biodiesel production to added-value products. Thus, different possibilities of glycerol conversion to added-value products should be first indentified based on the reported literature. In this way, two main routes for glycerol transformation are available, chemical conversion and fermentative transformation. Glycerol can be chemically transformed by many ways such as: oxidation, hydrogenolysis, etherification, pyrolysis, and gasification. In this sense, many catalysts such as: Pt, Pd, Au, Ru, Cu, Pt, zeolites, and mesoporouses materials, have been widely reported for glycerol conversion. Otherwise, many wild and metabolically engineered strains have been analyzed for the glycerol uptake as substrate in order to produce a wide spectrum of metabolites such us: 1,3-propanediol, ethanol, poly-3-hydroxibitirate, lactic acid, propionic acid, succinic acid, and rhamnolipids.
In order to achieve this objective it was required to: (i) classify all the information available on glycerol transformation by chemical or fermentative routes; (ii) organize these information based on the specific used way for either route, chemical (e.g., oxidation, hydrogenolysis, etherification, pyrolysis, and gasification) or fermentative (e.g., production of: 1,3-propanediol, ethanol, poly-3-hydroxibitirate, lactic acid, propionic acid, succinic
18
acid, and rhamnolipids); (iii) compare and analyze each transformation possibility based on operational criteria such as: conversion, yield, and productivity; and (iv) choose the conversion possibilities with the higher potential to be commercialized. This first stage corresponds to both the literature review and the choosing of the most attractive possibilities for glycerol conversion to added-value components. Moreover, it was found that not only pure glycerol has been used as feedstock to its transformation but also crude glycerol has been widely analyzed. Thus, in order to homogenize the feedstock used for this study, raw glycerol obtained from the biodiesel production process was considered as the unique feedstock. In this way, the influences of several feedstocks used for biodiesel production were analyzed on the composition of the glycerol layer and an average composition for the raw glycerol stream was chose. Due to this raw glycerol stream contains low quantity of glycerol, a purification process was analyzed in order to obtain the three most important qualities of commercially available glycerol. Under this view of point, only one feedstock (i.e., raw glycerol) is always considered and different qualities of glycerol (crude glycerol, technical glycerol and USP glycerol) can be used for its transformation. Additionally, the fact of work with raw glycerol as the unique feedstock, allows considering any designed process as an adjacent biorefinery to the biodiesel production process.
Because of many final products and transformation routes are considered though out this study, each alternative requires both a specific process analysis and a process design using a strategy based on knowledge.
The synthesis of technological schemes by mean of a strategy based on knowledge allows generating systematically alternatives which consider the specific characteristics of each process. Thus, it is possible to design technological configurations of high performance considering mainly techno-economic criteria. Here, the traditional
hierarchical decomposition methodology based on the onion diagram (see Figure 3.1.) for process design is applied [1]. This sequential procedure allows designing and comparing different alternatives for the same objective.
The process design starts analyzing the reaction step which is the fundamental stage in this study, and then the analysis continues to the external layer of the onion diagram adding stages such us: the separation and recycle system according to the Figure 3.1.
3. Methodology
19
This heuristic and hierarchical methodology emphasizes in both the decomposition and analysis of different process alternatives, allowing a quick selection of technological configurations that are often close to the best solution. Furthermore, the nature of this approach, allows discarding many configurations easily which in general do not lead to "good" designs. In addition, tiered design allows the use of process simulators and thus the process diagram can be completed in an evolutionary manner. This methodology has been applied primarily to processes of chemical or petrochemical industry.
Figure 3.1. Hierarchical decomposition according to the "onion diagram"
On the other hand, for the downstream process design the method so called breadth-first was applied in order to analyze different alternatives for the products recovery. This method allows both screening the best alternative for a specific purpose of the downstream process, and evaluating of process alternatives at the next level of hierarchical decomposition (see Figure 3.2).
Most of the alternatives for glycerol conversion to added-value compounds are analyzed during first hierarchical decomposition levels (1 and 2) by mean of the economic potential criteria which also involves operational variables such as conversion and yield. Thus, the alternatives with the highest economic potential are selected to continue the synthesis of technological schemes, while the alternatives showing unfavorable economic potential are
20
discarded. Thus, base structures are obtained for the chosen conversion possibilities which are later complemented through detailed process information for the main stages of processing.
Figure 3.2. The process design method based on the so called breadth-first
The last hierarchical levels of analysis (3-6) require more detailed information for the main process variables which are performed by mean of sensitivity analysis and subsequent rigorous economic evaluation. Details of the processes simulation and economic assessment are given below.
3.2 Processes simulation

Aspen Plus (Aspen Technology, Inc., USA) is the main used tool for defining, structuring, specifying, and simulating the technological schemes for either chemical or biochemical conversion of glycerol to added-value components.
Information required for simulating the most basic technological schemes such as: physical and chemical properties, parameters of design, and operation of processing units, are mainly obtained from secondary sources (e.g., articles, technical reports, databases, patents, among others). Then, the most complex and detailed technological schemes are obtained by mean of rigorous simulations which involve sensitivity analysis and search of optimal operation conditions.
Because of both the petrochemical character of Aspen Plus and its modular-sequential approach, there are not available kinetic models describing the biotechnological processes such as fermentations or enzymatic reactions. Therefore, it is required to work with the available interface between Aspen Plus and Excel. Additionally, the study of
3. Methodology
21
complex fermentation kinetics describing inhibition phenomena, in some cases requires generating more complex calculation routines using other kind of softwares (e.g., MatLab).
Although several possibilities for glycerol conversion to added-value products have been reported, a few publications describe kinetic models fitted good enough. Therefore, the stoichiometric approach is here considered as a completely valid and relevant approach for analyzing the reaction stage of different technological schemes.
On the other hand, specific compounds involved in the different processes of raw glycerol conversion to added-value products such as: free fatty acids, alkyl esters, proteins, salts, cell mass strains, enzymes, and other complex molecules produced by reactive-extractive process are not available on the Aspen Plus Database. Thus, these compounds should be created for each simulation as follows: conventional (by mean of group contribution methods), solids (e.g., biomass), or non-conventional (e.g., enzymes).
All processes are designed and analyzed using the same calculation base which is 1000 Kg/h of raw glycerol always fed to the glycerol purification process. As the simulation results, mass and energy balances are obtained for the technological schemes. Thus, it is possible to obtain requirements of additional raw material, solvents, utility fluids, and energy.
The analysis of conventional separation methods in the distillation process was carried out with the help of the corresponding modules of the process simulators. For this, both short-cut methods and rigorous models available in the simulation package were employed. For simulation of the different technologies involving the operation of distillation, the short-cut method DSTWU incorporated in the package Aspen Plus was applied. This method uses the equations and correlations of Winn-Underwood-Gilliland in order to provide an initial estimation of the minimum number of theoretic stages, minimum reflux ratio, location of the feed stage, and components distribution. The rigorous calculation of the operating conditions in the distillation columns was performed using the module RadFrac based on the equilibrium method that employs the MESH equations (Mass balance equations, phase Equilibrium equations, Summation of the compositions, and Heat balance equations) using the inside-out algorithm. Residue curve maps were
22
used for the conceptual design of the distillation schemes applying the principles of topological thermodynamics (analysis of the statics) [2]. Sensitivity analyses were performed in order to study the effect of the main operating variables (reflux ratio, temperature of the feed stream, ratio between the distillate and the feed, etc.) on the biodiesel purity and the energy consumption of this operation. The final result is the determination of operating conditions that allow developing energetically efficient processes. The objective of this procedure was to generate the mass and energy balances from which the requirements for raw materials, consumables, service fluids and energy needs are calculated.
On the other hand, because of the significant differences involved in the reaction of glycerol to different products, the reaction conditions are specific for each technological scheme. Besides, the downstream process is designed based on the products distribution obtained after the reactive stage. Thus, detailed information about reaction stage and downstream process is given according to each case of study.
Estimation of the energy consumption is performed based on the results of the mass and energy balances generated during the simulation process. Thus, the thermal energy required in the heat exchangers and re-boilers was taken into account, as well as the electric energy needed by pumps, compressors, mills, and other equipments. The energy demand was calculated from the mass and energy balances generated by the simulator. The balances included the energy consumption of reboilers and condensers used in distillation columns, and the energy consumption of the reactors.
3.3 Processes assessment

The capital and operating costs were calculated using the software Aspen Icarus Process Evaluator (Aspen Technologies, Inc., USA). This software estimates the capital costs of process units as well as the operating costs, among other valuable data, utilizing the design information provided by Aspen Plus and the data introduced by the user for specific conditions such as project location among others. Also, analyses are based on the strategy designed by Cardona et al [3-7] for process assessment.
3. Methodology
23
This analysis was estimated in US dollars for a 10-year period at an annual interest rate of 16 %, considering the straight line depreciation method and a 33% income tax [8]. The cost for raw glycerol, crude glycerol, and refined glycerol as well as the labor cost for operatives and supervisors, and the prices for electricity, water and low pressure vapor are showed in the Table 3.1. Additionally, the commercial price for other required compounds such us raw materials and solvents are listed in the Table 3.1.
Table 3.1. Used costs and prices for the economic assessment Costs Operatives Supervisors Electricity Water Low pressure vapor Raw glycerol Crude glycerol (85 wt %) Refined glycerol (98 wt %) Succinic acid Lactic acid Acetic acid Dichloromethane Trioctylamine Methanol 1-Octanol Iso-butylaldehyde 1,3-Propanediol Propionic acid DES Glucose PHB Propionic acid Value 2.14 4.29 0.03044 1.252 8.18 132.45 540.84 706.41 2492.2 1552.2 591.8 850 2550 290 1835 XX 1766 1220 3050 480 3500 1800 Units UDS$/h UDS$/h UDS$/kwh UDS$/m3 UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton UDS$/Ton
24
References
[1] Smith R.M. The nature of chemical process design and integration. Chemical Process: Design and Integration. John Wiley & Sons: Hoboken, NJ, USA, (2005), pp 1-15. [2] Pisarenko Y.A., Serafimov L.A., Cardona C.A., Efremov D.L., Shuwalov A.S., Reactive distillation design: analysis of the process statics. Reviews in Chemical Engineering 17 (4), 2001. [3] Cardona, C.A., Snchez, O. J., Energy consumption analysis of integrated flowsheets for production of fuel ethanol from lignocellulosic biomass. Energy. 2006. 31:2447-2459. [4] Cardona, C.A., Snchez, .J., Fuel ethanol production: Process design trends and integration opportunities. Bioresource Technology. 2007. 98 2415-2457. [5] Quintero, J.A., Montoya, M.I., Snchez, O.J., Giraldo, O.H., Cardona, C.A., Fuel ethanol production from sugarcane and corn: Comparative analysis for a Colombian case. Energy. 2008. 33:385-399. [6] Cardona, C.A., Gutirrez, L.F., Snchez, O. J., In Energy Efficiency Research Advances (D. M. Bergmann, ed.). 2008. Pp:173-212. Nova Science Publishers, Hauppauge, NY, USA. [7] Gutirrez, L.F., Snchez, .J., Cardona, C.A., Process integration possibilities for biodiesel production from palm oil using ethanol obtained from lignocellulosic residues of oil palm industry. Bioresource Technology. 2009. 100:1227-1237. [8] Posada J.A., Cardona C.A., Rincn L.E., Sustainable biodiesel production from palm using in situ produced glycerol and biomass for raw bioethanol. In 32nd symposium on biotechnology for fuels and chemicals. Clearwater Beach, Florida. April 19-22. 2010.
4. Separation and Purification of Glycerol

This chapter presents commercial, technical, and technological aspects related to the glycerol purification process such as the most important commercially qualities of glycerol, the influence of the feedstock used for biodiesel production on the glycerol layer, and the conventional and non-conventional purification processes. The most important qualities of commercial glycerol are: crude glycerol (80-88 wt %), technical glycerol (98 wt %), and refined glycerol (USP or FCC grades, 99.7 wt %). Thus, a flowsheet able to purify raw glycerol up to these three qualities was designed, simulated, and economic assessed. Simulation results showed that is possible to reach the quality requirements while economic results showed that is a profitable process. Also, recovering of anhydrous methanol at 99 wt % could represent an additional incoming for the purification process which could reduce the purification costs among 19 to 26 %. Simulation process is carried out using Aspen Plus software, while the economic evaluation is performed by Aspen Icarus Process Evaluator package.
4.1 Commercial qualities of glycerol

Most of the marketed glycerol is manufactured to satisfy the strict requirements of the United States Pharmacopeia (USP) and the Food Chemicals Codex (FCC) (The Soap and Detergent Association). However, technical grades of glycerol which have not been certified as USP or FCC are also available in the market. The three main qualities of glycerol commercially available depend on their purity, these are: raw glycerol, technical glycerol, and refined glycerol (USP or FCC grade). Raw glycerol usually has between 40 and 88 wt % of glycerol, and contains high amount of methanol, soaps, and salts. This glycerol is commonly obtained as by-product on biodiesel production. Technical glycerol is a high purity product where most of its pollutants have been totally removed. This glycerol is free of methanol, soaps, salts, and other components. Refined glycerol is a pharmaceutical quality product which can be used in foods, personal care, cosmetics,
26
pharmaceutical products, and other special applications. Also, these products must complete the specifications of Pharmacopeia of the USA (USP 30) and the Food and Drug Administration (FDA) of the USA. Table 4.1 shows the main quality specifications and the thresholds for the pollutants present in this glycerol [1].
Table 4.1. Quality specifications for the main qualities of glycerol Properties Raw Glycerol 40-88% 2.0% Max N/A N/A N/A N/A N/A N/A Technical Glycerol 98.0% Min N/A 2.0% Max 10 ppm Max 40 Max (Pt - Co) 1.262 (@25C) N/A N/A Refined Glycerol (USP) 99.70% N/A 0.3% Mx. 10 ppm Mx. 10 Max. (APHA) 1.2612 Min 20 ppm Mx 99.0 - 101.0% (dry base) 5 ppm Mx. 30 ppm Mx.
Glycerol Content Ash Moisture Chlorides Color Specific Gravity Sulfate Analysis
Heavy Metals Chlorates Components Ignition Residues Fatty acids and Esters Water pH (solution 10%) Organic Residues
N/A N/A
5 ppm Mx. 30 ppm Mx.
N/A N/A
N/A 1.00 Mx
100 ppm Mx. 1000 Mx
12.0% Max 4.0 - 9.0 2.0% Mx
5.0% Mx 4.0 - 9.1 2.0% Mx
0.5% Mx N/A N/A
4. Glycerol Purification
27
4.2 Effect of the feedstock for biodiesel production on glycerol composition

Raw glycerol has a very low value in the market because of its impurities. Also, composition of glycerol highly depends on both the family of used raw material and the process conditions for biodiesel production. This fact occurs because the chemical compositions of the feedstocks used for biodiesel production could change significantly. Fats and oils usually contain more than ten types of fatty acids, which have between 12 and 22 carbons. But, often the higher proportion of fatty acids has between 16 and 18 carbons. Although these fatty acids are saturated, monounsaturated or polyunsaturated [2], different degrees of saturation affect the properties of the biodiesel fuel. Thus, a "perfect" biodiesel should be only obtained from monounsaturated fatty acids.
The composition profile of fatty acids was presented by He and Thompson [3] for six vegetable oils (i.e., IdaGold mustard, PacGold mustard, rapeseed, canola, crambe, and soybean) and for waste vegetable oil (WVO) used as feedstocks on biodiesel production as shown in Table 4.2. Additionally, based on the reported information by He and Thompson [3], the composition of glycerol layer was calculated for each used feedstock for biodiesel production. The results are shown in Table 4.3.
Table 4.2. Fatty acid profile of vegetable and used oils [3] Composition (wt %) Fatty acids Palmitic (16:0) Estearic (18:0) Oleic (18:1) Linoleic (18:2) Linolenic (18:3) Eicosic (20:1) Erucic (22:1) MW Average (Kg/Kmol) IdaGold PacGold Rape Canola Soybean Crambe WVO 2,8 1 24,8 10,3 9,4 10,7 34,7 946,3 3,1 1,6 23,9 21,6 9,9 12,1 22,1 924,6 2,9 1 13,7 11,8 7,5 8,7 48,5 4,5 1,8 60,7 19,1 9,5 1,8 0,9 10,7 4,3 24,9 51,6 7,3 0,2 --872,8 2 0,9 17,9 8,1 4,5 3,7 54,1 978,5 18,7 6,3 40,5 28 1,5 ----867,2
968,5 882,1
28
Table 4.3. Composition of the glycerol layer obtained by decantation during the biodiesel production from different feedstocks Oil Component Methanol (wt %) Glycerol (wt %) NaOCH 3 (wt %) Proteins (wt %) Fats (wt %): Palmitic (16:0) Estearic (18:0) Oleic (18:1) Linoleic (18:2) Linolenic (18:3) Eicosic (20:1) Erucic (22:1) Ash (wt %) IdaGold PacGold Colza Canola Soja 32,59 60,05 2,62 0,13 1,94 0,054 0,019 0,480 0,200 0,182 0,207 0,672 2,67 32,68 61,39 2,82 0,18 1,08 0,030 0,011 0,269 0,112 0,102 0,116 0,376 1,85 28,20 25,07 59,94 60,38 2,27 0,06 8,88 2,24 0,05 11,68 Crambe WVO 11,72 46,41 1,99 0,14 36,41 1,020 0,364 9,030 3,750 3,423 3,896 12,635 3,33
26,06 23,17 61,67 65,01 2,56 0,05 7,17 2,69 0,46 8,42
0,249 0,327 0,089 0,117 2,203 2,896 0,915 1,203 0,835 1,098 0,951 1,250 3,082 4,052 0,64 0,58
0,201 0,236 0,072 0,084 1,779 2,087 0,739 0,867 0,674 0,791 0,767 0,901 2,489 2,921 2,48 0,26
4.3 Conventional purification process

At laboratory scale the purification of the system containing biodiesel, glycerol, soaps, and salts (mainly sodium methoxide, NaOCH3), is preformed using separation funnels, which allow to soaps being remained in the crude glycerol layer. Layer containing esters must be heated up to 85 C in order to recover the unreacted methanol. While industrially raw glycerol is refined through a filtration process, followed by mixing with chemical additives which allow the precipitation of salts and finally different qualities of commercial glycerol are obtained by a vacuum fractional distillation process.
29
Distillation is the most commonly used method for glycerol purification. This technology produces high purity glycerol at high yields. However, the glycerol distillation is an energy intensive process because of its high heat capacity, requiring a high supply of energy for vaporization [4]. Ion exchange has also been used to purify raw glycerol [5], but this technique is not economically viable from an industrial view of point due to the high content of salts. Also, when contents of sales are above 5 wt % which is tipically found in the glycerol stream obtained from the biodiesel industry, the chemical regeneration cost of these resins becomes very high. Figure 4.1 shows the flow diagram for the two above described conventional techniques for glycerol purification.
Purification by vaccum distillation
P-7
Concentration by evaporation Filtration
Purification by ionic exchange
Evaportation and refining of crude glycerol

Figure 4.1. Flowsheet of conventional schemes for glycerol purification.
4.4 Alternative purification processes

A commercially available technology for raw glycerol purification obtained during the biodiesel production was jointly developed by Rohm and Haas, a provider of functional polymers by ion exchange technologies and catalysts, and by Novasep Process, a supplier of purification solutions which includes chromatography, ion exchange, membranes, crystallization, and evaporation. The process is the so called Ambersep BD50 [6]. In principle, this process uses a chromatographic separator in order to remove large amount of salts and free fatty acids. Refined stream is then processed in an evaporator / crystallizer unit, which removes the salts in a crystalline form. This fact
30
avoids the effluents production in the glycerol purification plant. Thus, a glycerol stream at a purity of 99.5 wt % is obtained. But if a high quality glycerol is required, (e.g., 5 to 10 parts per million of salt content) it is possible to use a ion exchange demineralization unit. This process has lower energy requirements compared to the traditional distillation process. The block diagram for the Ambersep BD50 process is shown in Figure 4.2 which illustrates the different steps for the raw glycerol purification process.
Crude
Pre-Heating Heating up to 90C Filtration Heating up to 90C Degasification Filtration Refined Crystallization NaCl Effluent Light Water Secondary Glycerol
NaCl
Degasification CHR
Cooling at 40C
(Optional)
IEX (Optional) Concentration
Refined Glycerol
Figure 4.2. Flowsheet of the Ambersep BD50 process
31
4.5 Simulation of the glycerol purification processes

Based on the calculated compositions for the glycerol layer obtained from different feedstocks (see Table 4.3.), the profile compositions obtained from IdaGold mustard represents the average values among the first use oils analyzed. Thus, this stream was chosen to design the purification process of the raw glycerol.
Figure 4.3.a. shows the simplified flowsheeet for raw glycerol purification to 88 wt % (crude glycerol) and to 98 wt % (technique glycerol). In order to obtain glycerol at 99.7 wt % (glycerol USP grade), it is required a further refining process throughout a ion exchange resin which removes the triglycerides still contained in the mixture, as shown in Figure 4.3.b.
Water waste 1
Methanol
1
Raw Glycerol
Water
Water waste 2
RII-1
Glicerina USP
Solids
Organic Phase
Aqueous Glycerol
Glycerol
Adsorbato
Figure 4.3. Simplified flowsheet for raw glycerol purification. a) purification at 88 and 98 wt %. b) Purification at 99.7 wt %. 1. First evaporation column, 2. Neutralization tank, 3. Centrifuge, 4. Decantation tank, 5. Second evaporation column, 6. Distillation column. 7: Ionic exchange resine.
The raw glycerol stream is initially evaporated, where 90 % of methanol at 99 wt % is recovered. Also, since glycerol is the unique impurity present in the recovered stream, this steam of anhydrous methanol is appropriate to be reused in the transesterification process. Bottom stream obtained from Evaporator I is neutralized using an acid solution. Then both salts produced during the neutralization process and remaining ashes and proteins are retired by centrifugation. The clarified product obtained from the centrifuge is washed with water using a weight ratio of 2.4 (water/glycerol stream). Thus, 50% of the triglycerides remaining in the mixture are withdrawn with a glycerol lost of 1.8 %. The resulting aqueous glycerol stream and free of salts, solids, and protein but with a low content of both methanol and triglycerides, is subjected again to an evaporation process which removes more than 90% of water and most of the remaining methanol, with a
32
glycerol lost of 0.2%.Thus, the glycerol purity reached is 80 wt %. Then, the glycerol stream is purified through a distillation column to reach the required purity, either 88 wt % or 98 wt %. Although in all cases the used flowsheet is the same, the operational conditions change depending on the required purity.
In order to obtain glycerol at USP grade, the process conditions adjusted for glycerol at 98 wt % are in general preserved, but both the reflux ratio and the ratio of distillate/feed are increased for the distillation tower. Also, a final refinement stage through an ion exchange resin is required to remove 95 % of the triglycerides contained still in the mixture. Table 4.4 summarizes the simulation results obtained for the purification processes of raw glycerol. Also, it can be observed that the obtained products meet the quality requirements shown in Table 4.1.
Table 4.4. Simulation results for raw glycerol purification process. Streams Variable Raw Glycerol 25 973,3 Mass flow free of ash (kg/hr) Mass fraction: Triglycerides Methanol Water Glycerol NaOCH 3 Protein Mass flow of ash (kg/hr) 0,02 0,335 0 0,617 0,027 0,001 27,6 0 0,99 0 0,01 0 0 0 0,105 0,88 4,3 ppm 0,2 ppm 0,003 0,004 0,98 4,3 ppm 0,2 ppm 0,003 0,002 0,997 1,62 ppm 75 ppb 0,00015 0,015 0,014 0,016 0,001 Methanol Glycerol at 88% 104,7 665,25 Glycerol at 98% 189,2 596,595 Glycerol at 99,7% 204 586,179
Temperature (C)
144,2 301,981
33
4.6 Economical assessment for glycerol purification processes

Because of the fed glycerol stream contains 32.6 wt % of methanol it is required to consider two different scenarios. The first one considers that the withdrawn methanol is not recovered while the second scenario considers that the withdrawn methanol from the raw glycerol stream is recycled and reused as feedstock during the transesterification process since this stream is composed of 99 wt % methanol and 1 % glycerol. Thus, under the light of the second scenario, methanol is considered as a by-product stream which has an economical value. Economic assessment results for raw glycerol purification to 88, 98, and 99.7 wt % are shown in Table 4.5. The purification costs (PC) for each purification process are in the first column discriminated by raw materials, utilities, operating labor, maintenance and operating charges, plant overhead, general and administrative costs, capital depreciation, and co-products credit. The second column contains the share of each item by each purification process.
Table 4.5. Purification costs (PC) of raw glycerol (US$/L) Item (US$/L) Raw materials Utilities Operating labor Maintenance Operating charges General costs Administrative costs Capital depreciation Co-products credit CT without sale of metanol CT with sale of metanol CP Glycerol % CP CP Glycerol % CP CP Glycerol (88%) 88 wt % 98 wt % (98%) 99.7 wt % 0.05539 0.03741 0.01889 0.00721 0.00472 0.01305 0.01093 0.07595 0.05900 0.22356 0.16456 24.78 16.73 8.45 3.22 2.11 5.84 4.89 33.97 0.05539 0.07290 0.01889 0.00793 0.00520 0.01436 0.01179 0.07595 23.55 31.00 8.03 3.37 2.21 6.11 5.01 32.30 0.05539 0.13544 0.02173 0.00979 0.00543 0.01576 0.01257 0.08983 % CP (99.7%) 23.11 56.51 9.07 4.08 2.27 6.57 5.24 37.48 -19.34
-26.39 0.06574 0.26241 0.19666
-25.05 0.06691 0.34593 0.27902
34
Due to the low commercial prices of raw glycerol obtained from the biodiesel production, raw materials represent less than 25% of the total purification cost. The capital cost accounts the most share of the purification cost, being it about 35%. Moreover, an increase in the glycerol purity represents an increase in the utilities cost, reaching up to 56.5% of total purification cost when glycerol at 99.7% is obtained. On the other hand, the recovery of anhydrous methanol at 99 wt % represents a significant reduction in the total purification cost, with a decreasing between 19 26 % of the total costs.
The purification cost of raw glycerol obtained from biodiesel production was reported by Johnson and Taconi [7] as 0.15 USD$/lb or 0.26 USD$/L. This value is close to the total purification cost obtained for glycerol at 98% at the scenario that no considers the sale of anhydrous methanol as co-product. This scenario is the most standard analyzed since it is a technical quality of glycerol with no by-products production.
Commercial sale prices for different qualities of glycerol are as follows: 0.28 USD$/L for glycerol at 88 wt %, 1.39 USD$/L for vegetable glycerol at 98 wt %, 1.11 USD$/L for tallow glycerol at 98 wt % and 3.48 USD$/L for glycerol USP grade or at 99.7 wt %. For the assessed production scale, the purification and refining of raw glycerol is profitable since the purification costs are lower than their selling prices.
4.7 Conclusions
Commercially three qualities of glycerol were identified as the most important ones. Crude glycerol with a purity ranging from 80-88 wt %, technical glycerol mainly found at 97 wt %, and refined glycerol (USP or FCC grades) at 99.7 wt %. These three types of glycerol differ significantly in the content of water, fatty acid residues, esters, and other organic wastes. Also, some differences were found for the use of diverse feedstocks for biodiesel production on the composition of the glycerol layer. Although, most of the first use oils lead to not big differences in the glycerol layer, a completely different behavior was observed for the glycerol obtained from WVO represented by low concentration of glycerol and methanol with a high content of fats. On the other hand, based on the traditional purification of glycerol, a flowsheet able to purify raw glycerol up to the three commercial qualities above described was designed, simulated and economically assessed. Results showed that not only quality requirements were successfully obtained
35
but also for the analyzed purification scale all the processes were profitable. Thus, a homogenized raw material and purification process was obtained in order to continue the analysis of different possibilities of glycerol transformation to added-value products.
References
[1] Posada, J.A. and C.A. Cardona, Anlisis de la refinacin de glicerina obtenida como co-producto en la produccin de biodiesel. Ingeniera y Universidad, 2010. 14(1) [2] Beln-Camacho, D.R, Snchez, E.D., Garca, D., Moreno-lvarez, M.J., Linares O. Caractersticas fisicoqumicas y composicin en cidos grasos del aceite extrado de semillas de tomate de rbol (Cyphomandra betacea Sendt) variedades roja y amarilla. Grasas y Aceites. 2004, 55(4):428-433 [3] Thompson, J.C., He, B.B., Characterization of crude glycerol from biodiesel production for multiple feedstocks. Appl. Eng. Agric. 2006, 22(2):261-265. [4] Posada, J.A., Cardona, C.A., Rincn L.E., Sustainable biodiesel production from palm using in situ produced glycerol and biomass for raw bioethanol. En: Society for Industrial Microbiology. 32nd symposium on biotechnology for fuels and chemicals. Clearwater Beach, Florida. April 19-22. 2010. [5] Berriosa, M., Skelton, R.L., Comparison of purification methods for biodiesel. Chem. Eng. J. 2008, 144:459-465. [6] AMBERSEP BD50 Technology. <www.amberlyst.com/glycerol.htm> [Consulted: 2909-2009]. [7] Johnson, D.T., Taconi, K.A., The Glycerin Glut: Options for the Value-Added Conversion of Crude Glycerol Resulting from Biodiesel Production. Environ. Prog. 2007, 26(4):338-348.
5. Chemical Conversion of Glycerol

This chapter presents different ways of glycerol transformation to added-value products. Different reactions are described, such as: (i) oxidation on metallic catalysts like Pt, Pd, Au, and on promoters as Bi and Pb; (ii) hydrogenolysis to glycols on Ru, Cu and Pt catalysts; (iii) etherification to polyglycerols on zeolites and mesoporous materials; (iv) pyrolysis and gasification, where the objective is to produce syn-gas.
Glycerol is a potentially important feedstock for biorefineries, available as a byproduct in the biodiesel production by transesterification of vegetable oils or animal fats. Also, due to its high functionality, there are many transformation ways to produce added-value compounds using glycerol as sole feedstock for its conversion. On the other hand, new uses for glycerol need to be found since the biodiesel production cost vary inversely with the glycerol cost.
The high differences between the price of raw glycerol and refined glycerol, added to its chemical versatility have carried out an intense research for developing alternative uses and practical technologies to utilize the raw glycerol. In this sense chemical possibilities of glycerol transformation are reviewed as follows. Thus, several transformation possibilities to added value products have been found by chemical or biochemical ways.
5.1 Oxidation
Glycerol oxidation on metallic catalysts is carried out by mean of oxidative dehydrogenation mechanism on the metal surface [1]. First step is the alcohol dehydrogenation, followed by the oxidation of intermediate formed [2]. Due to the potential complexity of products distribution (Figure 5.1) the selectivity control on the process oxidation is key [3].
38
The main derived oxygenated products from glycerol (GLY), are: glyceric acid (GLYAC), dihydroxyacetone (DHA), hydroxypyruvic acid (HYPAC), tartaric acid (TARAC), mesoxalic acid (MESOXAC), oxalic acid (OXALAC), besides some intermediates as glyceraldehyde (GLYAL), glycolic acid (GLYCAC), and glyoxylic acid (GLYOXAC) as is shown in the Figure. 5.1.
Figure 5.1. Possible products for glycerol oxidation
The most studied metallic catalysts are palladium (Pd), platinum (Pt), and gold (Au), although the main disadvantage of Pd and Pt are their deactivation with the reaction time increment [2]. To improve the activity, selectivity, and stability of the reactive system, promoters are used on Pt and Au for redox reactions; there are particularly heavy metals from groups IV (lead, Pb) and V (bismuth, Bi) [4].This fact allows preventing the products over-oxidation on the metal surface, avoiding the products degradation until total oxidation to carbon dioxide, also promoters favors the secondary alcohols oxidation. Primary alcohols are oxidized to carboxylic acids (GLYAC, TARAC and HYPAC via DHA) in absence of promoters or under basic pH, and secondary alcohols are selectively oxidized on Pt-Bi metallic catalyst at acid pH (DHA, HYPAC via GLYAC and MESAC) [1-2].
Gallezot et al. [1, 5-8], Hutchings et al [9-11], Prati et al [12-18], Claus et al [2, 19] and Davis [20-22] have studied the selective glycerol oxidation on mono - or bi - metallic catalysts of Pd, Pt, and Au, using oxygen as oxidizer agent. Gallezot et al showed that
5. Chemical Conversion
39
GLYAC and TARAC are obtained under basic pH, while HYPAC is obtained under not very acid pH via DHA and, DHA and HYPAC are obtained under acid pH via GLYAC and MESAC [4, 6, 8, 23]. Total glycerol conversion is achieved for Pd and Pt catalysts with selectivities of 70% and 35% to GLYAC and HYPAC respectively. Also, for Pt-Bi catalyst selectivities of 83%, 74%, 37%, and 39% to TARAC, HYPAC, DHA, and MESOXAC are obtained respectively, with conversions upper to 75%, except for MESOXAC which was 53%. On the other hand, proofs carried out with activated coal (AC) as support showed that 5% Pd/CA catalyst has higher potential redox than 5% Pt/CA [4], and for Au catalyst was found that activity and selectivity increase when the particle diameter diminishes. Hutchings et al [3, 9-11] and Prati et al [12-18, 24] studied the glycerol reaction on Au catalysts. Au supported on carbon (Au/C) is extremely selective to GLYAC (>82%), with conversion higher to 60% [3]. Also, in systems at basic pH the selectivity to GLYAC is increases with both, pH and oxygen pressure. Bi metallic catalysts of Pd-Au take to total GLY conversion with high GLYAC selectivity (>45%), which was increased when bimetallic catalysts were immobilized on graphite.
GLY oxidation by Au catalysts supported on graphite, activated coal, and carbon nanoparticles was studied by Claus et al [2, 19] , who found that the last one support is the most chemically active, also confirmed the dependence among selectivity to GLYAC and particle size. Other mono- and bi-metallic nanoparticles of Au-Pd were evaluated by Davis et al [20-22] for GLY oxidation in liquid phase. The highest turnover frequency (TOF) was exhibits for the Au mono-metallic catalyst and the highest selectivity to GLYAC was reached by Pd. Also, activity and selectivity for bimetallic catalyst Au-Pd was dependent of the Au quantity.
5.2 Reduction
Glycerol reduction produces mainly 1,2- propileneglycol (12-PG), 1,3-propileneglycol (13PG), ethyleneglycol (ETGLY), and other by-products such as lactic acid (LACAC), acetol (ACET), acroleine (ACRO), besides degradation products such as propanol (PROPOH) , methanol (METOH), methane (MET), and carbon dioxide (CO2) [25]. Among glycerol reduction products the most important is propyleneglycol because of its high functionality which can be used in unsaturated polyester resins, functional fluids (antifreezes and heat
40
transfer), pharmaceuticals, foods, cosmetics, liquid detergents, tobacco humectants, flavors and fragrances, personal care, paints, and animal feed.
Several technological schemes for propyleneglycol production from glycerol have been patented [26-29] in which are used different catalysts such as copper, zinc, ruthenium, cobalt, magnesium, molybdenum, nickel, palladium and platinum; under a widely operation conditions for pressure (2000 5000 psi) and temperature (200 350 C). On the other hand Shanks and Lahr [30, 31] studied the interactions among reactants and catalyst for dehydrogenation/hydrogenation process with Ru supported on activated carbon (5% wt Ru/CA), thus pH effects, competitive adsorption, and products degradation (ethylene glycol and propylene glycol) were analyzed under high pressure, meddle temperature, and high glycerol concentration (1450 psi, 205 C and 10 wt % of glycerol). Due to the high catalytic activity of Ru was found that ethyleneglycol and propyleneglycerol degradation rate is independent of the initial glycol concentration, although propylene glycol is less competitive than ethylene glycol to active sites. Also, while selectivity to propyleneglycol is independent of pH, selectivity to ethyleneglycol increases at low basic conditions.
Lahr and Shanks [30] purposed a model for glycerol reduction, in which glycerol is adsorbed and dehydrogenated reversibly on the metallic catalyst where glyceraldehyde is formed, which is then desorbed and could react through four different way in a basic media: (i) retro-aldol mechanism to produce glycol aldehyde as precursor of ethylene glycol; (ii) oxidation and subsequent descarboxylation to produce also glycol aldehyde; (iii) dehydration to 2- hydroxypropionaldehyde as precursor of propylene glycol; or (iv) degradation to unwanted side products which is also a possible way to produce the glycols precursors. Finally, the respective precursors are hydrogenated to glycols. A new mechanism for glycerol reduction under moderate operation conditions was proposed by Dasari et al [25] where hydroxyacetone is formed by dehydrogenation of glycerol, which after react with hydrogen to produce propylene glycol and water.
The highest selectivities to propylene glycol have been reported for Cu-based catalysts which exhibit low selectivities to ethylene glycol and other degradation by-products [25]. While Ru- and Pd- based catalysts have low selectivities to propylene glycol (< 50% generally) because of competitive hydrogenolysis where C-C and C-O bonds are taken to
41
an excessive degradation to produce lower alcohols and gases [25, 32]. In general terms glycerol conversion is significantly increased by temperature, while yield has a maximum near to 200C due to the degradation products which occurs at high temperature. Also, propyleneglycol selectivity can be improved increasing the water contend in the glycerol mixture which reduce the glycerol conversion, but however the net yield increase.
5.3 Etherification
Glycerol etherification takes to polyglycerols which are oxygenated compounds used as surfactants, lubricants, cosmetics, and food preservatives. Polyglycerols have low polymerization level, and these can be obtained in lineal, cyclic, or branched chains, but researching effort s are focused on selective production of di- and/or tri-glycerols. Selectivity of glycerol etherification is similar to pseudo-polymerization where generally a mixture of lineal and cyclic polyglycerols is obtained, especially in presence of homogeneous catalysts such as sodium, potassium or carbonate hydroxide [32-34].
Etherification selectivity in the first reaction step on acid catalysts is not really controlled and a mixture of di- to hexa- glycerols (lineal or cyclic), esters of polyglycerol, and acroleine as by-products is obtained. Although selectivity in the first step could be slightly improved modifying the pseudo-pore size in the mesoporous materials [35]. On the other hand, glycerol conversion was improved by Na2CO3, although low selectivities to di- and tri-glycerols were as obtained. Then alkaline exchange zeolites were studied and selectivity was increased [32]. Incorporation on mesoporous catalytic structure of elements such as Al, Mg, and La, modifying only the activity, and selectivity is hold almost constant. Clacens et al [35] found that impregnation method takes to materials most stable and selective than incorporation method. Among the impregnated materials La is the most active but its selectivity was the worst; a positive behavior was found to Mg which is highly selective
5.4 Pyrolysis and gasification

Pyrolysis process produces liquid fuels at temperatures between 400 600 C, and gas products at temperatures upper to 750 C. Although gasification process is similar to
42
pyrolysis the main difference is that gasification is carried out in presence of oxygen like: air, or pure oxygen, or vapor.
Reactions catalyzed by protons or hydroxyl ions can be performed under almost- or super-critical water conditions (P> 22.1 MPa and T> 647 K) because of water is not only a solvent it is also a catalyst due to the self-dissociation which takes to formation of hydroxyl ions and protons. Under these conditions two competitive ways have been identified. The first one consists in a series of ionic reactions which occur at high pressure and/or low temperature. The second is a degradation reaction of free radicals, which occurs at low pressure and/or high temperature. On the other hand, temperature increases the reaction rate until critical temperature is reached, then reaction rate decreases drastically related to subcritical conditions.
The main products of glycerol degradation are: methanol, acetaldehyde, propionaldehyde, acroleine, allylic alcohol, ethanol, formaldehyde, carbon monoxide, carbon dioxide, and hydrogen. Acetaldehyde and formaldehyde formation increase with the pressure which indicates that these compounds are mainly formed by the ionic reaction, while methanol and allylic alcohol formation decrease with the pressure which indicates that these compounds are formed by the free radicals way [36]. Formation of gasses products happens to high temperature; also gases formation decrease with the pressure, this indicates a production by a reaction mechanism of free radicals. Gases formation occurs at high temperature; also gases formation decrease with the pressure which indicates that these are produced by a free radicals mechanism.
Syngas is the main product of pyrolysis and gasification processes. Syngas is a mixture of hydrogen (H2) and carbon monoxide (CO). A wide range of conversions and selectivities have been reported depending on the operational conditions such as temperature, pressure, and glycerol concentration [36-41], also the pollutants presence for raw glycerol, such as methanol and KOH [41]. Low glycerol concentration and high temperature takes to high CO2 concentrations in the gas product and the most products remain in the liquid phase [42]. On the other hand when temperature is increased the H2 and CO2 production are improved, which takes to a low CO concentration in the gas product.
43
Nitrogen (N2) is used like carrier gas for glycerol pyrolysis. High amounts of N2 takes to high liquid phase yield and gas production diminished. This process has a yield of 93% to syngas (H2 + CO) at 800 C like showed Valliyappan [41]. On the other hand gasification is carried out with vapor without any carried. Total glycerol conversion was reported for a initial mixture of 50 wt % of glycerol by Valliyappan from pure and raw glycerol [41].
References
[1] Gallezot, P., Selective oxidation with air on metal catalysts. Catalysis Today, 1997. 37(4): p. 405-418. [2] Demirel-Glen, S., M. Lucas, and P. Claus, Liquid phase oxidation of glycerol over carbon supported gold catalysts. Catalysis Today, 2005. 102-103: p. 166-172. [3] Hutchings, G.J., Catalysis by gold. Catalysis Today, 2005. 100(1-2): p. 55-61. [4] Garcia, R., M. Besson, and P. Gallezot, Chemoselective catalytic oxidation of glycerol with air on platinum metals. Applied Catalysis A: General, 1995. 127(1-2): p. 165-176. [5] Fordham, P., et al., Selective catalytic oxidation with air of glycerol and oxygenated derivatives on platinum metals, in Studies in Surface Science and Catalysis. 1996, Elsevier. p. 161-170. [6] Fordham, P., et al., Selective oxidation with air of glyceric to hydroxypyruvic acid and tartronic to mesoxalic acid on PtBi/C catalysts, in Studies in Surface Science and Catalysis. 1997, Elsevier. p. 429-436. [7] Fordham, P., M. Besson, and P. Gallezot, Selective oxidation with air of glyceric to hydroxypyruvic acid and tartronic to mesoxalic acid on PtBi/C catalysts. Stud. Surf. Sci. Catal., 1997. 108: p. 429-436. [8] Fordham, P., M. Besson, and P. Gallezot, Selective Catalytic Oxidation of Glyceric Acid to Tartronic and Hydroxypyruvic Acids. Applied Catalysis A: General, 1995. 133: p. L179-L184. [9] Carrettin, S., et al., Selective oxidation of glycerol to glyceric acid using a gold catalyst in aqueous sodium hydroxide. Chem. Commun., 2002. 7: p. 696-697.
44
[10] Carrettin, S., et al., Oxidation of glycerol using supported gold catalysts. Top. Catal., 2004. 27: p. 131-136. [11] Carrettin, S., et al., Oxidation of glycerol using supported Pt, Pd and Au catalysts. Phys. Chem. Chem. Phys., 2003. 5: p. 1329-1336. [12] Bianchi, C.L., et al., Selective oxidation of glycerol with oxygen using mono and bimetallic catalysts based on Au, Pd and Pt metals Catalysis Today, 2005. 102-103: p. 203-212. [13] Dimitratos, N., et al., Effect of particle size on monometallic and bimetallic (Au, Pd)/C on the liquid phase oxidation of glicerol. Catal. Lett., 2006. 108: p. 147. [14] Dimitratos, N., F. Porta, and L. Prati, Au, Pd (Mono and Bimetallic) Catalysts Supported on Graphite Using the Immobilization Method Synthesis and Catalytic Testing for Liquid Phase Oxidation of Glycerol. Applied Catalysis A: General, 2005. 291: p. 210214. [15] N. Dimitratos, A.V., D. Wang, F. Porta, D. Su, L. Prati,, Pd and Pt catalysts modified by alloying with Au in the selective oxidation of alcohols. Journal of Catalysis, 2005. 244(1): p. 113-121. [16] Prati, L. and F. Porta, Oxidation of alcohols and sugars using Au/C catalysts - Part 1. Alcohols. Appl. Catal., A., 2005. 291(199-203). [17] Prati, L. and M. Rossi, Chemoselective catalytic oxidation of polyols with dioxygen on gold supported catalysts. Stud. Surf. Sci. Catal., 1997. 110: p. 509-516. [18] Wang, D., et al., Single-phase bimetallic system for the selective oxidation of glycerol to glycerate. Chem. Commun., 2006. 18: p. 1956-1958. [19] Demirel, S., et al., Use of Renewables for the Production of Chemicals: Glycerol Oxidation over Carbon Supported Gold Catalysts. Appl. Catal. B: Environmental, 2007. 70: p. 637-643. [20] Ketchie, W.C., et al., Influence of gold particle size on the aqueous-phase oxidation of carbon monoxide and glycerol. Journal of Catalysis, 2007. 250: p. 94-101.
45
[21] Ketchie, W.C., M. Murayama, and R.J. Davis, Selective oxidation of glycerol over carbon-supported AuPd catalysts. Journal of Catalysis. , 2007. 250: p. 264-273. [22] Ketchie, W.C., M. Murayama, and R.J. Davis, Promotional Effect of Hydroxyl on the Aqueous Phase Oxidation of Carbon Monoxide and Glycerol over Supported Au Catalysts. Topics in Catalysis, 2007. 44: p. 307-317. [23] Fordham, P., M. Besson, and P. Gallezot, Selective oxidation with air of glyceric to hydroxypyruvic acid and tartronic to mesoxalic acid on PtBi/C catalysts. Stud. Surf. Sci. Catal. , 1997. 108: p. 429-436. [24] Porta, F. and L. Prati, Selective oxidation of glycerol to sodium glycerate with goldon-carbon catalyst: an insight into reaction selectivity. Journal of Catalysis, 2004. 224(2): p. 397-403. [25] Dasari, M.A., et al., Low-Pressure Hydrogenolysis of Glycerol to Propylene Glycol. Applied Catalysis A: General, 2005. 281: p. 225-231. [26] Casale, B. and A.M. Gomez, Method of hydrogenating glycerol. 1993: U.S. Patent N. 5,214,219. [27] Casale, B. and A.M. Gomez, Catalytic method of hydrogenating glycerol. 1994: U.S. Patent N. 5,276,181. [28] S. Ludwig, E.M., Preparation of 1,2-propaned. 1997: U.S. Patent N. 5,616,817. [29] Tessie, C., Production of propanediols. 1987: U.S. Patent N. 4,642,394. [30] D. G. Lahr, B.H.S., Kinetic Analysis of the Hydrogenolysis of Lower Polyhydric Alcohols: Glycerol to Glycols. Ind. Eng. Chem. Res. , 2003. 42: p. 5467-5472. [31] Lahr, D.G. and B.H. Shanks, Effect of Sulfur and Temperature on RutheniumCatalyzed Glycerol Hydrogenolysis to Glycols. Journal of Catalysis, 2005. 232: p. 386394. [32] Barrault, J., et al., Catalysis and Fine Chemistry. Catalysis Today, 2002. 75: p. 177181.
46
[33] Gerald, J., S. Werner, and D. Helmut, Process for the Preparation of Diglycerol and/or Polyglycerol, S.W. GMBH, Editor. 1993: U.S. Patent N. 5 243 086. [34] Lutz, J., et al., Process for the production of diglycerol, H. KGAA, Editor. 1998: U.S. Patent N. 5 710 350. [35] Clacens, J.M., Y. Pouilloux, and J. Barraultn, Selective Etherification of Glycerol to Polyglycerols over Impregnated Basic MCM-41 type Mesoporous Catalysts. Applied Catalysis A: General, 2002. 227: p. 181-190. [36] W. Buhler, E.D., H.J. Ederer, A. Kruse, C. Mas, , Ionic Reactions and Pyrolysis of Glycerol as Competing Reaction Pathways In Near- and Supercritical Water. Journal of Supercritical Fluids, 2002. 22: p. 37-53. [37] Matsumura, Y., et al., Biomass Gasification in Near- and Super-Critical Water: Status and Prospects (Review). Biomass and Bioenergy, 2005. 29: p. 269-292. [38] Mozaffarian, M., E.P. Deurwaarder, and S.R.A. Kersten, ECN-C--04-081 Green Gas (SNG) Production by Supercritical Gasification of Biomass. November 2004. [39] Antal, M.J., et al., Biomass Gasification in Supercritical Water. Ind. Eng. Chem. Res., 2000. 39: p. 4040-4053. [40] Xu, X., et al., Carbon-Catalysed Gasification of Organic Feedstocks in
Supercritical Water. Ind. Eng. Chem. Res., 1996. 35: p. 2522-2530. [41] Valliyappan, T., Hydrogen or Syn Gas Production from Glycerol Using Pyrolysis and Steam Gasification Processes, in Department of Chemical Engineering. 2004, University of Saskatchewan Saskatoon, Saskatchewan [42] Van Swaaij, W., Technical Feasibility of Biomass Gasification in Fluidized Bed with Supercritical Water. 2003.
6. Biochemical conversion of glycerol

This chapter studies different possibilities for glycerol bioconversion to added value products: 1,3-propanediol, ethanol, poly-3-hydroxybutirate, lactic acid, succinic acid, propionic acid, and rhamnolipids. Also, the influence of the main process variables on the fermentation behavior (conversion, selectivity, and products distribution) is discussed.
6.1 1,3-propanediol
In the early of 90s a biotechnological route which uses glycerol to produce 1,3propanediol by mean a fermentation process was developed [1]. 1,3-propanediol is a commercially important compound because it can be used as adhesive, antifreeze, cosmetics moisturizing, stabilizing detergents, and as additive for painting, printing inks, and high pressure lubricants. Also it can be used as monomer for polyesters synthesis such as polytrimethylene terephthalate (PTT) and polyethylene terephthalate (PET) which can improve the chemical and mechanical properties in comparison with other similar monomers.
Fermentative production of 1,3-propanediol (PD) under anaerobiosis takes place in two parallel ways. In the first one, a fraction of glycerol is oxidezed by glycerol-dehydrogenase (Glyc-DH) to dihydroxy-acetone (DHA), and then phosphorrylated by DHA kinase to enter glycol-lysis. The remaining glycerol is then dehydrated to 3-hydroxypropionaldehyde (3HPA) by glyceroldehydratase, where reduction con-tinues by
propanedioldehydrogenase (PPD-DH) and by a dependent NAD oxidorreduc-tase to 1,3propanediol [2-3]. 1,3-propanediol production can be performed biologically by several bacterial strains such as Klebsiella pneumoniae, Citrobacter freundii, Enterobacter agglomerans, Clostridium butyricum, and Clostridium acetobutylicum [4-5]. K.
pneumoniae and C. butyricum are commercially the most promising bacterial strains because of their high yield, productivity, and resistance to both substrate and product
48
inhibition. Among these two bacteria, K. pneumoniae DSM-2026 has been presented as one of the most appropriate bacterial strain for glycerol fermentation to 1,3-propanediol [6] and it was selected as the main process microorganism in this article. The purpose of this article is to analyze the glycerol fermentation to 1,3-propanediol by K. pneumoniae in one and two continuous fermentation stages.
Studies performed under batch and fed-batch cultures have showed low productivities of 1,3-propanediol, about 2-3 g L-1h-1 with a maximum 1,3-propanediol concentration of 5060 g L-1. In continuous cultures the productivity can be increased, but the maximum concentration reached is the half (about 30 g L-1) of the obtained under fed-batch or batch culture conditions. Due to glycerol bioconversion to 1,3-propanediol is a complex biological mechanism which is subject to inhibitions by substrate and products [7], process analysis become an important tool to develop efficient configuration process that allows obtaining the metabolite at high yield, concentration, and productivity. In this sense Posada et al [8], studied four culture configurations (batch, fed batch, continuous, and two continuous stages) for 1,3- propanediol production from glycerol, and each configuration process was optimized.
6.2 Ethanol
E. coli has showed the ability for metabolizing glycerol in presence of an external electron acceptor. Glycerol degradation process begins with the GlpF incorporation in the cytoplasm. Later phosphorylation process is carried out, which is catalyzed by GlpK kinasa. This phosphorylated carbohydrate (glycerol 3-phosphate) starts an oxidereduction process which is accelerated by different enzymes. The anaerobic process is catalyzed by the dehydrogenases GlpC, GlpB, and GlpA, while the aerobic process is catalyzed by GlpD. This dehydrogenation process produces dihydroxyacetone 3phosphate and finally glycolysis process takes place to obtain pyruvate (see Figure 6.1.). Also microorganisms such as Klebsiella, Citrobacter, Enterobacter, Clostridium, Lactobacillus, Bacillus, Propionibacterium, and Anaerobiospirillum have been reported for glycerol degrading in fermentative way. Degradation process of these microorganisms is strongly linked to 1,3-propanediol synthesis with Citrobacter freundii and Klebsiella pneumonia. However these microorganisms present diverse problems for their industrial use such as pathogenicity level, requirements of strict anaerobic conditions, and complex
6. Biochemical Conversion
49
cultivation media. In this way, it is necessary to search microorganisms able to metabolize glycerol without pathogenic effects as occurs with E. coli. Also, E. coli can use glycerol as carbon source without any external electron receiver. This process is regulated by GldA dehydrogenase and DHAK dihydroxyacetone kinase for obtaining ethanol, succinate, acetate, and formate (see Figure 6.2.) [9].
Figure 6.1. Schematic representation of glycerol degradation process on the part of Escherichia coli, on non fermentative process.
Deletions in E. coli have been carried out to increase formiate and ethanol yields from glycerol at a concentration of 10 g/L [10]. Thus, from glycerol dehydrogenase (gldA) and dihydroxyacetone kinase (dhaKLM) over expression a yield of 95% to ethanol from
50
glycerol was achieved. Also, a genomic analysis was carried out for determining the genes effect on the change from aerobic to anaerobic conditions in E. coli. A metabolic characterization to evaluate succinate, acetate, formiate, lactate, and ethanol yields was carried out [11].
Figure 6.2. Main metabolic pathways for fermentative degradation of glycerol by Escherichia coli.
51
A mixture of ethanol and formiate can be produced by glycerol fermentation using Klebsiella planticola isolated from the rumen [12]. Dharmadi et al [13], reported the glycerol fermentation by E. coli, the authors have evaluated the pH-dependence and CO2 availability. Ito et al [14], showed that glycerol at 10 g/L was almost completely consumed within 84 h; the main products were ethanol and succinic acid with molar yields of 86% and 7%, respectively. According to the authors, E. coli is already a good biocatalyst for glycerol conversion into ethanol and hydrogen.
E. aerogenes can be used for ethanol production at high yield from biodiesel wastes containing glycerol. In this way, a synthetic medium containing biodiesel wastes of glycerol at 80 mM was analyzed and glycerol was consumed in 24 h, producing 0.89 mol of H 2 and 1.0 mol of ethanol per mol of glycerol [14].
6.3 poly-3-hydroxybutirate
Polyhydorxyalcanoates are attractive substitute biopolymers for conventional petrochemical plastics which have similar physical properties to thermoplastics and elastomers. PHAs are homo or heteropolyesters synthesized and stored intracellularly by several Gram Negative bacteria [15]. PHAs can be produced from renewable resources through a fermentation process under restricted growth conditions for nitrogen, phosphorus, sulfurs and/or oxygen in the presence of an excess carbon source, and they can also be completely biodegraded by many microorganisms [16]. PHAs are stored in form of granules by bacteria and can account for up to 80% of the total bacterial dry weight [17]. On the other hand, polyhydoroxybutyrates (PHBs) were the first type of PHAs discovered and the most widely studied. PHB has similar properties to conventional plastics like polypropylene or polyethylene, and it can be extruded, molded, spun into fibers, made into films, and used to make heteropolymers with other synthetic polymers [18-19].
Wild strains such as Cupriavidus necator [20], Methylobacterium rhodesianum or recombinant microorganism such as E. coli recombinant [21, 22] can produce PHB using glycerol as a carbon and energy source. Bacteria used for PHAs production can be divided into two groups based on culture conditions. The first group requires limitation of an essential nutrient such as N, P Mg, K, O or S, and excess of a carbon source; some
52
examples are B. megaterium, C. necator, A. eutrophus, P. extorquens, and Ps. oleovorans. In the second group, nutrient limitation is not required and the polymer can be accumulated during the growth phase [23]; some examples are E. coli recombinant, Az. vinelandii recombinant, and A. latus. PHB producer strains which use glycerol as the carbon source are in the first group of bacteria. Polyhydorxyalcanoates are attractive substitute biopolymers for conventional petrochemical plastics which have similar physical properties to thermoplastics and elastomers.
PHAs are stored in the form of granules by bacteria and can account for up to 80% of the total bacterial dry weight [24]. On the other hand, polyhydorxybutyrates (PHBs) were the first type of PHAs discovered and the most widely studied. PHB has similar mechanical properties to conventional plastics like polypropylene or polyethylene, and it can be extruded, molded, spun into fibers, made into films, and used to make heteropolymers with other synthetic polymers [25].
The fermentation stage can be performed in different operational modes. Batch PHB production is normally induced by co-culturing the cells [26] or by limiting them with nitrogen availability using an excess of carbon source in the stationary phase [27]. To induce the desired nutrient limitation and to achieve a high cell density, a fed-batch process is the most commonly used method [28-29]. Thus, cell growth is maintained without nutrient limitation until a desired concentration is achieved. Then, an essential nutrient is limited to allow an efficient PHB synthesis. During this nutrient limitation stage the residual cell concentration (i.e., the difference between cell concentration and polymer concentration) remains almost constant and cell concentration increases only by polymeric intracellular accumulation [30]. For bacteria requiring an essential nutrient limitation a two-stage chemostat should be employed thus resulting in a 1.7 fold higher productivity compared to the one-stage chemostat [30]. Culture performance is affected by several variables including temperature, pH, fed carbon-to-nitrogen ratio, concentration of substrates and trace elements, ionic strength, agitation intensity, and dissolved oxygen. To substantially enhance the yield and productivity of many bioprocesses, optimization [31-32] and control [33] of the fermentation conditions has been used.
53
6.4 D-Lactic acid

The exhibited heterofermentative behavior of glycerol metabolism under anaerobic and microaerobic conditions by wild-type E. coli was recently reported [34-35]. And it was found that significant amounts of ethanol, acetic acid, succinic acid, and formic acid were produced, while a negligible amount of D-lactic acid was obtained. Besides the ability of E. coli to metabolize glycerol under anaerobic and microaerobic conditions, the corresponding pathways involved in the glycerol utilization were recently elucidated [36] and under these conditions, ethanol was identified as the primary fermentation product. Later, based on metabolic engineering strategies, an engineered E. coli for the efficient conversion of glycerol to D-lactic acid in a minimal medium was reported [37]. Thus, the homofermentative route to produce D-lactic acid was engineered by overexpressing the pathways involved in the glycerol conversion to D-lactic acid and blocking the pathways leading to the synthesis of by-products. In general terms, the enzymes involved in the pathways for glycerol conversion to glycolytic intermediates (i.e., GlpK-GlpD and GldADHAK) and the enzyme involved in the pathway for D-lactic acid synthesis from pyruvic acid were overexpressed (i.e., D-lactate dehydrogenase). Meanwhile, the by-products formation was minimized by inactivation of enzymes such as: pyruvate-formate lyase (pflB), fumarate reductase (frdA), phosphate acetyltransferase (pta), and alcohol/acetaldehyde dehydrogenase (adhE). Also, a mutation which blocks the aerobic D-lactate dehydrogenase (dld) was introduced in order to prevent the utilization of Dlactic acid. The Figure 6.3 shows both the pathways involved in the microaerobic utilization of glycerol in E. coli and the genetic modifications performed by metabolic engineering strategies for gene overexpressions or disruptions.
Although lactic acid bacteria have been used for D-lactic acid production from carbohydrate rich feedstocks, it has also been reported the use of alternative biocatalysts which are mainly engineered Escherichia coli strains able to produce D- or L-lactic acid [38-42]. But only a few papers have been published on the use of glycerol as carbon source for lactic acid production [37, 43]. For instance, Hong et al. [43] compared eight bacterial strains for lactic acid production from glycerol. Thus, the strain named AC-521 and a member of E. coli, showed the best performance for a fed-batch fermentation process. On the other hand, Mazumdar et al. [37] engineered several E. coli strains by overexpressing pathways involved in the conversion of glycerol to lactic acid and blocking
54
those leading to the synthesis of by-products as it was above described. In all cases they used a minimal medium supplemented with sodium selenite, Na2HPO4, (NH4)2SO4, NH4Cl, and 20 (or 40 or 60) g/l of pure (or crude) glycerol.
Figure 6.3. Pathways involved in the microaerobic utilization of glycerol in E. coli and Genetic modifications supporting the metabolic engineering strategies employed by Mazumdar et al [37]. Thicker lines (overexpression of gldA-dhaKLM, glpK-glpD, and ldhA) or cross bars (disruption of pflB, pta, adhE, frdA, and dld). Broken lines illustrate multiple steps. Relevant reactions are represented by the names of the gene(s) coding for the enzymes.
6.5 Succinic acid

Succinic acid is a C4 dicarboxylic acid produced as both intermediate of the tricarboxylic acid cycle (TCA) and one of the fermentation products of energy metabolism [44]. This metabolite can be used for the manufacture of industrially important chemicals including adipic acid, 1,4-butanediol, tetrahydrofuran, N-methyl pyrrolidinone, 2-pyrrolidinone, succinate salts and gamma-butyrolactone (see Figure 6.4); and for the synthesis of
55
biodegradable polymers such as polybutyrate succinate (PBS) and polyamides (Nylonx,4) and green solvents [45].
Figure 6.4. Products that can be synthesized from succinic acid.
Succinic acid is currently produced from crude oil by either catalytic hydrogenation of maleic anhydride to succinic anhydride and subsequent hydration, or direct catalytic hydrogenation of maleic acid [46]. The commercial price of petrochemically produced succinic acid is about 5.98.8 USD$/kg depending on its purity. Also, for its production from maleic anhydride, the raw material costs are about 1 USD$/kg of succinic acid [45].
Even though the production of chemicals based on succinic acid accounts to about 16.000 Ton/year [47], the market potential was estimated to be about 270,000 Ton/year if succinic acid replaced maleic anhydride for all uses [48-49]. Thus, because of these predictions, the dramatic raising in petroleum price, and the increasing environmental concerns, the fermentative production of succinic acid from renewable resources has recently received much attention. In this way several microorganisms including
56
Actinobacillus succinogenes [50-51], Anaerobiospirillum succiniciproducens [52-53], and Mannheimia succiniciproducens [54], recombinant Escherichia coli strains [55-56] and Corynebacterium glutamicum [57-58] have been found to produce succinic. Also, during fermentative production of succinic acid some by-products such as acetic acid, formic acid, lactic acid, and ethanol are also obtained. By products formation limits the possibility of its fermentative production in industrial scale, since the succinic acid yield is reduced and a more complex and costly downstream process is required [45, 59-60].
The biological production of succinate from glycerol occurs through a redox-balanced pathway in the presence of excess carbon dioxide. Unlike glycerol, succinate production from glucose is not redox balanced and can provide a maximum theoretical molar yield of 1.71 (carbon yield of 1.14) without external reducing power. Although ethanol and succinate are the only two products resulting from redox-balanced pathways of glycerol fermentation in E. coli [61-62], succinic acid is minor product [61]. In order to improve the succinate production, Blankschien et al. [63] engineered an E. coli strain by blocking pathways to competing metabolic products and thus leaving only the succinate pathway achieving redox balance during glycerol utilization (see Figure 6.5.)
Glycerol dissimilation in E. coli to dihydroxyacetonephosphate (DHAP) can proceed through two respiratory routes: the aerobic GlpKGlpD and the anaerobic GlpKGlpABC, or through the fermentative route GldADhaKLM (see Figure 6.5.). The last one has been reported to use glycerol efficiently under both anaerobic and microaerobic conditions [61, 64].
Because net ATP is typically not generated by substrate-level phosphorylation when succinate is produced from glycerol in wild-type E. coli (i.e., through ppc, see below and Fig. 6.5), use of the fermentative GldADhaKLM route is preferred because higher energy NADH is generated in glycerol dissimilation through GldA as opposed to a reduced flavin through GlpD or GlpABC [65-66]. However, DhaKLM uses phosphoenolpyruvate (PEP) as a cofactor, impacting the metabolic nodes available for glycerol fermentation to succinate (See Figure 6.5).
Production of succinate from glycerol involves fixing CO2 onto a 3-carbon intermediate, which is stepwise converted to succinate by the reductive branch of the TCA cycle [67]
57
(See Figure 6.5). E. coli uses PEP carboxylase (ppc) as its main carboxylation enzyme for succinate generation; however, this is not ideal as PEP levels will be decreased when the fermentative route of glycerol dissimilation (GldADhaKLM) is used (See Figure 6.5). An analogous argument can be made for the use of the primarily gluconeogenic PEP carboxykinases (from E. coli or natural succinate producers) [67]. Succinate synthesis from pyruvate, which is readily available, is limited because E. coli lacks a native pyruvatecarboxylase (pyc) and the conversion of pyruvate to malate by the gluconeogenic malic enzymes is not kinetically favored [67]. An effective way to retain the GldADhaKLM route and generate succinate is to introduce a pyruvate carboxylase (pyc) into E. coli, creating an efficient node for the step wise conversion of pyruvate to succinate.
Figure 6.5. Pathways involved in the micro aerobic utilization of glycerol and the generation of phosphoenol pyruvate and pyruvate, which can be carboxylated to intermediates leading to succinate [63].
58
Use of a heterologous pyruvate carboxylase (pyc) in E. coli to drive succinate production from glycerol leaves one remaining obstacle, the lack of net ATP production by substratelevel phosphorylation. Such a complication can be effectively overcome by the use of microaerobic conditions. ATP will be gained through oxidative phosphorylation resulting from the reducing equivalents generated during the utilization of glycerol, including those generated by the incorporation of glycerol into cell mass (i.e. cell mass is less reduced on average than glycerol) [61] (See Figure 6.5)
6.6 Propionic acid

Propionic acid and its calcium, sodium, and potassium salts are widely used as preservatives in animal feed and human foods, and propionic acid is also an important chemical intermediate in the synthesis of cellulose fibers, herbicides, perfumes and pharmaceuticals [68-69]. Currently, almost all propionic acid is produced by chemical synthesis from petroleum feedstocks. The acid also could be produced by
propionibacteria via the dicarboxylic acid pathway with acetic acid and succinic acid as byproducts [70-74], but low yield and productivity due to the inhibition of propionic acid on cell growth and propionic acid synthesis [72, 75] is a problem. To alleviate the inhibition of propionic acid on microbial growth and propionic acid synthesis, two approaches, extractive propionic acid fermentation [76-78] and propionic acid production with propionic acid-tolerant bacteria obtained via adaptive evolution [72, 79-80] have been developed. Despite such advancements, current microbial propionic acid production cannot economically compete petrochemical routes. Producing propionic acid from agricultural and industrial wastes may make microbial propionic acid production economically competitive. Glycerol is a main by-product of the biodiesel industry [81] and could thus be a low-cost feedstock to produce propionic acid. While most studies on propionic acid production by Propionibacterium acidipropionici have focused on glucose and whey lactose [77, 82-85], some studied have explored glycerol as the carbon source [86-87], and it was observed that glycerol might be advantageous since less acetic acid was produced during the consumption of glycerol [70, 86].
Since optimal conditions for the use of glycerol in propionic acid production have not yet been established, we optimized propionic acid production by propionic acid-tolerant P. acidipropionici CGMCC 1.2230 with glycerol as the carbon source in batch cultures and
59
then scaled-up production in a 10 m3 fermentor using the optimized conditions. The results obtained here may be helpful for industrial production of propionic acid.
References
[1] Ya-Nan, Z., C. Guo, and Y. Shan-Jing, Microbial production of 1,3-propanediol from glycerol by encapsulated Klebsiella pneumoniae. Biochemical Engineering Journal, 2006. 32: p. 93-99. [2] Barbirato, F., Soucaille P., and Bories. A.: Physiologic mechanisms involved in accumulation of 3-hydroxypropionaldehyde during fermentation of glycerol by
Enterobacter agglomerans. Appl. Environ. Microbiol. 62, 4405-4409 (1996). [3] Papanikolaou, S., Ruiz-Sanchez, P., Pariset, B., Blanchard, F., and Fick. M.: High production of 1,3-Propanediol from industrial glycerol by a newly isolated Clostridium butyricum strain. J. Biotechnol. 77, 191208 (2000). [4] Gonzalez-Pajuelo, M., Meynial-Salles, I., Mendes, F., Andrade, J.C., Vasconcelos, I., and Soucaille, P.: Metabolic engineering of Clostridium acetobutylicum for the industrial production of 1,3-Propanediol from glycerol. Metab. Eng. 7, 329-336 (2005). [5] Ito, T., Nakashimada, Y., Senba, K., Matsui, T., and Nishio N.: Hydrogen and ethanol production from glycerol-containing wastes discharged after biodiesel manufacturing process. J. Biosci. Bioeng.. 100, 260265 (2005). [6] Zhao, Y.N., Chen, G., and Yao, S.J.: Microbial production of 1,3-propanediol from glycerol by encapsulated Klebsiella pneumoniae. Biochem. Eng. J. 32, 9399 (2006). [7] Ya-Qin, S., et al., Mathematical modeling of glycerol fermentation by Klebsiella pneumoniae: Concerning enzyme-catalytic reductive pathway and transport of glycerol and 1,3-propanediol across cell membrane. . Biochemical Engineering Journal, 2008. 38: p. 22-32. [8] Posada, J.A., J.J. Jaramillo, and C.A. Cardona, Glycerol Fermentation to 1,3Propanediol: Modeling, Simulation, and Optimization. . 1st International Conference on Modelling and Simulations (ICOMOS 2010 - VF) VIRTUAL FORUM, July 12th - 26th, 2010.
60
[9] Murarka, A., et al., Fermentative utilization of Glycerol by Escherichia coli and its implications for the production of fuels and chemicals. Appl. Environ. Microbiol., 2008. 74: p. 1124-1135. [10] Yazdani, S.S. and R. Gonzalez, Anaerobic fermentation of glycerol: a path to economic viability for the biofuels industry. . Curr. Opin. Biotechnol., 2007. 18: p. 213-219. [11] Nikel, P.L., et al., Poly(3-hydroxybutyrate) synthesis from glycerol by a recombinant Escherichia coli arcA mutant in fed-batch microaerobic cultures. Appl. Microbiol. Biotechnol., 2008. 77: p. 1337-1343. [12] G.N.Jarvis, E.R.B. Moore, and J.H. Thiele, Formate and ethanol are the major products of glycerol fermentation produced by a Klebsiella planticola strain isolated from red deer. . J. Appl. Microbiol., 1997. 83: p. 166-174. [13] Dharmadi, Y., A. Murarka, and R. Gonzalez, Anaerobic Fermentation of Glycerol by Escherichia coli: A New Platform for Metabolic Engineering. Biotechnology & Bioengineering, 2006. 94(5): p. 821-829. [14] Ito, T., et al., Hydrogen and Ethanol Production from Glycerol-Containing Wastes Discharged after Biodiesel Manufacturing Process. J. Biosci. Bioeng., 2005. 100(3): p. 260-265 [15] Mahishi LH, Tripathi G, Rawal SK. Poly(3-hydroxybutyrate) (PHB) synthesis by recombinant Escherichia coli harbouring Streptomyces aureofaciens PHB biosynthesis genes: Effect of various carbon and nitrogen sources. Microbiol Res 2003;158;1927. [16] Steinbchel A. Perspectives for biotechnological production and utilization of biopolymers: metabolic engineering of polyhydroxyalkanoate biosynthesis pathways as a successful example. Macromol Biosci 2001;1;1-24. [17] Khanna S, Srivastava AK. Recent advances in microbial polyhydroxyalkanoates. Process Biochem 2005;40;607619. [18] Savenkova L, Gercberga Z, Muter O, Nikolaeva V, Dzene A, Tupureina V. PHBbased films as matrices for pesticides. Process Biochem 2002:37;719-722. [19] Ali AQ, Kannan TP, Ahmad A, Samsudin AR. In vitro genotoxicity tests for polyhydroxybutyrate A synthetic biomaterial, Toxicol in Vitro 2008;22(1);57-67.
61
[20] Cavalheiro, J.M.B.T., et al., Poly(3-hydroxybutyrate) production by Cupriavidus necator using waste glycerol. Process Biochemistry, 2009. 44(5): p. 509-515. [21] Pablo, I.N., et al., Poly(3-hydroxybutyrate) synthesis from glycerol by a recombinant Escherichia coli arcA mutant in fed-batch microaerobic cultures. . Appl. Microbiol. Biotechnol., 2008. 77: p. 1337-1343. [22] Mahishi, L.H., G. Tripathi, and S.K. Rawal, Poly(3-hydroxybutyrate) (PHB) synthesis by recombinant Escherichia coli harbouring Streptomyces aureofaciens PHB biosynthesis genes: Effect of various carbon and nitrogen sources. Microbiol. Res., 2003. 158: p. 1927. [23] Lee, S.Y., Plastic bacteria? Progress and prospects for polyhydroxyalkanoate production in bacteria. Tibtech, 1996. 14: p. 431-438. [24] Khanna, S. and A.K. Srivastava, Recent advances in microbial
polyhydroxyalkanoates. Process Biochem., 2005. 40: p. 607-619. [25] Ali, A.Q., et al., In vitro genotoxicity tests for polyhydroxybutyrate - A synthetic biomaterial. Toxicol in Vitro 2008. 22(1): p. 57-67. [26] Ganduri VSRK, Ghosh S, Patnaik PR. Mixing control as a device to increase PHB production in batch fermentations with co-cultures of Lactobacillus delbrueckii and Ralstonia eutropha. Process Biochem 2005:40; 257-264. [27] Khanna S, Srivastava AK. Computer simulated fed-batch cultivation for over production of PHB: A comparison of simultaneous and alternate feeding of carbon and nitrogen. Biochem Eng J 2006;27;197203. [28] Shahhosseini S. Simulation and optimisation of PHB production in fed-batch culture of Ralstonia eutropha. Process Biochem 2004:39;963-969. [29] Khanna S, Srivastava AK. Optimization of nutrient feed concentration and addition time for production of poly(-hydroxybutyrate). Enzyme Microb Technol 2006;39;1145 1151. [30] Lee SY. Plastic bacteria? Progress and prospects for polyhydroxyalkanoate production in bacteria. Tibtech 1996;14;431-438.
62
[31] Grothe E, Moo-Young M, Chisti Y. Fermentation optimization for the production of poly(-hydroxybutyric acid) microbial thermoplastic. Enzyme Microb Technol 1999; 25:132141. [32] Khanna S. Srivastava AK. Statistical media optimization studies for growth and PHB production by Ralstonia eutropha. Process Biochem 2005:40;2173-2182. [33] Dias JML,, Pardelha F, Eusbio M, Reis MAM, Oliveira R. On-line monitoring of PHB production by mixed microbial cultures using respirometry, titrimetry and chemometric modeling. Process Biochem 2009:44;419-427. [34] Gonzalez, R., A. Murarka, Y. Dharmadi, and S. S. Yazdani. 2008. A new model for the anaerobic fermentation of glycerol in enteric bacteria: trunk and auxiliary pathways in Escherichia coli. Metab. Eng. 10:234245. [35] Murarka, A., Y. Dharmadi, S. S. Yazdani, and R. Gonzalez. 2008. Fermentative utilization of glycerol in Escherichia coli and its implications for the production of fuels and chemicals. Appl. Environ. Microbiol. 74:11241135. [36] Durnin, G., J. Clomburg, Z. Yeates, P. J. J. Alvarez, K. Zygourakis, P. Campbell, and R. Gonzalez. 2009. Understanding and harnessing the microaerobic metabolism of glycerol in Escherichia coli. Biotechnol. Bioeng. 103:148161. [37] Mazumdar, S., Clomburg, J.M., and Gonzalez, R. 2010. Escherichia coli Strains Engineered for Homofermentative Production of D-Lactic Acid from Glycerol. Applied and Environmental Microbiology. 76: 43274336. [38] Dien, B. S., N. N. Nichols, and R. J. Bothast. 2001. Recombinant Escherichia coli engineered for production of L-lactic acid from hexose and pentose sugars. J. Ind. Microbiol. Biotechnol. 27:259264. [39] Okano, K., T. Tanaka, C. Ogino, H. Fukuda, and A. Kondo. 2010. Biotechnological production of enantiomeric pure lactic acid from renewable resources: recent achievements, perspectives, and limits. Appl. Microbiol. Biotechnol. 85:413423. [40] Zhou, S., K. T. Shanmugam, and L. O. Ingram. 2003. Functional replacement of the Escherichia coli D()-lactate dehydrogenase gene (ldhA) with the L(+)-lactate
63
dehydrogenase gene (ldhL) from Pediococcus acidilactici. Appl. Environ. Microbiol. 69:22372244. [41] Zhou, S., T. B. Causey, A. Hasona, K. T. Shanmugam, and L. O. Ingram. 2003. Production of optically pure D-lactic acid in mineral salts medium by metabolically engineered Escherichia coli W3110. Appl. Environ. Microbiol. 69:399407. [42] Zhu, Y., M. A. Eiteman, K. DeWitt, and E. Altman. 2007. Homolactate fermentation by metabolically engineered Escherichia coli strains. Appl. Environ. Microbiol. 73:456464. [43] Hong, A.A.; Cheng, K.K.; Peng, F.; Zhou, S.; Sun, Y.; Liu, C.M.; Liu, D.H. Strain isolation and optimization of precess parameters for bioconversion of glycerol to lactic acid. J. Chem. Technol. Biotechnol., 2009, 84, 1576-1581. [44] Huh Y.S., Jun Y.S., Hong Y.K., Song H., Lee S.Y., Hong W.H. Effective purification of succinic acid from fermentation broth produced by Mannheimia succiniciproducens. Process Biochemistry 41 (2006) 14611465. [45] Song H, Lee S. Y., Production of succinic acid by bacterial fermentation. Enzyme and Microbial Technology 39 (2006) 352361. [46] Kurzrock T, Weuster-Botz D., Recovery of succinic acid from fermentation broth. Biotechnol Lett (2010) 32:331339. [47] Patel MA (2006) Medium and long-term opportunities and risks of the biotechnological production of bulk chemicals from renewable resourcesthe potential of white biotechnology. The BREW Project (http:www.chem.uu.nl/brew/) [48] Delhomme C, Weuster-Botz D, Kuhn F (2009) Succinic acid from renewable resources as a C4 building-block chemicala review of the catalytic possibilities in aqueous media. Green Chem 11:1326 [49] Wilke T, Vorlop K-D (2004) Industrial bioconversion of renewable resources as an alternative to conventional chemistry. Appl Microbiol Biotechnol 66:131142 [50] Guettler MV, Rumler D, Jain MK. Actinobacillus succinogenes sp. nov., a novel succinic-acid-producing strain from the bovine rumen. Int J Syst Bacteriol 1999;49:207 16.
64
[51] Urbance SE, Pometto AL, DiSpirito AA, Denli Y. Evaluation of succinic acid continuous and repeat-batch biofilm fermentation by Actinobacillus succinogenes using plastic composite support bioreactors. Appl Microbiol Biotechnol 2004;65:66470. [52] Lee PC, Lee WG, Kwon S, Lee SY, Chang HN. Succinic acid production by Anaerobiospirillum succiniciproducens: effects of the H2/CO2 supply and glucose concentration. Enzyme Microb Technol 1999;24:54954. [53] Samuelov NS, Lamed R, Lowe S, Zeikus JG. Influence of CO 2 HCO 3 levels and pH on growth, succinate production and enzyme activities of Anaerobiospirillum succiniciproducens. Appl Environ Microbiol 1991;57:30139. [54] Lee PC, Lee SY, Hong SH, Chang HN. Isolation and characterization of new succinic acid producing bacterium, Mannheimia succiniciproducens MBEL 55E, from bovine rumen. Appl Microbiol Biotechnol 2002;58:6638. [55] Hong SH, Lee SY. Metabolic flux analysis for succinic acid production of by recombinant Escherichia coli with amplified malic enzyme activity. Biotechnol Bioeng 2001;74:8995. [56] Lin H, San K-Y, Bennett GN. Effect of Sorghum vulgare phosphoenolpyruvate carboxylase and Lactococcus lactis pyruvate carboxylase coexpression on succinate production in mutant strain of Escherichia coli. Appl Microbiol Biotechnol 2005;67:51523. [57] Bechthold I, Bretz K, Kabasci S, Kopitzky R, Springer A (2008) Succinic acid: a new platform chemical for biobased polymers from renewable resources. Chem Eng Technol 5:647654. [58] Okino SH, Noburyu R, Suda M, Jojima T, Inui M, Yukawa H (2008) An efficient succinic acid production process in a metabolically engineered Corynebacterium glutamicum strain. Appl Microbiol Biotechnol 81:459464. [59] Huh YS, Hong YK, Hong WI, Chang HN. Selective extraction of acetic acid from the fermentation broth produced by Mannheimia succiniciproducens. Biotechnol Lett 2004;26:15814.
65
[60] Kim BS, Hong YK, Hong WH. Effect of salts on the extraction characteristics of succinic acid by predispersed solvent extraction. Biotechnol Bioprocess Eng 2004;9:20711. [61] Durnin, G., Clomburg, J., Yeates, Z., Alvarez, P.J.J., Zygourakis, K., Campbell, P., Gonzalez, R., 2009. Understanding and harnessing the microaerobic metabolism of glycerol in Escherichia coli. Biotech.Bioeng.103,148161. [62] Zhang, Y., Huang, Z., Du, C., Li, Y., Cao, Z., 2009. Introduction of an NADH regeneration system into Klebsiella oxytoca leads to an enhanced oxidative and reductive metabolism of glycerol. Metab. Eng. 11,101106. [63] Blankschien M.D., Clomburg J.M., Gonzalez R., Metabolic engineering of Escherichia coli for the production of succinate from glycerol. Metabolic Engineering 12 (2010) 409 419. [64] Gonzalez, R., Murarka, A., Dharmadi, Y., Yazdani, S.S., 2008. A new model for the anaerobic fermentation of glycerol in entericbacteria: trunk and auxiliary pathways in Escherichia coli. Metab.Eng.10,234245. [65] Tran, Q.H., Bongaerts, J., Vlad, D., Unden, G., 1997. Requirement for the protonpumping NADH dehydrogenase I of Escherichia coli in respiration of NADH to fumarate and its bioenergetic implications. Eur. J. Biochem. 244,155160. [66] Unden, G., D unnwald, P., 2008. The aerobic and anaerobic respiratory chain of Escherichia coli and Salmonella enterica, enzymes and energetics. In: Niedhardt, F.C. (Ed.),EcoSalEscherichia coli and Salmonella, Cellular and Molecular Biology. ASMPress, Washington,D.C. [67] Sauer, U., Eikmanns, B.J., 2005. The PEP-pyruvate-oxaloacetate node as the switch point for carbon flux distribution in bacteria. FEMS Microbiol. Rev. 29, 765794. [68] Hsu, S., Yang, S., 1991. Propionic acid fermentation of lactose by Propionibacterium acidipropionici: effects of pH. Biotechnol. Bioeng. 38, 571578. [69] Roberto, M.C., Mayra, T., 2002. Production of propionate by fed-batch fermentation of Propionibacterium acidipropionici using mixed feed of lactate and glucose. Biotechnol. Lett. 24, 427431.
66
[70] Coral, J., Karp, S.G., Vandenberghe, L.P.S., Parada, J.L., Pandey, A., Soccol, C.R., 2008. Batch fermentation model of propionic acid production by Propionibacterium acidipropionici in different carbon sources. Appl. Biochem. Biotech. 151, 333 341. [71] Goswami, V., Srivastava, A.K., 2000. Fed-batch propionic acid production by Propionibacterium acidipropionici. Biochem. Eng. J. 4, 121128. [72] Suwannakham, S., Yang, S.T., 2005. Enhanced propionic acid fermentation by Propionibacterium acidipropionici mutant obtained by adaptation in a fibrousbed bioreactor. Biotechnol. Bioeng. 91, 325337. [73] Zhang, A., Yang, S., 2009a. Propionic acid production from glycerol by metabolically engineered Propionibacterium acidipropionici. Process Biochem. 44, 13461351. [74] Zhang, A., Yang, S.T., 2009b. Engineering Propionibacterium acidipropionici for enhanced propionic acid tolerance and fermentation. Biotechnol. Bioeng. 104, 766773. [75] Gu, Z., Glatz, B., Glatz, C., 1998. Effects of propionic acid on propionibacteria fermentation. Enzyme Microb. Technol. 22, 1318. [76] Goswami, V., Srivastava, A.K., 2001. Propionic acid production in an in situ cell retention bioreactor. Appl. Microbiol. Biot. 56, 676680. [77] Lewis, V., Yang, S., 1992. A novel extractive fermentation process for propionic acid production from whey lactose. Biotechnol. Progr. 8, 104110 [78] Ozadali, F., Glatz, B., Glatz, C., 1996. Fed-batch fermentation with and without online extraction for propionic and acetic acid production by Propionibacterium acidipropionici. Appl. Microbiol. Biot. 44, 710716. [79] Woskow, S., Glatz, B., 1991. Propionic acid production by a propionic acid-tolerant strain of Propionibacterium acidipropionici in batch and semicontinuous fermentation. Appl. Environ. Microb. 57, 28212828. [80] Suwannakham, S., Huang, Y., Yang, S.T., 2006. Construction and characterization of ack knock-out mutants of Propionibacterium acidipropionici for enhanced propionic acid fermentation. Biotechnol. Bioeng. 94, 383395.
67
[81] Ito, T., Nakashimada, Y., Senba, K., Matsui, T., Nishio, N., 2005. Hydrogen and ethanol production from glycerol-containing wastes discharged after biodiesel
manufacturing process. J. Biosci. Bioeng. 100, 260265. [82] Babuchowski, A., Hammond, E., Glatz, B., 1993. Survey of propionibacteria for ability to produce propionic and acetic acids. J. Food Protect. 56, 493496. [83] Carrondo, M., Crespo, J., Moura, M., 1988. Production of propionic acid using a xylose utilizing Propionibacterium. Appl. Biochem. Biotech. 17, 295312 [84] Lewis, V., Yang, S., 1992b. Propionic acid fermentation by Propionibacterium acidipropionici: effect of growth substrate. Appl. Microbiol. Biot. 37, 437442. [85] Quesada-Chanto, A., Afschar, A., Wagner, F., 1994. Optimization of a Propionibacterium acidipropionici continuous culture utilizing sucrose. Appl. Microbiol. Biot. 42, 1621. [86] Himmi, E.H., Bories, A., Boussaid, A., Hassani, L., 2000. Propionic acid fermentation of glycerol and glucose by Propionibacterium acidipropionici and Propionibacterium freudenreichii ssp. Shermanii. Appl. Microbiol. Biot. 53, 435440. [87] Zhang, A., Yang, S., 2009a. Propionic acid production from glycerol by metabolically engineered Propionibacterium acidipropionici. Process Biochem. 44, 13461351.
7. Study Cases of Chemical Conversion of Glycerol

Three different technological schemes to transform the glycerol obtained as by-product in biodiesel industry to added-value products are here designed, simulated, and economically assessed. Dehydration, steam gasification, and hydrogenolysis were the analyzed processes where acrolein, hydrogen, and 1,2-propanediol are their respective products. For dehydration and gasification processes a glycerol conversion of 100% was reached, and the respective molar yields to acrolein and hydrogen were 85.2 % and 78.2 %. Also, these two processes were heat integrated. 175 and 67 W/(feeding kg) were recovered for dehydration and gasification respectively. Economic results showed that the three processes are economically viable, and the highest economical return was obtained for 1,2-propanediol.
7.1 Generalities
Both dehydration and hydrogenolysis reactors were simulated based on a stoichiometric approach in which acrolein, hydroxyacetone, acetaldehyde, formaldehyde, and water were considered as the dehydration products, meanwhile 1,2-propanediol,
hydroxyacetone, ethylene glycol, and methane were considered as hydrogenolysis products. On the other hand, the gasification reactor was simulated as an Rgibbs module and carbon monoxide, carbon dioxide, hydrogen, methane, and water were considered as the reaction products.
All processes were integrated in a basic level for a better economic performance. Thus, the dehydration process was heat integrated, meanwhile the gasification process was heat and mass integrated and finally the hydrogenolysis process was mass integrated.
70
In all cases the non-random two liquid (NRTL) thermodynamic model was utilized to calculate the activity coefficients of the liquid phases and Redlich-Kwong (RK) equation of state was used to model the vapor phase.
Process engineering looks into the design of high-performance processes that meet mainly two kinds of criteria: high conversion levels and low production costs. Thus comparing opportunities to convert the glycerol by-product into added-value products is the core of this analysis. Also, simulations of the technological schemes were used to generate their respective mass and energy balance sheets, which are the basic input for the techno-economic analysis.
7.2 Acrolein Production

Acrolein is used as raw material to treat fiber and to produce acrylic acid and medicines, even more it has been used as a growth control agent of microbes in feed process lines due to its antimicrobial activity.
Although commercial manufacturing of acrolein has been based on the petrochemicalpropylene oxidation process, this compound can also be produced by homogeneous catalytic dehydration of glycerol in presence of zinc sulfate. In the last case, for glycerol dehydration at 360 C and 25 MPa, a maximal selectivity of 75 mol % at a conversion level of 50% was reached [1].On the other hand, both higher selectivity and conversion were obtained using heteropolyacid catalysts supported on silica with presence of titanium, aluminium, and zirconium oxides. Thus, the H 3 PW 12 O 40 catalyst supported on ZrO 2 was able to produce acrolein at a selectivity of 70 % [2]. Also, complete conversion of glycerol with selectivities ranging from 75 to 86 % was reported at temperatures between 275 -325 C on these heteropolyacid catalysts [3-4]. Even more, silicotungstic acid [5] and Nb 2 O 5 [6] supported on activated carbon have also been reported to produce acrolein from glycerol at selectivity levels near to 50%.
In addition to heterogeneous catalysis, acrolein can also be produced by glycerol conversion on hot-compressed water (HCW) with H 2 SO 4 as catalyst. In this sense, it was reported that yield to acrolein can be improved by increasing either the operational
7. Study Cases of Chemical Conversion
71
pressure or the concentration of glycerol or H2SO4. Using this conversion way, selectivity values up to 80% can be obtained [7].
The highly exothermic glycerol dehydration to acrolein is carried out by an acid catalyzed process as shown in Figure 7.1. An aqueous glycerol stream at 10 wt % is heated in two stages; in the first one the heat produced during the dehydration reaction is recovered in the Heat Exchanger I, meanwhile using the Heater I the reaction temperature is reached. Thus dehydrogenation reaction takes place at 275 C and 1 bar.
DC-1 Con-1 R-1 HE-1 Diluted Glycerol H-1 Reactives HE-2 Vapor Distillate 1 Distillate 2 DC-2
Bottoms 1 Products Condensate
Bottoms 2
Figure 7.1. Simplified flowsheet for acrolein production by glycerol dehydration. HE-1: Heat exchanger I; H-1: Heater; R-1: Dehydration reactor; Cond-1: Condenser; HE2: Heat exchanger II; DC-1: Distillation column I; DC-2: Distillation column II.
Equations (7.1) to (7.3) describe the reactive system for catalytic glycerol dehydration [4, 8], in which acrolein (C 3 H 4 O), hydroxyacetone (C 3 H 6 O 2 ), acetaldehyde (C 2 H 4 O), and formaldehyde (CH 2 O) are the main reaction products. The normalized yields for each reaction are 85%, 8% and 7%, respectively.
C 3 H 8 O 3 C 3 H 4 O + 2H 2 O (Equation 7.1)
C 3 H 8 O 3 C 3 H 6 O 2 + H 2 O (Equation 7.2)
C 3 H 8 O 3 C 2 H 4 O + CH 2 O + H 2 O (Equation 7.3)
After the dehydration reaction, products stream is cooled in the Heat Exchanger I, and thus this stream is thermally integrated with the fed diluted glycerol stream. Then, a share of water is condensed and the resulting mixture is cooled to 80 C in the Heat Exchanger II. Thus, the downstream process continues with a distillation column where both remaining water and hydroxyacetone are retired by the bottoms stream. Finally, to purify the acrolein stream from 92 to 98.5 wt %, a second distillation column could be used.
72
Thus, a mixture of acetaldehyde and formaldehyde is obtained by the distillated stream, but the condenser should be operated with a special coolant liquid since the distillated stream is obtained to -10 C.
During the heating of the fresh feed glycerol stream in the dehydration process, 175 W/(feeding kg) were recovered from the effluent reactor stream by mean of the Heat Exchanger I. Also, in the dehydration reactor not only glycerol was completely converted but also a yield to acrolein of 85.25% mol was achieved. Then, following the downstream process line, acrolein at 92.2 wt % was obtained in the distillated stream from the Distillation Column I and also 99.4 % of the produced acrolein was recovered. Additionally, in order to obtain a higher purity of acrolein, a further distillation column was analyzed. Thus, an acrolein steam at 98.5 wt % was obtained, but to reach the operation conditions a special coolant is required since the condenser must to operate at -10 C. In this way, when the second distillation column was used, the 98.7 % of the produced acrolein was recovered, this operational requirements surely increase the production costs. On the other hand, the most important energy consumptions were obtained the Heater I and the reboilers of both distillation columns, with net heat duties of: 591.1, 74.4, and 5.02 W/(feeding kg), respectively. The main simulations results for the dehydration process are shown in Table 7.1.
Table 7.1. Simulation results for dehydration process from glycerol Diluted Reactives Products Glycerol 298,1 548,1 548,1 1 1 1 0,1 0,9 0 0 0 0 100 0,1 0,9 0 0 0 0 100 0 0,936 0,052 0,007 0,003 0,002 100 Stream Condensate Vapor 372,6 1 0 0,978 0,002 0,019 0 0 17,75 352,7 1 0 0,927 0,063 0,004 0,004 0,003 82,25 Distillate 1 309,4 1 0 0,006 0,922 0 0,034 0,038 5,59 Bottoms 1 373,1 1 0 0,994 0 0,004 0,001 0 76,66 Distillate 2 262,5 1 0 0 0,087 0 0,371 0,542 0,39 Bottoms 2 325,4 1 0 0,006 0,985 0 0,009 0 5,20
Temp. (K) Pressure (atm) Comp. (wt%) Glycerol Water Acrolein Hydroxyacetone Acetaldehyde Formaldehyde Total Flow Rate (Kg/h)
73
7.3 Hydrogen production

Hydrogen is currently derived from nonrenewable natural gas and petroleum, but it could be produced from renewable resources such as biomass or its derivates [9]. Many applications in fields such as electricity generation, fuel cells, and automotive fuels have been found for hydrogen since it can be used in mobile and stationary applications. Besides, due to its high energy efficiency, sustainable, and nonpolluting character, hydrogen is considered as an obvious alternative to hydrocarbon fuels such as gasoline. Thus, it is expected that hydrogen plays a key role in the worlds energy future by replacing fossil fuels and storage energy [10].
Non-catalytic processes such as pyrolysis and steam gasification are technologies able to produce added-value products such as hydrogen and syn gas from glycerol. Pyrolysis is a thermal cracking process of organic liquids or solids at high temperature performed in oxygen absence; meanwhile steam gasification is carried out in presence of oxygen and produces fuel gases with higher hydrogen content than pyrolytic process.
Pyrolysis produces liquid fuels at low temperatures (400 to 600 C), but when this process is carried out at high temperatures (> 750 |C) gaseous products are obtained. Moreover, gasification is performed in presence of oxygen (i.e., air, pure oxygen, or steam) and a mixture of carbon monoxide and hydrogen is also produced [11]. In the case of steam gasification of glycerol at 600 - 700 C, a yield of 92.3 mol % to syn-gas with a H 2 /CO molar ratio of 2/1 was reported [12]. Meanwhile glycerol pyrolysis over carbonaceous catalysts at 800 C produces synthesis gas up to 81 vol % [13].
On the other hand, crude glycerol has been analyzed as raw material to produce hydrogen. For instance, yields ranging from 77 to 95 wt % were reported for catalytic steam reforming of crude glycerol on commercial Ni [14]. Besides, higher yields have been reached by steam gasification from crude glycerol such as 97 % to syn-gas and 65.7 % to H 2 [15]. Hydrogen is produced by supercritical water gasification (SCWG), with glycerol as carbon source. The simplified flowsheet for hydrogen production from glycerol is shown in Figure 7.2. A mixture containing diluted glycerol at 25 wt % is heated in two stages, and an intermediate compression process is required. Thus, the first heat exchanger produces
74
overheat vapor and then the reaction pressure is reached by mean of the compressor. During the compression operation a heat excess is produced, this is used to heat the fresh diluted glycerol stream in the Heat Exchanger I. Thus, compressed stream and fed glycerol stream are thermally integrated. Then, by mean of the Hater I the reaction temperature is achieved, and low heating requirement are needed. The resulting mixture is fed to the gasification reactor at 600 C and 300 bar (i.e. supercritical conditions).
HE-1 Diluted Glycerol
Comp-1
R-1 H-1 Reactives
Overheat Vapor Products 1 HE-2 Products 2
Sep-1 Vapor 1
Sep-2 Vapor 2
Recycle Fresh Water M-1
Condensate 1 Condensate 2
Figure 7.2. Simplified flowsheet for hydrogen production by gasification. HE-1: Heat exchanger I; Comp-1: Compressor: H-1: Heater I, R-1: Gasification reactor; HE-2: Heat exchanger II; Sep-1: High-Pressure (HP) gas-liquid separator; Sep-2: LowPressure (LP) gas-liquid separator; M-1: Mixer.
Molar distribution of reaction products after gasification has been reported as follows [16]: hydrogen 29.2 %, carbon dioxide 36.1 %, carbon monoxide 0.9 % and methane 33.8 %. Then, the reaction stoichiometry can be expressed as shows the equation (7.4).
10C 3 H 8 O 3 + H 2 O 13H 4 + 15CO 2 + CO + 14CH 4 (Equation 7.4)

Reaction products are cooled in the Heat Exchanger II and a two-phase stream (gas and liquid phases) is obtained. This mixture is fed to the high-pressure (HP) gas-liquid separator which operates at temperatures between 25-100 C, and at 300 bar. Then the liquid phase, obtained by bottoms, is further transferred to the low-pressure (LP) gasliquid separator which operates at 20 C and 1 bar. The HP separator produces a H 2 -rich gas stream, while the LP separator produces a CO 2 -CH 4 -rich gas. The liquid product from LP separator, which is mostly water, is mixed with fresh water and fed to the Heat Exchanger II as cooling fluid, and thus an overheat-vapor stream is obtained. Finally, the CO 2 -CH 4 -rich gas stream could be burned to generate process heat.
75
For the dehydration process, the feed stream to the gasification process was also heat integrated. Thus, the heat excess obtained during the isentropic compression operation (i.e., 67 W/(feeding kg)) was recovered in the Heat Exchanger I. Besides, the gasification reactor was able to reach a conversion of 100 % at a yield of 78.2 %. In addition to the heat integration of the fresh feed glycerol stream, the products reaction stream was also heat integrated and 784 W/(kg of reaction products) were recovered. Thus, an overheat vapor stream was obtained. In order to purify the hydrogen produced in the gasification reactor, two gas-liquid separation units were used. The first one is the High-Pressure (HP) gas-liquid separator in which H 2 at 90.9 mol % was obtained, meanwhile in the LowPressure (LP) gas-liquid separator the water was retired and a stream containing a mixture CO 2 -CH 4 was obtained. Also, this stream containing a mixture CO 2 -CH 4 could be used to generate process heat. The net heat duties in the gasification process were represented by the Compressor, Heater I, and Gasification Reactor; where the heat duty reported by the simulator were 673, 146.3 and -75.13 W/(feeding kg), respectively. The main simulations results for the gasification process are shown in Table 7.2.
Table 7.2. Simulation results for gasification process from glycerol Stream Product Diluted Glycerol Temp. (K) Pressure (atm) Comp. (wt%) Glycerol Water CO 2 CO CH 4 H2 Total Flow Rate (Kg/h) 100 100 100 100 0,994 99,01 20,69 78,32 107,99 0,061 0,939 0 0 0 0 0,061 0,939 0 0 0 0 0 0,804 0,074 0,002 0,080 0,041 0 0,804 0,074 0,002 0,080 0,041 0 0 0,032 0,007 0,052 0,909 0 0,841 0,076 0,002 0,082 0 0 0,023 0,451 0,011 0,515 0 0 0,978 0,013 0 0,009 0 0 0,984 0,009 0 0,006 0 298,1 1 873,1 300 Reactives s 1 873,1 300 Products Vapor 2 308,9 300 1 308,9 300 Condensate Vapor Condensat Recycle 1 308,9 300 2 293,1 1 e2 293,1 1 293,1 1
76
7.4 1,2-propanediol production

1,2-propanediol is produced commercially by the the hydration of propylene oxide derived from propylene by either the chlorohydrin process or the hydroperoxide process. But in the presence of metallic catalysts and hydrogen, glycerol can be hydrogenated to propylene glycol, 1,3 propanediol, or ethylene glycol.
1,2-propanediol is used in unsaturated polyester resins, functional fluids (antifreeze, deicing, and heat transfer), pharmaceuticals, foods, cosmetics, liquid detergents, tobacco humectants, flavors and fragrances, personal care, paints and animal feed [17]. The antifreeze and deicing market is growing because of concern over the toxicity of ethylene glycol-based products.
1,2-propanediol
is
obtained
from
glycerol
by
the
sequential
processes
of
dehydrogenation-hydrogenation via hydroxiacetone, also glycerol hydrogenolysis can be carried out in both liquid and vapor phases. Liquid process takes place at low pressure producing mainly 1,2-propanediol and 1,3-propanediol in presence of supported catalysts such as Rh [18], Ru [19], or Pt [20]. On the other hand, glycerol hydrogenolysis in vapor phase is catalyzed by Cu at high hydrogen pressure [21], but in this process lateral reactions occur and different reaction by-products are obtained. In order to overcome this problem a two-step process has been proposed, thus dehydrogenation is first performed under vacuum conditions and then hydrogenation is carried out at high hydrogen pressure [22-23].
An efficient two steps process for selective production of 1,2-propanediol from glycerol was developed [24]. The reaction is carried out in vapor phase on a copper metallic catalyst at ambient pressure of hydrogen. In the first step hydroxyacetone is produced by glycerol dehydrogenation and then 1,2-propanediol is produced by hydrogenation of hydroxyacetone. Also, due to both the high temperature required for the dehydrogenation reaction and the improved selectivity at low temperatures for the hydrogenation reaction, the reactor configuration has a gradient temperature ranging from 200 C (on the reactor top) to 120 C (on the reactor bottom). Thus, a molar selectivity of 96% to 1,2-propanediol with total glycerol conversion is achieved.
77
The simplified flowsheet for 1,2-propanediol production from glycerol is shown in Figure 7.3. Glycerol is diluted at 30 g/L and then this stream is heated up to the reaction temperature. This stream is fed to the dehydrogenation-hydrogenation reactor, besides a hydrogen stream at 200 C is fed in a volumetric ratio of 1/141 (glycerol/H 2 ) to the reactor. The gradient reactor is available to convert glycerol completely to 1,2-propanediol, hydroxyacetone, ethylene glycol, and methane, as is shown by the equation system (7.5) to (7.7).
Purge D-1 R-1 M-1 Glycerol Water M-2 Fresh H2 H-2 Reactive H2 Recycled H2 Condensate Bottoms 1 Bottoms 2 H-1 Reactives Products HE-1 Sep-1 Distillate 1 Distillate 2 DC-1 DC-1
Figure 7.3. Simplified flowsheet for 1,2-propanediol production by hydrogenolysis. M-1: Mixer I; H-1: Heater I; R-1: Hydrogenolysis reactor; HE-1: Heat exchanger; Sep-1: Gas-lquid separator; D-1: Divisor; DC-1: Distillation column I; DC-2: Distillation column II; M-2: Mixer II; H-2: Heater III.
C 3 H 8 O 3 C 3 H 6 O 2 + H 2 O (Equation 7.5)
C 3 H 6 O 2 + H 2 C 3 H 8 O 2 (Equation 7.6)
C 3 H 6 O 2 + 2H 2 C 2 H 6 O 2 + CH 4 (Equation 7.7)
1,2-propanodiol purification process starts with a flash operation at 30 C and 50 bar. Then, the gas stream containing mainly H 2 is recycled and mixed with fresh H 2 . The resultant stream of H 2 must be heated up to reaction temperature and then fed to the reactor. The liquid stream obtained from the flash operation is purified using two distillation columns. In the first one, most of the water quantity is retired meanwhile in the second distillation column the remaining water and the non-converted hydroxyacetone are retired.
78
During the simulation of the hydrogenolysis process, the fresh feed glycerol stream was heated up to 200 C. Then, a hydrogen stream at the reaction temperature was fed to the reactor at a volumetric ratio of 1/141 (glycerol/H 2 ) and glycerol is completely converted in the hydrogenolysis reactor. The purification process for 1,2-propanediol requires an evaporation process at 30 C and 50 bar, in which a gaseous stream containing mainly H 2 (99.9 mol %) was recycled and mixed with fresh H 2 . The liquid stream obtained by the bottom stream from the evaporation process was purified by mean of two distillation columns. Thus, 99.96 % of water was discarded using the Distillation Column I, and both the remaining water and no-converted hydroxyacetone were obtained by the distillated stream in the Distillation Column II. In this way, 1,2-propanediol at 99 wt % was achieved. The main simulation results for the hydrogenolysis process are shown in Table 7.3.
Table 7.3. Simulation results for hydrogenolysis process from glycerol Stream Reactives Temp. (K) 200 Products 200 1 Condensate Bottoms 1 30 50 183.6 1 Bottoms 2 186.7 1 Recycled H 2 30 50 Reactive H 2 200 1
Pressure (bar) 1 Comp. (wt%) Glycerol 1,2-PD Hydroxyacetone Ethylene glycol Water H2 Total Flow Rate (Kg/h) 33333 0,3 0 0 0 0,7 0
0 0,020 0,001 0 0,069 0,911
0 0,240 0,007 0 0,753 0
0 0,969 0,029 0,001 0,001 0
0 0,99 0,01 0,001 0 0
0 0 0 0 0,008 0,992
0 0 0 0 0,007 0,992
408099
33398
8263
7997
355966
374765
7.5 Economic assessment

Operational and capital costs were disaggregated by raw material, utilities, operation labor, maintenance, operating charges, plant overhead, general and administration costs, and depreciation cost, as shows the Table 7.4 for each glycerol conversion process [2532]. Besides, all costs were normalized and their shares are shown in Table 7.5.
79
Table 7.4. Production costs for glycerol conversion to added-value Production costs (US$/L of product) Raw materials Utilities Operating labor Maintenance Operating charges Plant Overhead General and Administrative Depreciation of capital Total costs Sale Price 0,0555 0,8274 1,110 0,0646 1,7935 1,779 1,069E-4 2,368E-4 2,498E-4 0,0192 0,2682 0,4200 Acrolein at 92 wt % 0,2927 0,3067 0,0033 0,0116 0,0008 0,0069 0,1499 Acrolein at 98.5 wt % 0,2920 1,2006 0,0042 0,0123 0,0011 0,0082 0,2105 4,777E-05 3,731E-05 1,129E-05 1,031E-05 2,823E-06 1,080E-05 9,625E-06 0,1203 0,0813 0,0073 0,0066 0,0081 0,0063 0,0191 Hydrogen 1,2-propanediol
Table 7.5. Percentage of Production costs for glycerol conversion to added-value Production costs (US$/L of product) Raw materials Utilities Operating labor Maintenance Operating charges Plant Overhead General and Administration Depreciation of capital Acrolein at 92% 35.38 37.07 0.40 1.40 0.10 0.83 18.12 6.71 Acrolein at 98.5% 16.28 66.94 0.23 0.69 0.06 0.46 11.74 3.60 20.17 15.75 4.77 4.35 1.19 4.56 4.06 45.14 44.85 30.31 2.72 2.46 3.02 2.35 7.12 7.16 Hydrogen 1,2-propanediol
Although glycerol at technical grade was considered as raw material for the three technological schemes, some differences in the raw material costs can be observed due to the differences among yield, selectivity, and products density (see Table 7.4). For
80
instance, because of the dehydration process had the lowest selectivity; the highest raw material cost was obtained for acrolein. Also, due to hydrogen is only product here obtained in gas phase not only the lowest raw material cost but also the lowest production cost in US$/L were obtained for the gasification process. Then, in order to compare the three technological schemes, it was necessary to include the commercial sale price for each product as shown Table 4. Even more, the shares of each item allow identifying the main economical resource consumers, as shown Table 7.5.
Two qualities of acrolein are observed in Table 7.4, there are 92 and 98.5 wt %. These assessment were performed because of the acrolein production process at 98.5 wt % requires a powerful coolant system which implies high operational costs; and thus its total production cost is higher than the commercial sale price. On the other hand, since the service cost to produce acrolein at 92 wt % is only the 25 % required to obtain acrolein of high purity, the total production cost for this process is lower than its commercial sale price. During the acrolein production at 92 wt %, most of the production costs are represented by raw materials and services which totaling 72 % of the total production cost. Meanwhile for acrolein production at 98.5 wt %, only the services contributing the 67 % of total production cost. Thus, the production process of acrolein at high purity is not economically viable.
Although in most of the chemical process the raw material cost represents near to 50 % of the total production cost, in the case of hydrogen production from glycerol the sum of both raw material and services costs were almost 36 % of the total production cost. But, for this process the main investment is represented by the process units since the equipment depreciation is 45.14% of the total production cost. The high depreciation cost occurs because extreme operational conditions (i.e., high temperatures and pressures) are required during this process.
Finally for 1,2-propanediol production the raw material and services costs represent the most share of the total production cost, which add 75%. And then, the equipment depreciation cost is 16.7 % of the total production cost. These values are typical for a chemical process.
81
By comparing the ratio of the commercial sale price respect to the obtained total production cost (i.e., sale/production costs), a ratio of 1.055 was found for hydrogen production, followed by the acrolein process with a ratio of 1.34, and the highest value obtained was for 1,2-propanediol production at a ratio of 1.57. Thus, the production of 1,2propanediol could generate the highest economical return for glycerol conversion into added-value products among the analyzed processes.
7.6 Conclusions
Acrolein, hydrogen, and 1,2-propanediol, are three of the most commercially important products obtained from glycerol, due to their applications, established market, and sale prices. Here the technological schemes to produce these compounds were designed, simulated, and economically assessed. Thus, simulation results showed that all the processes are technologically feasible reaching high purity of product. Also, acrolein production was found to be viable at a purity of 92 wt %, but do not at a purity of 98.5 wt %. Finally, both hydrogen and 1,2-propanediol production processes are also economically viable, where the last one generates the highest profit margin.
References
[1] Ott, L., Bicker, M., Vogel, H., 2006. Catalytic dehydration of glycerol in sub- and supercritical water: a new chemical process for acrolein production. Green Chem. 8, 214220. [2] Chai, S.-H., Wang, H.-P., Liang, Y., Xu, B.-Q., 2008. Sustainable production of acrolein: Preparation and characterization of zirconia-supported 12-tungstophosphoric acid catalyst for gas-phase dehydration of glycerol. Appl. Catal., A 353, 213-222. [3] Atia, H., Armbruster, U., Martin, A., 2008. Dehydration of glycerol in gas phase using heteropolyacid catalysts as active compounds. J. Catal. 258, 71-82. [4] Tsukuda, E., Sato, S., Takahashi, R., Sodesawa, T., 2007. Production of acrolein from glycerol over silica-supported heteropoly acids. Catal. Commun. 8, 13491353.
82
[5] Lili, N., Yunjie, D., Weimiao, C., Leifeng, G., Ronghe, L., Yuan, L., Qin, X., 2008. Glycerol dehydration to acrolein over activated carbon- supported silicotungstic acids. Chin. J. Catal. 29, 212-214. [6] Chai, S.-H., Wang, H.-P., Liang, Y., Xu, B.-Q., 2007. Sustainable production of acrolein: Gas-phase dehydration of glycerol over Nb 2 O 5 catalyst. J. Catal. 250, 342-349. [7] Watanabe, M., Iida, T., Aizawa, Y., Aida, T. M., Inomata, H., 2007. Acrolein synthesis from glycerol in hot-compressed water. Bioresour. Technol. 98, 12851290. [8] Pathak, K.D., 2005. Catalytic conversion of glycerol to value-added liquid chemicals. Thesis for the degree of Master of Science, University of Saskatchewan, Canada. [9] Cortright, R.D., Davda, R.R., Dumesic, J.A., 2002. Hydrogen from catalytic reforming of biomass-derived hydrocarbons in liquid water. Nature 418, 964-966. [10] Demirbas, A., 2002. Fuel properties of hydrogen, liquefied petroleum gas, and compressed natural gas for transportation. Energ. Source. 24, 601610. Franco, C., Pinto, F., Gulyurtlu, I., Cabrita, I., 2003. The study of reactions influencing biomass steam gasification process. Fuel 82, 835-842. [11] Franco, C., Pinto, F., Gulyurtlu, I., Cabrita, I., 2003. The study of reactions influencing biomass steam gasification process. Fuel 82, 835-842. [12] Chaudhari, S.T., Bakhshi, N.N., 2002. Steam gasification of chars and bio-oil, In: Report to bioenergy development program renewable energy branch: Energy, mines, and resources. Canada, Ottawa, 396-436. [13] Fernndez, Y., Arenillas, A., Dez, M.A., Pis, J.J., Menndez, J.A., 2009. Pyrolysis of glycerol over activated carbons for syn-gas production. J. Anal. Appl. Pyrolysis 84, 145150. [14] Garcia, L., French, R., Czernik, S., Chornet, E., 2000. Catalytic steam reforming of bio-oils for the production of hydrogen: effects of catalyst composition. Appl. Catal., A 201, 225239. [15] Valliyappan, T., Bakhshi, N.N., Dalai, A.K., 2008. Pyrolysis of glycerol for the production of hydrogen or syn gas. Bioresour. Technol. 99, 44764483.
83
[16] Mozaffarian, M., Deurwaarder, E.P., Kersten, S.R.A, 2004. Project: green gas (SNG) production by supercritical gasification of biomass, pp. 71, 2004. [17] Dasari, M.A., Kiatsimkul, P.-P., Sutterlin, W.R., Suppes, G.J., 2005. Low-Pressure hydrogenolysis of glycerol to propylene glycol. Appl. Catal., A. 281, 225-231. [18] Chaminand, J., Djakovitch, L., Gallezot, P., Marion, P., Pinel, C., Rosie, C., 2004. Glycerol hydrogenolysis on heterogeneous catalysts. Green Chem. 6, 359-361. [19] Maris, E.P., Davis, R.J., 2007. Hydrogenolysis of glycerol over carbon-supported Ru and Pt catalysts. J. Catal. 249, 328-337. [20] Kurosaka, T., Maruyama, H., Naribayashi, I., Sasaki, Y., 2008. Production of 1,3propanediol by hydrogenolysis of glycerol catalyzed by Pt/WO 3 /ZrO 2 . Catal. Commun. 9, 1360-1363. [21] Wang, S., Liu, H., 2007. Selective hydrogenolysis of glycerol to propylene glycol on CuZnO catalysts . Catal. Lett. 117, 62-67. [22] Chiu, C.-W., Tekeei, A., Ronco, J.M., Banks, M.-L., Suppes, G.J., 2008a. Reducing byproduct formation during conversion of glycerol to propylene glycol. Ind. Eng. Chem. Res. 47, 6878-6884. [23] Chiu, C.-W., Tekeei, A., Sutterlin, W.R., Ronco, J.M., Suppes, G.J., 2008b. Lowpressure packed-bed gas phase conversion of glycerol to acetol. AIChE J. 54, 2456-2463. [24] Akiyama, M., Sato, S., Takahashi, R., Inui, K., Yokota, M., 2009. Dehydration hydrogenation of glycerol into 1,2-propanediol at ambient hydrogen pressure. Appl. Catal., A 371, 60-66. [25] Cardona, C.A, Snchez, O.J., 2006. Energy consumption analysis of integrated flowsheets for production of fuel ethanol from lignocellulosic biomass. Energy 31, 24472459. [26] Cardona, C.A., Posada, J.A., Quintero, J.A., 2010. Use of agroindustrial subproducts and wastes: Glycerin and Lignocellulosics, first ed. Artes Grficas Tizn, Manizales. (In Spanish).
84
[27] Gutirrez, L.F., Snchez, O.J., Cardona, C.A., 2009. Process integration possibilities for biodiesel production from palm oil using ethanol obtained from lignocellulosic residues of oil palm industry. Bioresour. Technol. 100, 1227-1237. [28] Posada, J.A., Cardona, C.A., 2010. Design and analysis of fuel ethanol production from raw glycerol. Energy, doi:10.1016/j.energy.2010.07.036. [29] Posada, J.A., Cardona, C.A., 2010b. Validation of glycerin refining obtained as a byProduct of biodiesel production. Ingeniera y Universidad. 14, 2-27. (In Spanish). [30] Posada, J.A., Naranjo, J.M., Lpez, J.A., Higuita, J.C., Cardona, C.A., 2010a. Design and analysis of poly-3-hydroxybutyrate production processes from crude glycerol. Process Biochem., doi:10.1016/j.procbio.2010. 09.003.. [31] Posada, J.A., Cardona, C.A., Rincn, L.E., 2010b. Sustainable biodiesel production from palm using in situ produced glycerol and biomass for raw bioethanol. In 32nd symposium on biotechnology for fuels and chemicals. Clearwater Beach, Florida. [32] Quintero, J.A., Montoya, M.I., Snchez, O.J., Giraldo, O.H., Cardona, C.A., 2008. Fuel ethanol production from sugarcane and corn: Comparative analysis for a Colombian case. Energy 33, 385399.
8. Study Cases of Biochemical Conversion of

Glycerol
This chapter presents the process design, simulation, and economical assessment of six different possibilities for glycerol transformation by fermentation. For the production of 1,3propanediol, a kinetic model was used allowing optimizing the fermentation stage by three different approaches. For ethanol, poly-3-hydroxybutirate, lactic acid, succinic acid, and propionic acid production, a yield approach was used in all cases. Thus, several scenarios were analyzed in each case depending on the glycerol fermentation stage or on the downstream process.
8.1 1,3-Propanediol production

Although 1,3-propanediol could be biologically produced from glycerol by several bacterial strains such as: Klebsiella pneumoniae, Citrobacter freundii, Enterobacter agglomerans, Clostridium butyricum, and Clostridium acetobutylicum [1-2]; the K. pneumoniae and C. butyricum strains are the most promising bacterial because of their high yield, productivity, and resistance to both substrate and product inhibition. Among these two bacteria, K. pneumoniae DSM-2026 has been presented as one of the most appropriate bacterial strain for glycerol fermentation to 1,3-propanediol [3]. First of all, here the fermentation process of glycerol to 1,3-propanediol by K. pneumoniae is analyzed in one and two continuous stages.
The material balances for continuous glycerol fermentation in one single stage are solved using two independent variables namely the glycerol concentration in the feed stream and the dilution rate. Since, acetic acid and ethanol are also produced during glycerol fermentation to 1,3-propanediol by K. pneumoniae, the material balances in a dynamic state needed to be solved for biomass, substrate, and all the obtained products as shown in equations (8.1) to (8.5).
86
dX iOut = Di ( X iIn X iOut ) + i X iOut dt

dC
Out G ,i In Out = Di ( C G ,i C G ,i ) q G ,i X iOut
(8.1)
dt
Out dC PD ,i
(8.2)
dt
dC
Out HAc , i
In Out = Di ( C PD ,i C PD ,i ) + q PD ,i X iOut
(8.3) (8.4) (8.5)
dt
dC
Out EtOH ,i
In Out = Di ( C HAc ,i C HAc ,i ) + q HAc ,i X iOut
dt
In Out = Di ( C EtOH ,i C EtOH ,i ) + q EtOH ,i X iOut
Where: X, C G , C PD , C HAc , and C EtOH , are the concentrations for biomass (g/L), glycerol (mol/L), 1,3-propanediol (mol/L), acetic acid (mol/L), and ethanol (mol/L), respectively. D is the dilution rate (i.e., ratio between volumetric flow and reactor volume) (h-1), is the specific rate of cellular growth (h-1), q G is the specific rate of glycerol consumption, and q PD , q HAc , and q EtOH are the generation rates of each product (h-1). Subscript i, indicates the fermentation stage for a multistage system. In the i fermentation stage, In and Out superscripts indicate the in and the out conditions respectively.
The kinetic model of glycerol fermentation by K. pneumoniae has been previously explained [4-6]. Specific rates of cell growth, substrate consumption, and products formation are given in the equations (8.6) to (8.11).
i = max
C G ,i C 1 G,i C G ,i + K S CG C 1 PD ,i C PD C 1 HAc ,i C HAc C 1 EtOH ,i C EtOH (8.6)
q G ,i = m G +
i
Y
m G
m + q G
C G ,i
C G ,i + K S
(8.7)
m m q PD ,i = m PD + i * YPD + q PD
C G ,i
C G ,i + K PD
(8.8)
m m q HAc ,i = m HAc + i * YHAc + q HAc
C G ,i
C G ,i + K HAc
(8.9) (8.10)
q EtOH ,i = qG ,i Y(m / G ),i EtOH
Y(m / G ),i = EtOH
b1 b2 + c1 + Di CG ,i c 2 + Di CG ,i
(8.11)
8. Study Cases of Biochemical Conversion
87
Equation (8.6) describes the specific rate of cell growth which represents a kinetic model
* * * with inhibition by both substrate and products. The kinetic parameters C G , C PD , C HAc and * C EtOH
are the critical concentrations (i.e., the concentration where biological activity is
stopped). The required parameters to solve the kinetic model (valid at 37 C and at neutral pH [4, 6]) are: Maximum Specific Growth Rate, max =0.67 h-1 and the Monod Saturation Constant, Ks=2.8e-4 gmol/L.
Constants for specific rate of substrate consumption and product formation are: m G = 2.20 e-3, m PD =-2.69e-3, m HAc =-9.7e-4, Ym Gly = 8.2, Ym PD =6.769e-2, Ym HAc =3.307e-2, qm Gly =2.858e-2, qm PD =2.659e-2, qm HAc =5.74e-3, K* Gly =1.143e-2, K* PD =1.55e-2, K* HAc = 8.571e-2. Also, b 1 , b 2 , c 1, and c 2 constants in equation (8.11) are: 2.5e-5, 5.18e-3, 6e-5, and 5.045e-2 mol/(L*h), respectively.
The critical concentrations required in equation (8.6) were taken from [2]. These concentrations have an average global deviation of 8.6% for 29 stable states of glycerol fermentation by K. pneumoniae and C. butyricum [5-6]. The critical concentrations for glycerol, 1,3-propanediol, acetic acid, and ethanol are: 2.012, 0.8975, 0.7798, and 0.3975 mol/L, respectively.
On the other hand, glycerol fermentation to 1,3-propanediol by K. pneumoniae presents a metabolic overflow of products causing dynamic phenomena of non-lineal behavior such as multiplicity of steady states, hysteresis, and oscillations [5]. Thus, the operational conditions that take to multiplicity of steady states were determined.
In order to obtain the best performance in the first fermentation stage, the volumetric productivity was optimized using the Levenverg-Marquardt method [7]. Volumetric productivity is one of the most important functions to be optimized from an operative point of view, since it implies a high production in a small reactive volume on a short period of time. The volumetric productivity is shown in equation (8.12), where Pr 1 is the volumetric productivity in (mol/(L*h)), C PD1 is the outlet 1,3-propanediol concentration in (mol/L),and D 1 is the dilution rate in (h-1). The first fermentation stage in this equation is indicated by the subscript 1.
88
Pr1 = C PD1D1
(8.12)
In this case, when multiplicity occurs, the steady states with the highest concentration of 1,3-propanediol were considered. This is to ensure that the steady states with the higher volumetric productivities are evaluated. Regardless of the inhibition phenomenon caused by substrate or products, fermentation systems could be successfully performed in two continuous stages reaching simultaneously a high productivity and a high product concentration in the first and second fermentation stages respectively [4, 8].
To assess the glycerol fermentation process in two continuous stages, three optimization models were analyzed. The first optimization model is a sequential procedure in which the outlet stream from the first fermentation stage (operated at optimal conditions) is directly fed on the second one. The volumetric productivity on the second fermentation stage was calculated as a function of both the dilution rate and the achieved change on the 1,3propanediol concentration as shown in equation (8.13). Also, the maximum outlet concentration of 1,3-propanediol on the second fermentation stage was determined by the Levenverg-Marquardt method.
Pr2 = D2 ( C PD 2 C PD1 )
(8.13)
Although in the second optimization model the objective function is the volumetric productivity in the second fermentation stage (see equation 8.13), only the optimal dilution rate in the first fermentation stage was kept unchanged. Thus, the volumetric productivity was calculated as a function of both the dilution rate on the second stage and the feed glycerol concentration on the first stage.
Finally, in the third optimization model the productivity of both fermentation stages were simultaneously considered as a function of the dilution rate in both stages and the feed concentration of glycerol in the first fermentation stage. The objective function in this model is the product of productivities between both fermentation stages (P r3 ), as shown in equation (8.14).
Pr3 =Pr1 Pr2
(8.14)
89
Figure 8.1 shows the multiple steady states, hysteresis loops, and the wash out line for the continuous glycerol fermentation. Multiplicity of steady states and hysteresis loops were studied considering the dilution rate as a parameter from a low glycerol concentration up to the wash out conditions. In the hysteresis loops when the feed glycerol concentration increases, low 1,3-propanediol yields are obtained. Moreover, when the feed glycerol concentration decreases, high 1,3-propanediol yields are acquired. The latter condition corresponds to the upper curves in the hysteresis loops. The "wash out" line indicates the extreme operational conditions where the dilution rate equals to the cellular growth rate (
i = Di ).
Thus, volumetric productivity was calculated as a function of both the feed glycerol concentration and the dilution rate, using a polygon mesh (partition) from low feed glycerol concentration up to the wash out line (see Figure 8.2.a.). The conditions that generate the higher 1,3-propanediol concentration were selected at the multiplicity of steady states region. As a consequence, a discontinuity between A and B was obtained and the higher volumetric productivity values were located at the right side of these points.
Figure 8.1. Hysteresis loops and multiple steady states. Concentrations of: a) Biomass (g/L). b) Residual Glycerol (mol/L). c) 1,3-Propanediol (mol/L). d) Acetic Acid (mol/L) and e) Ethanol (mol/L). Vertical lines indicate the limits of the multiple steady states region. Dotted lines show the "wash out" conditions for each dilution rate.
90
Figure 8.2.a) 1,3-propanediol volumetric productivity, (the column in the right side gives the scale). b) Region of multiplicity of steady states, optimal productivity for each dilution rate, global optimal productivity, and wash-out line.
For each dilution rate exists a feed glycerol concentration that generates a maximum volumetric productivity as obtained by the polygon mesh distribution shown in Figure 2.a. In order to obtain the global optimum for volumetric productivity both independent variables (i.e., feed glycerol concentrations and dilution rate) must be simultaneously considered. Since the highest volumetric productivity is close the steady states region, this area must be considered when selecting the initial estimated to apply in the optimization method.
In order to find the conditions for feed glycerol concentration and the dilution rate that generates the highest volumetric productivity, the Levenverg-Marquardt optimization method was employed [7]. Since volumetric productivity is a non-continuous function, the initial estimated for both the feed glycerol concentration and the dilution rate must be higher than the conditions obtained in the A point.
The multiple steady states region for glycerol fermentation was reported by Xiu et al. [26], but the used critical concentration parameters have a smaller fitting than the ones used in this work. The optimal conditions for glycerol fermentation in one continuous stage are as follows: 0.2821 h-1 for the dilution rate and 0.6882 mol/L for the feed glycerol concentration, with a volumetric productivity of 0.1076 mol/(L*h).
91
Additionally, the obtained outlet concentration of 1,3-propanediol is 0.3811 mol/L. This optimal volumetric productivity is outside the multiple steady states region as shown in Figure 8.2.b. Since this volumetric productivity is very close to the multiple steady states region, minimum requirements in the automatic control of the equipments are recommended. Additionally, Figure 8.2.b. shows the wash out conditions and the optimal volumetric productivity for each dilution rate.
The kinetic model given for the fermentation system by equations (8.1) to (8.12) is equally applicable for simulation of a fermentation process with two continuous stages. Thus, three models to optimize the second fermentation stage were used.
In the first model, optimal conditions for the first fermentation stage were used to calculate the volumetric productivity on the second fermentation stage, but this function increases proportionally to the dilution rate, contrary to the behavior shown by the concentration of 1,3-propanediol (see Figure 8.3). The optimal concentration of 1,3-propanediol was 0.4126 mol/L at a dilution rate of 1.9850 h-1, with a productivity of 0.0625 mol/(L*h).
Figure 8.3. 1,3-Propanediol productivity and concentration in the second fermentation stage.
92
In the second optimization model the dilution rate for the first stage was kept from the first model and the productivity in the second fermentation stage was optimized as a function of the feed glycerol concentration. The reached productivity is 0.1128 mol/(L*h), at a dilution rate on the second stage of 0.79 h-1, with a feed glycerol concentration on the first stage of 0.8817 mol/L. The outlet concentration of 1,3-propanediol from the first and second stages are 0.3405, and 0.4833 mol/L, respectively. Since each dilution rate has its own optimal productivity, it was necessary to calculate the global optimal productivity using the Levenverg-Marquardt method (see Figure 8.4). Optimal productivities for each dilution rate in the second fermentation stage are represented by the discontinuous curve. Also, the P point indicates the global optimal productivity in the second stage.
Figure 8.4. Volumetric productivity in the second fermentation stage using the optimal dilution rate obtained by the model 1 for the first fermentation stage, (the column in the right side gives the scale).
Finally, in the third optimization model the productivities of both fermentation stages were simultaneously optimized considering the two dilution rates and the feed glycerol concentration as independent variables. The obtained results using the optimal dilution rate in the first fermentation stage are shown in Figure 8.5. The optimum product of productivities for each dilution rate in the second fermentation stage is represented by the discontinuous curve.
93
Figure 8.5. Product of productivities of both fermentation stages using the optimal dilution rate obtained by the model 1 for the first fermentation stage, (the column in the right side gives the scale).
The Q point indicates the global optimum for the product of productivities. This point corresponds to a dilution rate of 0.2821 h-1 and 3.08 h-1 in the first and second stages respectively, and a feed glycerol concentration of 0.7362 mol/L in the first stage. The reached product of productivities was 0.0116 (mol/(L*h))2 where the outlet concentration of 1,3-propanediol was 0.4124 mol/L and the global molar yield was 0.5602 1,3-PD/ Gly.
Table 8.1 shows the results of the three different models used to optimize the fermentation of glycerol in two stages. The highest global yield and volumetric productivity in the first fermentation stage were generated using the sequential optimization model. On the other hand, the volumetric productivity in the second fermentation stage was optimized using the combined optimization model under the optimal dilution rate in the first fermentation stage. But, when the dilution rate in the first fermentation stage was decreased at 0.25 h-1, the best final concentration of 1,3-propanediol was obtained as shown in Table 8.1.
94
Table 8.1. Results summary for each optimization model. Optimization CS 0 a Model Sequential. Combined Combined Combined
a
D1
D2c
CPD 1 d
CPD 2 e Y PD/S f
Pr 1 g
Pr 2 h
Pr 3 i
0,688 0,282 1,985 0,381 0,882 0,282 0,790 0,341 0,932 0,250 0,940 0,396 0,852 0,300 0,720 0,311
0,413 0,483 0,512 0,469 0,412
0,599 0,548 0,549 0,551 0,560

b
0,107 0,096 0,099 0,093 0,106
0,063 0,113 0,109 0,114 0,109
6,72e-3 1,09e-2 1,08e-2 1,06e-2 1,15e-2

d
Simultaneous 0,736 0,282 3,080 0,377

-1 c
Feed glycerol concentration in the first stage in (mol/L),
Dilution rate in the first 1,3g e
fermentation stage in (h ),
Dilution rate in the second fermentation stage in (h-1),

f
propanediol concentration in the first fermentation stage in (mol/L), concentration in the second fermentation stage in (mol/L),
1,3-propanediol
h
Global fermentation yield,

i
Volumetric productivity in the first fermentation stage in (mol/(L*h)), (mol/(L*h))2.
Volumetric
productivity in the second fermentation stage in (mol/(L*h)), Product of productivities in
Also, when the dilution rate in the first fermentation stage was increased to 0.30 h-1, the highest volumetric productivity in the second stage was obtained as shown in Table 8.1. Then, the volumetric productivity in the second fermentation stage can be optimized only at a specific dilution rate in the first fermentation stage.
Using the simultaneous optimization model a high volumetric productivity in both fermentation stages and the highest product of productivities were obtained. Also, the obtained value for the optimal dilution rate in the first stage was the same using the sequential optimization model. Thus, the use of a sequential optimization model allowed obtaining the highest global yield for 1,3-propanediol (0.599) and the maximum volumetric productivity in the first fermentation stage (0.1075 mol/(L*h)), whereas the highest 1,3propanediol outlet concentration (0.512 mol/L) was observed when the combined optimization model was employed. Meanwhile, using the simultaneous optimization model showed both: high volumetric productivities in the two fermentation stages and the highest product of productivities (0.01157 (mol/(L*h))2). In this way, for the fermentation of glycerol in two continuous stages, three different operational configurations are available depending on the desired process objective namely global yield, 1,3-propanediol outlet concentration, or high simultaneous productivity.
95
On the other hand, the main problem for designing the downstream process for 1,3propanediol recovery and purification from the fermentation broth is the high hydrophilicity and high boiling point of 1,3-propanediol. The purification scheme proposed here is based on integrated reaction-separation units to carry out the recovery of 1,3-propanediol. The first integrated stage is a reactor-extraction process where the hydrophilic nature of 1,3propanediol is changed by the acetylation reaction with iso-butyl aldehyde, which produces 2-iso-propyl-1,3-dioxane as shown in Figure 8.6.
H3C HC H3C O O CH2 CH2 CH2
H3C
HO OH
+
H3C
H2O
Figure 8.6. Acetylation reaction of 1,3-propanediol with iso-butyl aldehyde to 2-iso-propyl1,3-dioxane
The 2-iso-propyl-1,3-dioxane has a hydrophobic character which is dragged into the organic phase containing mainly iso-butyl aldehyde. This aldehyde acts as both reagent and solvent for the reactive-extraction process [9-11]. Subsequently, the 1,3-propanediol is recovered by reactive distillation of 2-iso-propyl-1,3-dioxane and water though out the reverse reaction of cyclical acetylation. Thus, 1,3-propanediol at high purity is obtained by bottom while the iso-butyl aldehyde is recovered by distillated and it is able to be reused in the downstream process. Then, the aldehyde is not consumed in the purification process.
Based on the fermentation results obtained previously for the fermentation stage, three scenarios are selected in order to perform the process design and analysis for glycerol conversion to 1,3-propanodiol.
The first scenario considers conditions of optimal volumetric productivity in the first fermentation stage and optimal final concentration of 1,3-propanediol according to the sequential model presents in Table 8.1. The second scenario considers conditions of the highest final concentration of 1,3-propanediol and the highest productivity in the second fermentation stage according to the second combined model presented in Table 8.1. And finally, the third scenario considers the optimal global productivity having into account both fermentation stages according to simultaneous model presented in Table 8.1.
96
Calculated results of the two-stage fermentation process for each scenario are shown in Table 8.2, where the column named Feed indicates the fed stream to the first fermentation stage, Products 1 corresponds to the fermentation products stream from the first fermentation stage, and Products 2 is the fermentation products stream from the second fermentation stage.
The normalized stoichiometry for the fermentative reactions are shown in Table 8.3 according to fermentation results obtained for each scenario (Scen. in Table 8.3) and each fermentation tank (Ferm. in Table 8.3). Also, the molecular formula used for K. pneumoniae is CH 1.75 O 0.46 N 0.23 . Table 8.2. Fermentation results for the three considered scenarios Concentration mmol/L Feed Products 1 Products 2
Scenario 1 Glycerol 13PD AcAc EtOH Biomass (g/L) 688.2 0.0 0.0 0.0 0.0 Scenario 2 Glycerol 13PD AcAc EtOH Biomass (g/L) 932.0 0.0 0.0 0.0 0.0 Scenario 3 Glycerol 13PD AcAc EtOH Biomass (g/L) 736.2 0.0 0.0 0.0 0.0 147.8 376.8 111.7 33.4 2.6267 92.1 412.2 121.7 37.3 2.8725 330.1 395.1 118.2 23.6 2.4924 153.6 511.4 149 35.1 3.1489 77.6 381.1 109.6 43.4 2.7741 8.9 419.7 122.5 51.3 3.1788
97
Table 8.3. stoichiometric reactions for each scenario and each fermentation stage Reactions Molecular Weight Scen 1, Ferm 1 Scen 1, Ferm 2 Scen 2, Ferm 1 Scen 2, Ferm 2 Scen 3, Ferm 1 Scen 3, Ferm 2 Glycerol 92.09 1.0 1.0 1.0 1.0 1.0 1.0 Residual Gly 13PD 92.09 0.1067 0.1045 0.3399 0.4523 0.1896 0.4832 76.09 AcAc 60.05 EtOH 46.07 Biomass 23.94 0.1593 0.1984 0.1072 0.0807 0.1408 0.0891
0.5238 0.1506 0.0597 0.4530 0.1514 0.0927 0.4069 0.1217 0.0243 0.3424 0.0907 0.0339 0.4834 0.1433 0.0429 0.3071 0.0868 0.0338
The simplified flowsheet for 1,3-propanediol production from raw glycerol is shown in Figure 8.7. Cell mass contained in the fermentation is withdrawn throughout a centrifugation process. Then, the clarified fermentation broth is mixed with iso-butyl aldehyde in a weight ratio of 5/1 (iso-butyl aldehyde /1,3-propanediol). The reactiveextraction process takes place at 10 C since better distribution coefficients are obtained at lower temperatures. From the reaction extraction unit two product streams are obtained.
Aqueous stream should be purified in order to recover significant amounts of both isobutyl aldehyde and 2-iso-propyl-1,3-dioxane. This stream is subjected to two distillation stages, in the first one the heterogeneous azeotrope iso-butyl aldehyde-water is obtained by distillated, and then iso-butyl aldehyde is obtained at 97.3 wt % by decantation. In the second distillation column the homogeneous azeotrope composed by 2-iso-propyl-1,3dioxane and water is obtained at the top of the distillation column. Then, the obtained azeotrope is mixed with the organic stream obtained during the reactive extraction process which contains mainly iso-butyl aldehyde and 2-iso-propyl-1,3-dioxane. The resulting mixture is distilled and the heterogeneous azeotrope iso-butyl aldehyde-water is once again obtained as the top product, and thus a mixture containing 2-iso-propyl-1,3dioxane and water is obtained. This stream is directly fed to the reactive distillation column as follows.
98
Metanol E-1 Raw Glycerol R-1 C-1 Water Dec-1 E-2
Waste DC-1 Water 1
Waste RII-1 Water 2 Glycerol (98wt%)
Water M-1
F-1
F-2
Glycerol (85wt%) Solids Organic Phase Aqueous glycerol Adsorbate Diluted Glycerol Dec-3 IBuAld Waste water Distillated-4 Distillated-3 DC-4 Water M-2 Waste water DC-3 Distillated-2 M-2 Aqueous Phase Solids DC-2 Organic Phase RE-1 Fresh IBuAld Cen-1
Fermentation Broth
Dec-2 Distillated-6 Distillated-5 DC-5 RDC
13PD
Bottoms-5
Bottoms-4
Bottoms-3
P-3
Bottoms-2
Figure 8.7. Simplified flowsheet for 1,3-propanediol production from raw glycerol. E: evaporator, R: reactor, C: centrifuge, Dec: Decanter, DC: distillation column, M: Mixer, F: fermentator, RE: Reactor extractor, RDC: Reactive distillation column.
In order to determine both the operational viability and the best configuration in the reactive-distillation tower (localization of the reaction zone), the Static Analysis is applied [12-16]. This methodology is the main tool for the qualitative study of the reactive distillation process which requires minimum initial information. Also it is based on both the thermodynamic topological behavior of reactive system and on the selection of stable state limits of highest conversion.
Static Analysis was developed by Serafimov et al [12] and has been sufficiently illustrated for Pisarenko et al [13] and validated in multiple reactive systems [14-16]. Some considerations should be made to carry out this analysis: (i) the reaction takes place under equilibrium conditions and (ii) the reactive distillation column operates to total both reflux and efficiency. In other words ( to conditions are considered. T /) localization of the reaction zone. he main
operation parameters are: the flow ratio of distillated to bottom (P/W) and the volume and
99
To carry out the thermodynamic topological analysis of the reactive system their singular points are characterized as shown Table 8.4 and the corresponding quaternary residue map curves is obtained as shown in Figure 8.8.
Table 8.4. Singular Points ** - Acetilation System of 1,3-PD* with 2iP13DO* Component Azeotrope_H 2 O-iBuAld* iBuAld* Azeotrope_H 2 O-2iP13DO* H2O 2iP13DO* 13PD* Type Heterogeneous Homogeneous Heterogeneous Homogeneous Homogeneous Homogeneous Temperature 61,35 C 64,10 C 92,87 C 100,00 C 138,12 C 214,40 C X1 0,2079 1 0,7449 1 1 1 X2 0,7921 --0,2551 -------
* iBuAld: iso-Butyraldehyde; 2iP13DO: 2-iso-Propil-1,3-Dioxano; 13PD: 1,3-Propanediol; H2O: Water. ** Singular Points at 1Atm.
Figure 8.8. Residue map curves for the reactive system.
Thus, only one distillation region is obtained with a bunch of residue curves starting from the azeotrope of minimum boiling point (iso-butyl aldehyde-water) and ending in corner of 1,3-propanediol. By direct separation (formulated distilled) three distillation subregions are obtained while in the case of indirect separation (bottoms formulated) two distillation
100
subregiones are founded. Thus, ten different possibilities for the feeding ratio of 2-isopropyl-1,3-dioxane to water to the reactive distillation column were analyzed (results no shown). Then, the feeding ratio that generates the highest products distribution (P/W) at a total conversion of 2-iso-propyl-1,3-dioxane (see Figure 8.9) corresponds to the 2-isopropyl-1,3-dioxane/water molar mixture of 0.3776/0.6224, as is shown in Figure 8.10. Also, based on both the chemical equilibrium and the residue curve maps, it was determined that the reactive zone must be localized in the stripping section of the reactive distillation column. Also, the product obtained in the distillated stream is the heterogeneous azeotrope iso-butyl aldehyde-water.
Then, in order to verify it was possible to obtain the conditions and the trajectory predicted by the Static Analysis method, simulations at both / (stages/ reflux ) and finite condition were carried out. Also, it was found that a self-extractive phenomenon (which can be understood as a non-lineal variation in the relative volatility of a system with the change of concentrations in a multicomponent mixture, see Figure 8.11) affects strongly the reactive distillation column performance. The analysis of the isovolatility curves allowed determining that a redistribution of the feeding streams lead to a high conversion keeping the configuration obtained by the Static Analysis method.
Figure 8.9. Direct separation with fed 0.377645/0.6223552iP13DO/Water
101
Figure 8.10. P/W ratio, Direct Separation (XF: 0.377645/0.622355-2iP13DO/water)
Figure 8.11. iso-Volatility curve (Wateriso-Butyraldehyde2-iso-Propil-1,3-Dioxane)
In this case, the water required as reactive to carry out the hydrolysis reaction of 2-isopropyl-1,3-dioxane is fed in five different stages to the reactive distillation column. For instance, in the Scenario 1 the reactive distillation column has 45 stages and the water stream is fed to the stages: 9, 15, 22, 29, 36, and 43. And the respective mass flows are: 7.56, 11.46, 15.36, 19.27, and 23.17 kg/h.
102
In this way high conversion levels of 2-iso-propyl-1,3-dioxane are obtained. But, because of the 1,3-propanediol purity ranges 83.3 and 93.9 wt %, a final distillation process is required in order to achieve a higher purity of 1,3-propanediol. A summary of the main simulation results for each scenario is given in Table 8.5.
Table 8.5. Summary of the main simulation results for 1,3-propanediol production from glycerol Scenario1 Dilgly Temperature K Mass Flow kg/hr WATER GLYCE-01 KPNEUMON 1,3-P-01 ACETI-01 ETHAN-01 ISOBU-01 2IP13DOX 9115.8 8536.2 578.9 0 0 0 0 0 0 8900.1 8536.2 6.45 27.16 273.67 62.95 20.15 0 0 920.0 14.28 0.273 0 1.633 0.387 1.335 491.32 410.77 Scenario2 Dilgly Temperature K Mass Flow kg/hr WATER GLYCE-01 KPNEUMON 1,3-P-01 ACETI-01 ETHAN-01 ISOBU-01 2IP13DOX 6766.1 6186.5 578.9 0 0 0 0 0 0 6594.4 6186.5 89.0 20.26 250.30 57.58 10.37 0 0 925.9 15.51 5.95 0 1.70 0.473 0.560 515.1 386.6 6847.5 6227.5 83.04 0 7.54 52.84 8.50 442.4 25.73 423.5 0 5.95 0 1.703 1.021 1.001 1.515 412.3 247.9 2.74 5.95 0 238.4 0.747 0 1 E-03 0.017 245.0 0.007 5.95 0 238.4 0.686 0 0 0.002 310 13PD2 310 Organic2 283.2 Aqueous1 IP13DO3 283.1 409.1 13PD3 444.4 13PD4 487.7 9270.3 8582.9 6.18 0 10.95 59.07 17.48 557.88 35.83 453. 8 0 0.273 0 1.63 1.08 3.74 0.449 446.60 272.6 0.752 0.273 0 251.1 1.08 0.003 0.183 19.24 252.7 0 0.273 0 250.83 0.980 0 0 0.656 310 13PD2 310 Organic2 283.2 Aqueous1 IP13DO3 283.1 404.2 13PD3 457.9 13PD4 486.8
103
Scenario3 Dilgly Temperature K Mass Flow kg/hr WATER GLYCE-01 KPNEUMON 1,3-P-01 ACETI-01 ETHAN-01 ISOBU-01 2IP13DOX 8531.1 7951.5 578.9 0 0 0 0 0 0 8338.9 7951.5 53.03 23.73 259.07 60.30 14.28 0 0 872.37 13.77 2.277 0 1.535 0.374 0.573 465.07 388.77 8687.6 7995.5 50.7 0 10.20 56.55 12.37 527.9 34.32 427.6 0 2.28 0 1.53 0.7 0 0 423.1 253.6 5.15 2.28 0 245.6 0.59 0 0.001 0.011 248.4 0.027 2.28 0 245.6 0.507 0 0 0 310 13PD2 310 Organic2 283.2 Aqueous1 IP13DO3 283.1 411.4 13PD3 424.7 13PD4 487
The final production of 1,3-propanediol is mainly related to the fermentation yield of both fermentation stages. Thus, while the decreasing order for the final concentration of 1,3propanediol after the second fermentation stage was Scenario 2 > Scenario 1 > Scenario 3 (see Table 8.2) and the decreasing order for the fermentative yield of glycerol to 1,3propanediol was Scenario 1 > Scenario 3 > Scenario 2 (see Table 8.3), the decreasing order for the actual production of 1,3-propanediol was Scenario 1 > Scenario 3 > Scenario 2 (see Table 8.5).
Otherwise, high recovery percentages were achieved for iso-butyl aldehyde, indicating that low requirements of fresh reactive are required. Higher differences are noticed when the whole technological scheme is analyzed. The maximum global molar yield from glycerol to 1,3-propanediol was obtained for the Scenario I, while the minimum was obtained for the Scenario 2. The relative difference between these two scenarios was 5.21 %, which was close to the relative difference for the fermentation yield obtained for the same both scenarios, 11.14 %. Thus, it can be stated that the technological performance of 1,3-propanediol production from raw glycerol depends mostly on the global conversion of substrate to the main product during the fermentation stage. See Table 8.6.
104
Table 8.6. Data representing the behavior of the downstream process Scenario 1 Reactive-Extraction Extraction efficiency (%) Distribution coefficient Loading (Z) Downstream Process Global 13PD recovery (%) IBuAld recovery (%) Global Process Global process yield from Glycerol to Lactic acid (%) 0.5244 0.4984 0.5134 91.66 98.66 95.26 99.48 94.80 99.54 96.00 204.09 0.177 96.99 203.39 0.182 96.06 204.91 0.177 Scenario 2 Scenario 3
8.2 Ethanol production

Ethanol can be produced from sugarcane [17], corn starch [17], sugar [18], molasses [19], cassava [20], wheat [21] or lignocellulosic biomass [22-25]. Glycerol fermentation by Escherichia coli produces a mixture containing predominantly ethanol, acetate, and succinate, also low amounts of formiate could be produced [26]. Succinate and acetate are competitive by-products which could eventually decrease the ethanol yield as was shown in Figure 6.2. Thus, glycerol can be converted into ethanol and either hydrogen or formiate. The resulting mixture can be easily purified due to the significant physicochemical differences among its compounds.
The analysis here performed is based on the results presented by Yazdani and Gonzalez, about glycerol conversion to ethanol by Escherichia coli SY04 (pZSKLMgldA) [27]. Their experimental study used two approaches: (i) ethanol and H 2 co-production, and (ii) ethanol and formiate co-production. It was found that the maximum theoretical yield in both cases was 1 mol of ethanol plus 1 mol of either formiate or hydrogen per each mol of consumed glycerol. As additional information to perform the simulation, the average molecular formula for E. coli of CH 1.9 O 0.5 N 0.2 was used [28]. Three different possibilities for ethanol production from glycerol were considered. The first and second possibilities use crude glycerol (88 wt %) in a fermentation stage at a dilution
105
of 10 g/L and 20 g/L, respectively. Meanwhile, the third possibility considered pure glycerol (98 wt %) at 10 g/L. The flowsheet of these three simulated bioprocess for fuel ethanol production from glycerol using E. coli is shown in Figure 8.12.
In all cases the flowsheet is the same, but the operational conditions are different. Obtained glycerol from the purification process (88 wt % or 98 wt %) was cooled at 37 C and diluted (10 g/L or 20 g/L) in fresh water at 37 C. Then the glycerol fermentation process was carried out by E. coli SY04 (pZSKLMgldA) [27] and a mixture of ethanol, formiate, and cells was produced. Cells were withdrawn by centrifugation and an aqueous stream of ethanol and formiate was obtained. This stream was distilled and ethanol concentrated in two distillation columns with 40 and 30 stages respectively. Then, an ethanol stream between 93 wt % and 94 wt % of purity was obtained (concentration near to ethanol-water azeotrope 95.6 wt %). Finally, ethanol was dehydrated in a molecular sieve and fuel ethanol was obtained at 99.5 wt %. The main results of this simulation are shown in Table 8.7.
Water
Broth
5
Distillate 1
6
Distillate 2
Ethanol
Glycerol Diluted Glycerol
Solids
Water waste 1
Adsorbate Water waste 2
Figure 8.12. Simplified flowsheet of fuel ethanol production from glycerol at 88 wt % and 98 wt %. 1. Mixed tank, 2. Fermentation tank, 3. Centrifuge, 4. First distillation column, 5. Second distillation column, 6. Molecular sieves.
In general terms the ethanol production has been described as a process composed of four main stages named: raw material conditioning, fermentation, separation and dehydration, and waste treatment. A comparison among the ethanol production from traditional feedstocks (i.e., sugar cane and corn) versus raw glycerol as raw material was performed. Figure 8.13 shows the required stages for ethanol production from these three feedstocks.
106
Table 8.7. Simulation results for fuel ethanol production from glycerol STREAM Diluted glycerol From Crude Glycerol at 10 g/L Temperature (C) Mass Flow (kg/hr) Mass Fraction: Water Glycerol E. coli Ethanol Formiate 0.99 0.01 0 0 0 0.9899 0.0002 0.0004 0.0048 0.0046 0.06 0 0 0.94 0 0.005 0 0 0.995 0 37 37 77.9 292 77.9 273.072 Broth Distillate 2 Ethanol
57633.551 57624.871
From Crude Glycerol at 20 g/L Temperature (C) Mass Flow (kg/hr) Mass Fraction: Water Glycerol E. coli Ethanol Formiate 0.98 0.02 0 0 0 0.9801 0.0000 0.0006 0.0103 0.0089 0.067 0 0 0.933 0 0.005 0 0 0.995 0 37 37 77.9 313 77.9 290.508
28896.25 28881.1230
From Pure Glycerol at 10 g/L Temperature (C) Mass Flow (kg/hr) Mass Fraction: Water Glycerol E. coli Ethanol Formiate 0.99 0.01 0 0 0 0.9896 0.0000 0.0004 0.0053 0.0047 0.07 0 0 0.93 0 0.995 0 0.005 0 37 37 77.9 317 77.9 293.383
57530.195 57538.901
107
Combustion gases Bagasse Co-generation Steam
Sugar cane
Corn
Crude Glycerol
Grinding
Grinding
Evaporation
Clarification
Cooking
Neutralization
CO2
Fermentation
Enzyme
Liquefaction
Washing
Centrifugation Dehydration by adsorption
Enzyme
Saccharification
Distillation
Ethanol
DIstillation train Whole Stillage
CO2 Dehydration by adsorption
Fermentation
Fermentation
Evaporation train
DIstillation train
Centrifugation
Biomass
Ethanol Concentrated stillage
Evaporation train Whole Stillage
DIstillation train
Centrifugation Thin Stillage
Dehydration by adsorption
Drying
Ethanol
Distillers dried grains
Figure 8.13. Stages for ethanol production from sugar cane, corn, and crude glycerol.
Although simulations were carried out for ethanol production from sugar cane and corn according to the flowsheets shown in Figure 8.14, the most relevant results were obtained from the economic assessments and then they are discussed in the Section 8.8.2.
On the other hand, based on the possibility of transforming both raw glycerol and biomass to ethanol, a sustainable production of biodiesel from oil palm was here proposed. Thus, sustainable biodiesel production can be performed using only oil palm as a single feedstock. Palm oil extraction produces mainly two lignocellulosic residues: empty fruit bunches (EFB), produced in the highest amount; and palm press fiber (PPF) resulting from press cake separation. They have an important lignocellulosic content and low moisture, thus both residues can be used as feedstock for bioethanol production [24]. Besides, glycerol is the main by-product of biodiesel production and it can also be transformed to ethanol. Thus, both oil extraction residues (EFB and PPF) and raw glycerol are used as feedstock to produce the ethanol required to carry out the transesterification reaction with palm oil.
108
Figure 8.14. Simplified flowsheet for ethanol production from: (A) sugar Cane: 1. Washing tank. 2. Mill. 3. Clarifier. 4. Rotary Filter. 5. Fermentator. 6. Centrifuge. 7. Absorption column. 8. Concentration column. 9. Rectifying column. 10. Molecular sieves. 11. Evaporator. 12. Boiler. 13. Turbo-generator. (B) Corn: 1. Washing tank. 2. Mill. 3. Liquefaction reactor. 4. SSF Reactor. 5. Absorption column. 6. Concentration column. 7. Rectification column. 8. Molecular sieves. 9. First evaporation train. 10. Centrifuge. 11. Second evaporation train. 12. Air dryer.
109
The first step is the overall extraction process of palm oil from fresh fruit brunches (FFB). The whole extraction process includes: Pretreatment stage, where FFB are first cooked using saturated steam in order to prepare fruit to a subsequent remotion from brunches, and then they are digested in a cylindrical vertical tank at 100C, obtaining a separation of pulps from nuts. Extraction stage, mashed fruits are passed through a screw pressing, where crude oil is splited up from cake. Refining stage, is made, first decanting crude with hot water at 90C, obtaining a decanter cake from additional ,oil is recover. Clarified oil contains 1% of water, for that reason, must be dried under vacuum conditions before be stored in oil tanks. Also, obtained press cake is treated, in order to obtain Palm Press Fiber (PPF) and nuts, these vegetable wastes are used to extract palm kernel oil (PKO) and palm kernel cake (PKC).
Then, EFB and PPF obtained in the oil extraction process is used in bioethanol production from lignocellulosic biomass, composed up to 75% of cellulose and hemicellulose. The overall process usually includes five main steps: biomass pretreatment, cellulose hydrolysis, fermentation of hexoses, separation and effluent treatment. In first step, feedstock is pretreated, because composition of this biomass, containing up to 75% of cellulose and hemicellulose, it should be broken down into fermentable sugars able to be converted into ethanol and other products [23]. Among available pretreatment methods, in this work is used a diluted acid pretreatment with sulphuric acid 1-10% at 121 C [29], in order to hydrolyze hemicellulose, producing Hexose and Pentose. In a previous work Cardona et al [18], showed a very promising integrated configuration for bioethanol production from an energy viewpoint [18, 30] known as simultaneous saccharification and co-fermentation (SSCF). In this configuration the hydrolysis of cellulose, the fermentation of glucose released, and the fermentation of pentoses present in the feed stream is simultaneously accomplished in a same single unit, using a genetically modified Zymomonas mobilis, Culture broth exiting SSCF bioreactor has an ethanol concentration of about 6% weight. This stream is concentrated up to 92% in two distillation columns. The dehydration of ethanol is made by adsorption with molecular sieves. Stillage obtained from the bottoms of concentration column is evaporated to reduce its volume and diminishing the costs of its further treatment and the lignin is separated using centrifugation.
110
Otherwise, biodiesel is produced from refined palm oil obtained in extraction section. This vegetable oil is composed by a mixture of triglycerides, being major compounds: Tripalmitin, Triolein and Trilinolein. They are transeterified, reacting with ethanol using potassium hydroxide as catalyst. This process is carried out with an integration approach named multistage reactor-extractor. This process combines the chemical reaction and liquidliquid extraction, achieving high selectivity, conversion, productivity, and purity [31]. In this way two main streams are obtained: biodiesel-enriched liquid phase (65% of ethyl esters), continuously removed from the reactor-extractor and sent to a separation unit where ethanol is recovered, and glycerol-enriched liquid phase (44% of glycerol) [32]. Finally, an additional amount of ethanol can be produced from raw glycerol as it was above described.
Simulations are based on Colombian palm industry conditions, reported by Gutierrez et al. [24] with an average installed processing capacity of crude palm oil of 122 tonnes per day of FFB. And the lignocellulosic residues are 28.06 tonnes per day of EFB, and 17.98 tonnes per day of PPF. Thus, the total crude palm oil is 21.76 tonnes per day. To carry out the simulation of integrated biodiesel production process some particularities of each stage must be considered, which are described as follows: Extraction process considers the composition data reported by Abdul Aziz et al. [33, 34], and Wan Zahari and Alimon [35], for both lignocellulosic residues, EFB and PPF, obtained during palm oil extraction, also the yield extraction was based on the reported data by Prasertsan and Prasertsan [36], for processing of FFB to crude palm oil.
Ethanol production process from lignocellulosic biomass was analyzed in a previous work [30], and a brief description of the main processing units is given: pretreatment of lignocellulosic biomass, enzymatic hydrolysis, and co-fermentation processes were simulated based on a stoichiometric approach. Thus, lignocellulosic biomass was converted into glucose and pentoses, which after were converted into cell biomass, ethanol, and fermentation by-products.
Biodiesel production process from palm oil was previously analyzed by Gutirrez et al. [24] in an integrated raw-ethanol production process from lignocellulosic biomass, where a kinetic approach and a multi-stage reactorextractor were used. This kinetic model considers a serial reactive system which transforms triglycerides and ethanol into an ethyl
111
ester (biodiesel) molecule and either diglycerides, monoglycerides, or glycerol. Figure 8.15 shows jointly the ethanol production process from lignocellulosic biomass, the biodiesel production process from palm oil, and the ethanol production process from glycerol.
Here, the process integration is made at different levels to increase efficiency and productivity. First level is given for two individual processes: i) Integration reactionreaction for the process of producing ethanol from biomass feedstock EFB and PPF. In this process the reaction of hydrolysis of cellulose is carried out simultaneously with the fermentation of pentoses and hexoses in the SSCF process. ii) Integration reactionseparation of the process of producing biodiesel from palm oil, which uses a multistage extractor reactor. Second level of integration takes place between the processes of biodiesel and ethanol production using lignocellulosic biomass and glycerol. Second level uses a totally integrated configuration, using oil palm residues EFB and PPF, which are proposed as raw materials for ethanol production. They enter the pretreatment reactor where react with dilute acid at high pressure. Then, the pretreated lignocellulosic biomass undergoes the transformations described in Bioethanol Production from lignocellulosic biomass Section obtaining dehydrated ethanol with purity greater than 99.5% by weight. This stream of ethanol, is mixed with an incoming one form (Bioethanol from glycerol) section, where process crude glycerol is first purified, finally, the crude glycerol is first refined, and then converted to ethanol by mean of an Escherichia coli strain in a fermentative process [37]. Final ethanol mixture with a purity of 99.5 % weight, along with crude oil, is fed to a multi-stage reactorextractor (Biodiesel from palm oil where transesterification reaction is continuously accomplished by reactive extraction process using KOH. Also, main input data and operation conditions used in the simulation process are shown in Table 8.8.
Thus, sustainable biodiesel production from oil palm was simulated considering jointly four processes named: palm oil extraction and refining, biodiesel production, and ethanol production from two feedstocks, lignocellulosic residues and raw glycerol. The corresponding simulation results for the main process streams are shown in the Table 8.9. Due to extraction process is not showed in the Figure 8.15 the main feed streams are EFB, PPF, and palm oil; and the other feed streams are service fluids and catalytic
112
agents. In this sense, the main product streams are biodiesel and ethanol, and also some waste water streams are obtained.
Figure 8.15. Flowsheet for the integrated process of combined biodiesel and bioethanol production. Syrup (concentrated sugars in water). (1) Pretreatment reactor, (2) Washing, (3) Ionic exchange, (4) Simultaneous saccharification and co-fermentation, (5) Concentration column, (6) Rectification column, (7) Molecular sieves, (8) Evaporation train, (10) Centrifuge, (11) Multi-stage reactorextractor, (12) Distillation column for biodiesel purification, (13) Distillation column for glycerol purification, (14) Neutralization tank, (15) Centrifuge, (16) First distillation column, (17) Washing tank, (18) Evaporation column, (19) Second distillation column, (20) Mixed tank, (21) Fermentation tank, (22) Centrifuge, (23) Third distillation column, (24) Fourth distillation column, (25) Molecular sieves.
113
Table 8.8. Main input data and operation conditions used in the simulation process. Feature Description FFB: Cellulose 18,38; Hemicellulose 12,52; Lignin 9,05; Others 9,05; Raw materials Moisture 56,84. composition EFB: Cellulose 15,47; Hemicellulose 11,73; Lignin 7,14; Ash 0,67; Moisture 65,00. PPF: Cellulose 24,00; Hemicellulose 14,40; Lignin 12,60; Ash 3,00; Oil 3,48; Others 2,52; Moisture 40,00. Ethanol production from Lignocellulosic biomass pretreatment: H 2 SO 4 diluted at 190C and 12.2 atm by 10 min. Hemicellulosic conversion 75%. Simultaneous saccharification and co-fermentation: T. reesei
lignocellulosi cellulases and recombinant Z. mobillis at 30C by 144 h. Cellulose and c biomass cellobiose conversion are 80% and 100% respectively. Ethanol and biomass yield are 92% and 2,7% of theoretical. Ethanol distillation: Two distillation columns at 1,77 atm, where the final ethanol concentration is 92,3 wt %. Ethanol dehydration: Molecular sieves at 116 C by 10 min at 1,7 atm. Final ethanol concentration: 99,5 wt %. Multi-stage reactorextractor: 5 counter-current stages at 60C with a Biodiesel production residence time of 6 h. Liquid phase equilibrium was considered. Biodiesel purification: Ethanol recovery by distillation from both light
from palm oil (biodiesel) and heavy (glycerol) phases.
Glycerol fermentation: E. coli SY04 (pZSKLMgldA) is used in a Ethanol production from crude glycerol fermentation broth of 20 g/L of glycerol (previously purified until 88 wt %) at 37 C and pH 7. Ethanol distillation: Two distillation columns at atmospheric pressure, where the final ethanol concentration is 94,1 wt %. Ethanol dehydration: Molecular sieves at 116 C by 10 min at 1,7 atm. Final ethanol concentration: 99,5 wt %.
114
Table 8.9. Main process streams for ethanol production from lignocellulosic biomass Recycle Lignocell Stream Biomass d Broth water for washing T (C) P (bar) Mass flow (Kg/h) Cellulose (%) Hemicellulose (%) Lignin (%) Glucose (%) Xylose (%) Water (%) Triolein (%) Diolein (%) Monoolein (%) Ethanol (%) Ethyloleate (%) Glycerol (%) 20 1 1913,1 18,38 12,52 9,05 --56,84 ------30 1 4185 1,52 1,31 3,72 0,56 0,67 80,47 ---5,59 --77,4 1,793 667,3 0,04 0,03 0,09 0,5 0,89 97,78 ---0,01 0,01 0,02 Rectific. Column Distillate 93,4 1,793 310,8 -----8,5 ---91,5 --BioEtOH Biodiese From Biomass 25 1000 230,7 -----0,5 ---99,5 --73,7 0,2 947,8 -----0,002 0,6 0,2 0,02 1,32 98,9 0,02 244,9 0,3 95,8 -----10,5 -1,2 0,298 0,002 -88 0,5 ---99,5 --47,11 l Refined Glycerol BioEtOH From Glycerol 77,9
This configuration considers two simultaneous processes: (i) saccharification and fermentation, where the cellulose hydrolysis produces glucose, which is assimilated by the microorganisms and converted into ethanol, and thus the inhibitory effect of glucose over cellulases are reduced. And (ii) reactive extraction where reaction and separation are integrated in only one processing unit and then conversion, yield and productivity are improved related to conventional processes because of the continuous products extraction. In this way, a high product concentration is obtained in both cases.
A high ethanol/palm oil ratio is used inside the multistage reactor-extractor since the excess of ethanol leads to a better conversion of feedstock during biodiesel production, as it was showed by Gutirrez et al [24], because of the ethanol fed to the reactorextractor is a mixture of the ethanol obtained from both production processes (lignocellulosic biomass and glycerol) and the recycled ethanol during biodiesel
115
purification. This mixture ensures a high ethanol flow inside the multi-stage reactor extractor. Thus, a 99,9 % of triolein conversion is reached with a final ethyloleate purity of 98,4 wt %.
Biodiesel production from oil palm and ethanol production from lignocellulosic biomass have been individually described and analyzed in many times. An important integration approaches was recently performed by Gutirrez et al [24], they considered different process configurations for heat and mass integration, and results were discussed based only on data obtained from process simulation. But the integration of these two processes leaves an unsolved problem, which is the production of low cost glycerol; since its sales do not represent a significant income for the integrated biodiesel production. In this way, the integration of a biorefinery that uses crude glycerol as feedstock to produce more rawethanol was analyzed. Thus, it is possible to have the oil palm as single raw material to produce biodiesel. Glycerol conversion to ethanol was 99,8 % and the yield was 99 % of theoretical.
8.3 PHB production

Glycerol purification, glycerol fermentation (cell growth and PHB accumulation), mass cell pretreatment, PHB isolation, and PHB purification are the five stages needed for the process of PHB production from raw glycerol. The purification process of raw glycerol was described in the Chapter 4. This process includes a methanol recovering which decreases the purification costs in 37.5% [38].C. necator JMP 134 can synthesize PHB up to 70 wt % of the cell dry mass from various carbon substrates [39]. The fermentation process is carried out in two stages, in the first stage cell growth occurs and in the second stage PHB is synthesized. Air and pure oxygen are fed at the first fermentation tank where the fermentation broth is saturated between 15 or 20 DOC %, thus PHB accumulation takes place inside the cell mass [40]. The detailed conditions for the fermentation process are shown in Table 8.10
After fermentation, the next step is PHB isolation and purification. PHB must be extracted from the cell cytoplasm. Cell membrane is broken and PHB is dissolved and separated from the residual biomass. The separation step can be divided in three parts: pretreatment, extraction, and purification. In the pretreatment step cell disruption is
116
carried out easily and some alternatives for this step are: heat, alkaline or salt pretreatment and freezing [41].
Table 8.10. Process conditions for glycerol fermentation. Feature Description Glycerol 88 wt% Feedstock Feed concentration: 170.8 g/L Feed flow rate: 1000 kg/h Pumped Outlet Pressure: 25 atm Net Work required: 8.1 KJ/Kg Sterilization Temperature: 139 C Heat duty: 443.96 KJ/Kg Heat exchange Heat duty: 415.96 KJ/Kg Required area: 63.62 m2 Temperature: 35 C pH: 7 Cell mass growth Residence time: 21 h. Aeration: 0.6 vol/(vol*min) Cell mass concentration: 4.91 wt % or 50.4g/L 4.24 wt % or 44.4 g/L PHB concentration: 0.44 wt % or 4.5 g/L Temperature: 35 C pH: 7 Residence time: 22.5 h. PHB accumulation Cell mass concentration: 7.1 wt % or 73.4 g/L PHB concentration: 2.7 wt % or 27.8 g/L Flow rate: 929.05 kg/h 8.7 wt % or 91.5 g/L 5.5 wt % or 57.1 g/L 893.173 kg/h 0.70 wt % or 7.3 g/L 429.31 KJ/Kg 402.29 KJ/Kg 51.34 m2 Glycerol 98 wt% 249 g/L 1000 kg/h 25 atm 7.9 KJ/Kg
Some of the different extraction methods to separate PHBs from the cell residual material are: solvent extraction, digestion, mechanical cell disruption, supercritical fluids extraction, cell fragility and spontaneous liberation. In Table 8.11 advantages and disadvantages of the most commonly used PHB extraction methods are listed. Solvent extraction modifies the cell membrane permeability and the PHB is then dissolved [41].
117
Some used solvents are: chlorinated hydrocarbon (e.g. chloroform), cyclic carbonates (e.g., propylene and ethylene carbonates), halogenated solvents (e.g., chloroethanes and chloropropanes), non-halogenated solvents (e.g., chain (410 carbons) alcohols, esters, amides, and ketones (both cyclic and acyclic compounds)). Digestion can be chemically or enzymatically performed. Chemical digestion uses different chemical agents to destroy lipids, carbohydrates, proteins and enzymes. According to the chemical agent used, the chemical digestion could be: digestion by surfactants (e.g. anionic sodium dodecyl sulfate (SDS) and synthetic palmitoyl carnitine), by sodium hypochlorite, by sodium hypochlorite and chloroform, surfactant-hypochlorite digestion, surfactant-chelate digestion, and selective dissolution of non-PHA cell mass by protons.
The enzymatic digestion uses enzymes to degrade the cell membrane. Some varieties of proteolytic enzymes have high activities on protein dissolutions and slight effects on PHB degradation. Enzymatic digestion can be complemented by other extraction methods. Mechanical cell disruption has been widely used to recover intracellular proteins by different ways [42-43] such as: bead mill disruption, high pressure homogenization, disruption by ultrasonication, centrifugation, and chemical treatment. Supercritical fluids have unique physicochemical properties such as high densities and low viscosities that make them suitable as extraction solvents. Due to its low toxicity and reactivity, moderate critical temperature and pressure (31C and 73 atm), availability, low cost, and nonflammability CO 2 is the most used fluid [44]. This extraction method can also be combined with NaOH or salt (NaCl) pretreatments to get higher disruption levels [41]. The cell fragility method takes advantage of the cell fragility shown by some bacteria after large amounts of PHB accumulation. Other extraction methods use air such as: air classification and dissolved-air flotation. Finally, purification methods involve a hydrogen peroxide treatment combined with action of enzymes or chelating agents [41].
Besides, based on the available methods for PHB extraction (see Table 8.11) three PHB production processes from either raw glycerol (88 wt %) or pure glycerol (98 wt %) were designed.
118
Table 8.11. PHB Extraction Methods (Adapted from Jacquel et al [25]) Extraction Methods Advantages Elimination of Endotoxine/high purity No polymer degradation Treatment of high cell densities No polymer degradation High purity Low polymer degradation high purity Limited degradation/low operating cost High purity/low environmental pollution High recovery and high purity low operating costs Good recovery No chemicals used No chemicals used Low cost, low toxicity Use of weak Disadvantages Break PHA granules morphology. Hazards connected with halogenated solvents. High price/Low recovery Low purity/Water waste treatment needed Degradation of the polymer Degradation of the polymer High quantity of solvent needed Results (wt %)
Solvent extraction
Purity: 99.5%; Recovery> 90%
Digestion by surfactants
Digestion by NaOCl Digestion by NaOCl and chloroform Digestion by NaOCl and surfactants Digestion by chelate and surfactants Selective dissolution of NPCM (Non PHB cell mas) by protons Enzymatic digestion Bead mill disruption High pressure homogenization Supercritical CO2 Using cell fragility
Surfactant: High cell density digestion by SDS)Purity >95%; release rate >90% Purity: 99%; Recovery: 94% Purity: >97%; Recovery: 91% Surfactant-EDTA disodium salt. Purity: 98%; Recovery: 86.6% Purity: 98.7%; Recovery: 93.3%
Large volume of wastewater Low degradation of the polymer
Purity: 98.7 wt% Recovery: 95.4% Purity: 92.6 wt% Recovery: 90% Yield: 98% Purity: 95% Recovery: 89% Purity: 99%
High cost of enzymes Require several passes Poor disruption rate for low biomass levels Low micronization Low recovery -
119
extracting conditions Air classification Dissolved air flotation Spontaneous liberation High purity No chemicals used No extracting chemicals needed Low recovery Require several consecutive flotation steps Low recovery (80% cells secretes PHB granules spontaneously
Recovery: 96% Yield: 90% Purity: 97% Purity: 86%
Yield: 80%
The flowsheets for PHB production are shown in Figure 8.16. The fermentation process begins with a sterilization of diluted glycerol. Pure glycerol was diluted at 249 g/L and raw glycerol at 170.8 g/L based on the total glycerol consumption by C. necator [40]. Glycerol conversion takes place in two continuous fermentation stages for cell growth and PHB accumulation, with operation times of 21 and 22.5 h, respectively. Total glycerol consumption is considered on the second fermentation stage. The sterilization and fermentation stages are common for the three PHB production processes. Table 8.12 shows the first downstream process (see Figure 8.16) and it is based on a variation of the BIOPOL flowsheet [45-46]. The first step is a thermal treatment at 85 C, and then a digestion process using the pancreatin enzyme Burkholdeira sp. PTU9 and NaOCl is carried out [47]. This pretreatment causes an appropriate cell disruption releasing the PHB to the fermentation broth. The digestion product containing between 7 to 9 wt % of biomass is filtrated and the residual cell mass is withdrawn. The mixture containing the resuspended PHB at 5.5 5.7 wt % is treated with a hydrogen peroxide solution. Then, using a flash process the majority of the water content is retired. Finally, PHB at 99.9 wt % is obtained by spray drying.
Figure 8.16 shows the second downstream process and the process conditions are given in Table 8.13. After passing through the high pressure homogeniser, the depressurized stream is centrifuged and the solid product is heated and mixed with diethyl succinate (DES) in a 1/20 ratio of biomass/solvent. The solvent extraction process takes place by modification of the cell membrane permeability and PHB dissolution [41]. Residual cell mass is withdrawn by centrifugation and a mixture of PHB-water is gelled by cooling and the DES is recovered. Finally PHB at 99.9 wt % is obtained by spray drying.
120
Figure 8.16 also shows the third downstream process for PHB production, which is described in Table 8.14. This downstream process uses an alkaline pretreatment with a NaOH solution. Then, a digestion process is carried out using NaOCl and sodium dodecyl sulfate (SDS) as detergent. The disrupted cells are centrifuged and PHB is washed with H 2 O 2 . The obtained mixture is subjected to an evaporation process and most of the water content is discarded. Finally, PHB at 99.9 wt % is obtained by spray drying.
Fermentation stage
1
GLYCEROL (88 wt % or 98 wt %)
4 3
GASES
O2
AIR ENZYME
Downstream Process I
12
WASTE WATER 2
WASTE WATER 1
11
10
H2O2
NaOCl
PHB
SOLIDS STEAM
Downstream Process II
14
WASTE WATER 2
11 13 12
10 STEAM
8
DES
7
|
9
STEAM DES SOLIDS 2 STEAM H2O2 WASTE WATER 1
PHB
Downstream Process III
12
WASTE WATER 2
11
10
SDS NaOCl
NaOH
PHB STEAM
SOLIDS
Figure 8.16. Flowsheets for PHB production from glycerol (88 or 98 wt %). Fermentation stage: 1. Pump; 2. Sterilizer; 3. Heat exchanger I; 4. Fermenter I; 5. Fermenter II. Downstream Process I: 6. Heater; 7. Digestor; 8. Centrifuge; 9. Washer tank; 10. Heat exchanger II; 11. Evaporator; 12. Spray drier. Downstream Process II: 6. Homogenizer; 7. Centrifuge I; 8. Heat exchanger II; 9. Heat exchanger III; 10. Extractor; 11. Centrifuge II; 12. Heat exchanger IV; 13. Decanter; 14. Spray drier. Downstream Process III: 6. Alkaline tank; 7. Digester; 8. Centrifuge; 9. Washer tank; 10. Heat exchanger II; 11. Evaporator; 12. Spray drier.
Fermentation broth
121
Table 8.12. Process conditions for PHB recovery: Downstream Process I Feature Feedstock Heat Pretreatment Description Fermentation broth from glycerol at : 88 wt % Temperature: 85 C Residence time: 15 min Heat duty: 109.95 KJ/Kg Temperature: 50 C NaOCl at 30 wt % Ratio NaOCl/cell mass: 1/2 Enzyme: Burkholdeira sp. PTU9 Residence time: 1 h. pH: 9 Enzyme concentration: 2 wt % Residence time: 20 min. Retired products: mass cell, mainly. Concentration: 1.2 v/v % Heat duty: 2.305 MJ/Kg Required area: 3.06 m2 PHB purity: 37.7 wt % Heat duty: 2.1542 MJ/Kg Product Purity: 99.9 wt % Flow rate: 24.98 kg/h 98 wt %
114 KJ/Kg
Chemical + enzymatic Digestion
Centrifugation H 2 O 2 -Water washing Heat Exchanging Water evaporation Spray Drying
2.227 MJ/Kg 2.95 m2 53.0 wt % 1.02 MJ/Kg 99.9 wt % 48.25 kg/h
Table 8.13. Process conditions for PHB recovery: Downstream Process II Feature Feedstock Pumped Description Fermentation broth from glycerol at : 88 wt % Pressure outlet: 70 Mpa Net work required: 239.73 KJ/Kg Pressure: 70 Mpa High pressure Temperature: 110 C Residence time: 45 min homogenizer Heat Duty: 218.40 KJ/Kg Residence time: 20 min. Centrifugation 1 Recovered products: solids at 62.5 wt % Heat duty: 0.831 MJ/Kg Heat Exchanging 2 Required area: 0.40 m2 Solvent to heat: Diethyl-succinate Heat duty: 0.158 MJ/Kg Heat Exchanging 3 Required area: 0.5 m2 98 wt % 249.68 KJ/Kg
250.02 KJ/Kg 65 wt % 0.829 MJ/Kg 0.45 m2 0.158 MJ/Kg 0.43 m2
122
Solvent extraction Centrifugation 1 Heat Exchanging 4 Decantation
Spray Drying
Temperature: 110 C Pressure : 1 atm Mass ratio of PHB/solvent: 1/20 Residence time: 20 min. Extracted products: cell mass Heat duty: 0.200 MJ/Kg Required area: 19.5 m2 Temperature: 25 C PHB purity: 38 wt % Heat duty: 1.13 MJ/Kg Product Purity: 99.9 wt % Flow rate: 25.36 kg/h
0.216 MJ/Kg 18 m2 41.7 wt % 1.09 MJ/Kg 99.9 wt % 48.74 kg/h
Table 8.14. Process conditions for PHB recovery: Process III Feature Feedstock Description Fermentation broth from glycerol at : 88 wt % Temperature: 35 C Concentration: 3 M Alkaline pretreatment Ratio: 0,4 (Kg of NaOH)/(Kg of mass cell) Temperature: 55 C NaOCl at 30 wt % Chemical + surfactant Ratio NaOCl/cell mass: 1/3 Surfactant: Anionic sodium dodecyl sulfate (SDS) Digestion Heat Duty: 109.20 KJ/Kg Residence time: 20 min. Residence time: 20 min. Centrifugation 1 Retired products: mass cell H 2 O 2 -Water washing Concentration: 1.2 v/v % Heat duty: 2.34 MJ/Kg Heat Exchanging 4 Required area: 6.6 m2 Water evaporation PHB purity: 25.2 wt % Heat duty: 2.53 MJ/Kg Product Purity: 99.9 wt % Spray Drying Feed flow rate: 25.13 kg/h 98 wt %
125.01 KJ/Kg
2.12 MJ/Kg 6.3 m2 37.2 wt % 2.14 MJ/Kg 99.9 wt % 48.46 kg/h
123
PHB production from crude glycerol starts with the glycerol purification process. Glycerol content in the feedstock is 60.05 wt % and the rest is mainly methanol, which is recovered at 99.9 wt % of purity using an evaporation process. The process continues with impurities treatment and water evaporation and then the stream containing 80.5 wt % of glycerol is distilled. Two different operation conditions were used for the molar distillated ratio: 0.11 and 0.40, to obtain glycerol at 88 and 98 wt %, respectively.
The glycerol fermentation process can be carried out by two ways, using glycerol at 88 or 98 wt %. Each way requires different glycerol concentrations in the fermentation media, which are 170.8 and 249 g/L, for 88 and 98 wt % of glycerol respectively. These differences account for the impurities of glycerol at 88 wt % which affect the metabolic process of C. necator. The diluted glycerol stream is sterilized at 139 C and 25 atm, in both cases. Then, temperature and pressure are fitted to operation conditions (i.e., 35 C and 1 atm). The first fermentation stage is called Cell mass growth, where air and oxygen are fed to reach the stress conditions. This process is carried out for 21 h at pH 7; then if glycerol at 88 wt % is used, 50.4 g/L of cell mass and 4.5 g/L of PHB are obtained. When glycerol at 98 wt % is used, 44.4 g/L of cell mass and 7.3 g/L of PHB are obtained. In the second fermentation stage the operation conditions are kept equal to the first fermentation stage, where the residence time is 22.5 h and PHB accumulation occurs. Thus, 27.8 g/L (for glycerol at 98 wt %) and 57.1 g/L (for glycerol at 98 wt %) of PHB are obtained in the fermentation outlet stream. The fermentation broth is mainly a mixture of incorporated PHB in the cell mass and water. To recover the PHB from this broth, three different downstream processes were considered, and each one was evaluated with glycerol at 88 and 98 wt % (see Figure 8.17).
In the Downstream Process I a heat pretreatment is carried out, which denatures the genetic material and proteins, and destabilizes the outer membrane of the bacterial cells. The cell mass is then disrupted and PHB is released using for 1 hour a combined digestion involving Burkholdeira sp. PTU9 enzyme (2 wt %) and a sodium hypochlorite solution (30 wt %) in a 1/2 mass ratio of NaOCl/cell mass. Furthermore, solids are removed by centrifugation, followed by water washing and a peroxide hydrogen treatment. The resulting mixture is heated and near 90 % of water is evaporated. When the fermentation process is carried out with glycerol at 88 or 98 wt %, a stream with 37.7 or 53.0 wt % of PHB is obtained respectively. Finally, this steam is spray dried until 99.9
124
wt % of PHB is obtained. Energetically wise the most efficient process is the one that uses glycerol at 98 wt % since the other one requires evaporating higher water quantities in the spray drying process (see Table 8.12). The total energy consumptions were 4.57 MJ/Kg and 3.36 MJ/Kg when glycerol at 88 or 98 wt % was used respectively.
Figure 8.17. Scheme for the simulation procedure to synthesize PHB from crude glycerol.
Raw Glycerol: 60 wt %
Glycerol Purification Process
Crude Glycerol: 88 t%
Pure Glycerol: 98 t%
Fermentation Process in Two Stages
Fermentation Process in Two Stages
Downstream
Downstream
Downstream
Downstream
Downstream
Downstream
Process III
Process III
Process II
Process II
Process I
PHB 99.9 wt %
Process I
125
The pretreatment step in the Downstream Process II is carried out in a high pressure homogenizer at 70 Mpa, and 110 C for 45 min and then the water excess is extracted by centrifugation requiring 218.40 KJ/Kg processed. The products stream and the stream containing the solvent Diethyl-succinate (DES) are mixed at 110 C, with a mass ratio PHB/solvent of 1/20. Solvent extraction takes place and the disrupted cell mass is retrieved by centrifugation. The resulting mixture is decanted at 25 C and the recovered DES is recicled in the extraction process. When fermentation is performed with glycerol at 88 or 98 wt %, a 38.0 or 41.7 wt % of PHB purity is achieved respectively. Finally, this PHB stream is spray dried up to 99.9 wt % of purity. Using glycerol at 98 wt % implies a higher energy consumption (2.79 MJ/Kg) than that for glycerol at 88 wt % (2.77 MJ/Kg). These energy consumptions were calculated as a sum of the main energy consumer units from Table 8.13.
The Downstream Process III differs from the Downstream Process I in the pretreatment and digestion steps. Pretreatment is carried out in alkaline media at 35 C using a solution of NaOH (3 M) in a NaOH/cell mass ratio of 0.4. Then, the combined digestion process takes place at 55 C for 20 min. This process involves anionic sodium dodecyl sulfate as surfactant and sodium hypochlorite (30 wt %) with a mass ratio NaOCl/cell mass of 1/3. After water evaporation, PHB at 25.2 or 37.2 wt % of purity are obtained when the fermentation process is carried out with glycerol at 88 or 98 wt %, respectively. Then spray drying is carried out and PHB is purified up to 99.9 wt %. Nevertheless, glycerol at 98 wt % takes lower energy consumption than glycerol at 88 wt % (i.e., 4.38 MJ/Kg and 4.98 MJ/Kg, respectively).
8.4 D-Lactic acid production

In order to analyze the production process of D-lactic acid from raw glycerol, three main stages have been distinguished: (i) glycerol purification, (ii) glycerol fermentation, and (iii) D-lactic acid recovery and purification.
Although lactic acid bacteria have been used for D-lactic acid production from carbohydrate rich feedstocks, it has also been reported the use of alternative biocatalysts which are mainly engineered Escherichia coli strains able to produce D- or L-lactic acid [48-52]. But only a few papers have been published on the use of glycerol as carbon
126
source for D-lactic acid production [53-54]. For instance, Hong et al. [54] compared eight bacterial strains for lactic acid production from glycerol. Thus, the strain named AC-521 and a member of E. coli, showed the best performance for a fed-batch fermentation process. On the other hand, Mazumdar et al. [53] engineered several E. coli strains by overexpressing pathways involved in the conversion of glycerol to lactic acid and blocking those leading to the synthesis of by-products as it was above described. In all cases they used a minimal medium supplemented with sodium selenite, Na 2 HPO 4 , (NH 4 ) 2 SO 4 , NH 4 Cl, and 20 (or 40 or 60) g/l of pure (or crude) glycerol. Conventional purification of lactic acid from the fermentation broth could be performed by two routes: (i) crystallizing and acidifying the previously clarified and concentrated (32 wt %) fermented liquor; or (ii) crystallizing, filtering, dissolving, and subsequently acidifying the previously precipitated calcium lactate. But, these conventional routes generate huge quantities of calcium sulphate cake which is difficult to dispose of [55]. The main impurities contained in the clarified fermentation broth are residual substrate, color, and other organic acids. Thus, in order to recover and purify the lactic acid, and also to remove these impurities from the fermentation broth, several processes have been proposed. The most remarkable purification processes are: adsorption, electrodialysis, reactive extraction, and reactive distillation. Here, a brief description of each one is given and some of their advantages and disadvantages are also discussed.
Recovery of lactic acid from a fermentation broth by adsorption requires special characteristics of extractants and solid sorbents such as: high adsorption capacity and selectivity, regenerability, and in some cases biocompatibility with microorganisms. Many carboxylic acid fermentations operate effectively at a pH > pKa of the acid product; for example lactic acid (pKa = 3.86) fermentation is typically produced at pH 56 [55]. In this case, agents sufficiently basic to retain a significant capacity several pH units above the acids pKa are recommended. Different basic extractants and polymeric sorbents have been investigated for the extraction and sorption of lactic acid, but the uptaking degree depends mainly upon the agent basicity and capacity [56]. For instance, weak base polymer adsorbents such as: IRA-35 [57], MWA- 1, and VI-15 [58], which did not show high final purity of lactic acid from fermentation broth. Better results were reported for other studies [59] which used adsorbents with a water-insoluble macro-reticular gel, or a
127
weak basic anionic exchange resin with a tertiary amine, or a pyridine functional group, or a strongly basic anionic exchange resin with quaternary functional groups.
Ion exchange technique has also been studied for the recovery of lactic acid from a fermentation broth by several ionic exchanger resins such as: poly(4-vinylpyridine) resin (PVP) [60], IRA-420 [61], IRA-400 [62-63], and IRA-92 [64]. Using IRA-92 (weakly basic exchanger) and under optimal conditions of the fermentation broth (pH 6.0), lactic acid was recovered with a yield, purity, and specific productivity of 0.826, 96.2%, and 1.16 g L.A./(g-resin day), respectively [64]. Otherwise, when IRA-400 (a strong anionic exchange resin) was used by different authors in an fluidized bed column, 0.18 g L.A./(g-resin) and 0.126 g L.A./(g-resin) [65] were recovered. Similar results have also been reported for lactic acid recovering by ionic exchange resins, such as: 0.1 [66], 0.18 [67], and 0.2 [68] g L.A./(g-resin).
By mean of a simulated moving bed (SMB) chromatography process (a continuous separation process consisting of a circle of chromatographic columns) with a PVP resin, high product purity (99.9 wt %) and high purification yield (>93%) were achieved [69]. On the other hand, the PVPs adsorption capacity was found to decrease about 14 % each time after base regeneration [70].
An advantage for adsorption on ion exchange resin is its possibility to couple it with the fermentation process. But in the same way, adsorption and ion exchange technologies require: (i) regeneration of the ion exchange resin, (ii) adjustment the pH of the fed stream in order to increase the sorption efficiency, (iii) large amounts of extra-chemicals, and (iv) treatment and disposal of large quantities of salts and effluents [55]. Besides, the decreasing on the adsorption capacity for the ion exchange resins has also been reported [55]. Additionally, because of the low adsorption selectivity of some ion exchange resins, a further purification by esterification is necessary.
Electrodialysis has been recognized as a promissory technology for lactic acid recovery from the fermentation broth since the product can be continuously removed maintaining constant the pH of the medium [71]. Recovery of lactic acid is performed from lactate salts in a two steps process, a conventional electrodialysis to concentrate and purify the product, and a bipolar electrodialysis to convert the of lactate salts into lactic acids [55]. In
128
situ lactate recovery electrodialysis has been used with free and immobilized cells in order to reduce the product inhibition. But, although the final amount of product was increased in the medium containing immobilized cells, problems related to deposition and fouling of bacteria on the membranes were found [55]. On the other hand, complete technological schemes have also been suggested around electrodialysis process for lactic acid purification. Bailly et al. [72-73] proposed a process in which a conventional electrodialysis is utilized prior to an electrodialysis stage which uses bipolar membranes to increase the concentration of organic acid salts. The same configuration of tow-stage electrodialysis was reported by Habova et al. [74] as an efficient technique for lactate ions recovering. The same configuration of two-stage electrodialysis was reported by Habova et al [74] as an efficient technique for lactate ions recovering, while Li et al. [75] combined both conventional electrodialysis and bipolar membrane electrodialysis in one laboratory scale bioreactor. These processes let to have a good pH control, lead to reduce the generation of troublesome salts, and seem to be economically and environmentally attractive. But, exploitation of electrodialysis with bipolar membranes will require two previous stages, named: micro-filtration and monopolar electrodialysis. Despite the several studies performed in order to improve the eletrodialysis fermentation method, commercialization of this process has not been reported [55].
Other widely studied alternative to recover lactic acid from a fermentation broth is the reactive extraction. This process uses the reaction between extractants which are in the organic phase and the extracted materials which are in the aqueous phase. Then, the complexes formed during the reactions are solubilized in the organic phase. The most used extractans for the reactive extraction of carboxylic acids are hydrocarbon, phosphorous, and aliphatic amine [76]. Thus, three categories of extractions are recognized: (i) extraction by solvation with carbon-bonded oxygen-bearing extractants, (ii) salvation with phosphorous-bonded oxygen-bearing extractants, and (iii) proton transfer or ion-pairing formation with high molecular weight aliphatic amines and their salts [77]. But, although only the last two extractants have been mainly used in the recovery of carboxylic acids, the best extractabilities have been noticed for aliphatic amines. This attribute is due to the behavior of the acid proton during the transfer from an aqueous phase to an organic solution. That is, meanwhile the measures of extractability in systems containing oxygen-bearing extractans are given by both the acid strength in the aqueous solution and the hydrogen bond in the organic solution; in the case of using aliphatic
129
amines, the extracted compounds are more stable ammonium salts. The reactive extraction process of carboxylic acids with tertiary amine extractants is composed of three sequential steps: dissociation of carboxylic acid, proton transfer to the amine, and recombination of ammonium salt [77]. The following reaction describes the overall process, but its stoichiometry varies with several factors, such as: the property and concentration of amine, acid, and diluent. R 3 N + HA- R 3 NHA
(8.15)
A successful reactive extractive process depends mainly on a high distribution coefficient for the lactic acid (K d ), and also the extractant should have both low water solubility and low distribution coefficient for the impurities. The distribution coefficient is defined, as ratio of the lactic acid concentration in the solvent phase to lactic acid concentration in the aqueous phase, as shown in equation (8.16).
Kd =
Concentration of LA in the in the organic phase Concentration of LA in the in the aqueous phase
(8.16)
Other two important parameters used to evaluate the performance of the reactive extraction process are the extraction efficiency (E) and the loading (Z), as shown in equations (8.17) and (8.18), respectively.
E (%) =
(Initial concentration of LA - Reffinate concentration of LA) 100 Initial concentration of LA
(8.17)
Z=
Concentration of LA in the organic phase Inital concentration of amine
(8.18)
Although long chain aliphatic amines are effective extractants of carboxylic acid from dilute aqueous solutions, these extractants must always be used dissolved in organic diluents due to their physical properties such as viscosity, density, and corrosivity [78]. Solvents containing functional groups which interact strongly with complex are called as Active Diluent (e.g., 1-octanol), while the solvents with low interaction level with complex
130
are called Inert Diluent (e.g. n-hexane). Thus, nature of the organic diluents affects not only the basicity of amines but also the behavior of the extraction process [78]. Extractants, mainly ternary amines, including tri-n-octylamine (TOA), tripropylamine (TPA), tributyl amine (TBA), trilauryl amine, tri-n-butyl phosphate (TBP), triisooctylamine, Alamine 336 have been reported for lactic acid recovery from aqueous medium; besides inert and active diluents such as: hexane, heptane, xylene, chloroform, chlorobenzene, chlorobutane, octanol, decanol, dodecanol, oleyl alcohol, tributyl phosphate,
methylisobutylketone, and methylene chloride-n-hexane, have also been used for this process [55].
Recovery of lactic acid by reactive extraction is performed through the formation of an acid-amine complex according to equation (8.15). A second step of regeneration is required in order to reverse this reaction and to recover both the acid product phase and the extractant phase available to be recycled. Regeneration could be carried out through backextraction into an aqueous phase [78] by two approaches, swing either temperature of diluent composition which leads to changes in the equilibrium relationship. Moreover, recovering of lactic acid from a loaded solvent phase can also be performed using solutions of NaOH and HCl [79-80]. One of the most suitable techniques for the regeneration process is the so-called temperature-swing regeneration [78], where the extracted stream is mixed with a fresh aqueous stream at a higher temperature to produce an acid-laden aqueous product and an acid-free organic phase.
Lactic acid can also be obtained by a sequential process containing esterification of crude lactic acid, distillation of ester, and hydrolysis of ester by reactive distillation in order to obtain the preceding alcohol and lactic acid. Among the different alcohols analyzed for the esterification process, methanol appears to be the most suitable one because of the relatively low boiling of both methanol and methyl lactate. This implies lower energy consumption for heating is the subsequent processes (i.e., distillation of ester and reactive distillation). A typical configuration analyzed for lactic acid purification through reactive batch distillation includes two columns reactants separation from the product and two reboilers for esterification reaction and hydrolysis reaction [81-83]. In most of the studies performed on lactic acid purification by reactive distillation, the cation exchange resin Dowex 50W has been used for both esterification reaction and hydrolysis reaction [81-84]. The main drawbacks for implementing this technology at industrial scale are not
131
only the low conversion levels on the esterification reaction (around 50%) but also the high energy requirements to evaporate a mixture of produced ester and the water present in the fermentation broth.
Joglekar et al. [55] drew four possible routes for lactic acid purification from a fermentation broth considering both different fermentation modes and downstream processes as shown in Table 8.15. Also, the purification costs were estimated, for an assumed production of 1000 Tons of lactic acid (100 wt %) per year, based on reported costs and on prices of raw materials and utilities for India. According to the authors, purification cost of Route 2 was not calculated since the data available on expanded bed ion exchange adsorption technology is not enough for estimating the costs.
Table 8.15. Downstream processes for lactic acid recovery from a fermentation broth Route Fermentation mode 1 Continuous Reactive extraction, re-extraction, esterification, and hydrolysis by reactive distillation 2 Continuous Adsorption/desorption using methanol as eluent, xxx 1.59 Downstream process Cost (USD$/LA kg)
esterification, and hydrolysis by reactive distillation Addition of lime, precipitation of calcium lactate, 3 Batch dissolution in methanol, acidification to separate calcium sulphate, esterification, and hydrolysis by Reactive distillation. 4 Batch Addition of ammonium hydroxide, micro filtration, monopolar electrodialysis, bipolar electrodialysis,
1.40
esterification with reactive distillation and hydrolysis.
1.74
Here, based on the above reviewed literature, a technological scheme for lactic acid production is proposed, simulated, and economically assessed as follows. Based on the results reported by Mazumdar et al. [53], five fermentative scenarios were identified to be analyzed for the D-lactic acid production from raw glycerol. The scenarios differ on: strain, substrate concentration, substrate purity, fermentation time, and fermentation stages. In this way, different values of yield, glycerol consumption, and productivity are reported. Detailed information for the fermentation stage of each scenario is given in Table 8.16.
132
Table 8.16. Base information for the glycerol fermentation to D-lactic acid Glycerol Scenario Strain Conc. (g/L) Glycerol purity for fermentation Glycerol consumption (%) Molar yield to D-LA Main by product: Ethanol. 1 LA01(pZSKLMgldA) 20 Pure 100 0.820 Fermentation time: 36 h Main by product: Succinic acid. 2 3 4 LA02(pZSglpKglpD) LA02dld(pZSglpKglpD) LA02dld(pZSglpKglpD) 20 40 40 Pure Pure Crude 100 100 100 0.812 0.833 0.859 Fermentation time: 36 h Fermentation time: 72 h Fermentation time: 72 h Two fermentation stages: 48 h and 36 h, 5 LA02dld(pZSglpKglpD) 60 Crude 90 0.934 each one. Details
The downstream process for D-lactic acid recovery and purification from the fermentation broth is based on a reactive-extractive process because of its good process characteristics, such as: low toxicity, low cost, low boiling point, extraction yield, and recovery yield. Tri-n-octylamine and dichloromethane are used as extractant and active diluent, respectively. A mixture of tri-n-octylamine diluted in dichloromethane at 0.6 M was considered for this process and the used weight ratio of fermentation broth to organic media was 2/1. During the reactive extraction process, a complex molecule is formed according to equation (8.15) (see Figure 8.18). The back-extraction process is carried out by a combined effect of changing the extractant concentration (Regeneration Swing Concentration Process) and changing the temperature profile (Regeneration Swing Temperature Process), which is reached by a distillation process under vacuum conditions in order to obtain a highly pure D-lactic acid. Also, distillated water is fed in a molar ratio of 1/1 respect to the formed complex.
The simplified flowsheet for D-lactic acid production is shown in Figure 8.19. The main differences among the five scenarios are determined by the fermentation stage, which leads to different: flows, equipment sizes, and operational conditions. The fermentation product is a mixture containing mainly organic acids and cell mass, where the cell mass is
133
withdrawn by centrifugation. The clarified broth is mixed with three organic streams, where two of those streams correspond to recycled tri-n-octylamine and dichloromethane. The third one is a mixture of fresh both extractant and diluent because of their lost during the purification process. Then, the reactive extraction process is performed at 20 C with a residence time of 1.5 h and with a mixture of 0.6 M of tri-n-octylamine dissolved in dichloromethane. Thus, the complex is produced and extracted to the organic phase, which is distillated in order to remove the remaining water. The distillation product is the heterogeneous azeotrope of dichloromethane-water. This azeotrope is also obtained from the distillation of the aqueous phase obtained during the reactive extractive process, which contains a significant amount of dichloromethane. Then, these two streams are mixed and treated by decantation in order to recover dichloromethane at 99.7 wt %. The bottom stream obtained after distillation of the organic phase from the reactive distillation process is also distillated and the remaining dichloromethane is recovered. This stream is mixed with the decantation product containing dichloromethane at 99.7 wt %, and the final obtained purity was 99.8 wt %. The new bottom stream, containing mainly the complex, is mixed with distillated water and then the back-extraction process takes place by distillation. The distilled product is tri-n-octylamine at 99.1 wt % while the bottom product is a mixture of D-lactic acid (85.5 wt %) and water. This last stream is finally purified by vacuum distillation and D-lactic acid at 99.9 wt % is obtained.
The fermentation processes were simulated using a yielding approach where glycerol is consumed and products including cell mass are formed according to Table 8.17. The molecular formula used for E. coli strains was CH 1.9 O 0.5 N 0.2 [85-86].
Figure 8.18. Complex formed during the reactive extraction process of D-lactic acid
134
Metanol E-1 Raw Glycerol Solids Organic Phase Aqueous glycerol R-1 C-1 Water Dec-1 E-2
Waste Water 1
DC-1 Waste Water 2
Water RII-1 Glycerol (98 wt %) M-1
Glycerol (85 wt %) Adsorbate Diluted Glycerol RE-1 Fresh TOA DCE Cen-1
Aqueous Phase DC-5 Distillated-5 Organic Phase
Dec-2
Distillated-2 DC-2
DC-3
Fermentation Broth
F-1
Distillated-3 Bottoms-2 M-3 Lactic Acid DC-4 DC-6 Aqueous Phase M-2 Solids
Distillated-4
Distillated-6 Bottoms-3 Bottoms-6
Bottoms-4
Figure 8.19. Simplified flowsheet for D-lactic acid production from raw glycerol. E: evaporator, R: reactor, C: centrifuge, Dec: Decanter, DC: distillation column, M: Mixer, F: fermentator, RE: Reactor extractor.
Table 8.17. Stoichiometry for glycerol fermentation to D-Lactic Acid by Engineered E. coli. Scenario Glycerol Residual Ac Ac Succ Ac EtOH D-Lac Ac For Ac CO 2 Glycerol 1 2 3 4 5 -1 -1 -1 -1 -1 0 0 0 0 0.1 0.021 0.044 0.048 0.045 0 0.008 0.0076 0.008 0.024 0 0 0 | 0.82 0.812 0.833 0.859 0.8406 0 0 0 0.006 0 0.045 0.044 0.048 0.045 0.0414 11.5 6.5 3.4 4.5 5 Biomass
0.0414 0.00855
D-lactic acid production process starts with the purification of raw glycerol up to the required purity according to each scenario as it was showed in Table 8.16. Detailed information about the glycerol purification process was previously reported [87-88]. A summary of the main simulation results for each scenario is given in Table 8.18. The final production of D-lactic acid is directly related not only to the fermentation yield but also to
135
the substrate consumption. In other words, while the decreasing order respect to the yield was: Scenario 5 > Scenario 4 > Scenario 3 > Scenario 1 > Scenario 2, the decreasing order respect to the D-Lactic acid production was: Scenario 4 > Scenario 5 > Scenario 3 > Scenario 1 > Scenario 2. This change in the order between the Scenarios 5 and 4 occurs due to the incomplete consumption of glycerol during the fermentation in the Scenario 5.
Table 8.18. Summary of the main simulation results for D-lactic acid production process Dilut gly Lactac1 Mass Flow (kg/hr) Water Glycerol Formic Acid Ethanol Acetic Acid Succ Acid Lact Acid Ecoli Tri-n-amine Dichlmethan Complex 28575.1 27995.5 578.9 0 0 0 0 0 0 0 0 0 28560.0 27995.5 0 0 6.95 7.9 0 464.3 72.2 0 0 0 Scenario 1 Lacacid4 Lacacid6 13959.4 68.2 0 0 0.855 0.612 0 0.273 0 986.1 10813.9 2085.2 Scenario 2 Dilut gly Lactac1 Mass Flow (kg/hr) Water Glycerol Formic Acid Ethanol Acetic Acid Succ Acid Lact Acid Ecoli Tri-n-amine Dichlmethan 28575.1 27995.5 578.9 0 0 0 0 0 0 0 0 28531.5 27995.5 0 0 16.6 5.9 459.8 40.8 0 0 0 Lacacid4 13954.7 68.0 0 0 1.28 0.08 0.27 0 1002.4 10813.5 2064.8 Scenario 3 Dilut gly Lactac1 Lacacid4 Lacacid6 Destwater Lactac9 Lactacp Lacacid6 3068.9 0 0 0 1.28 0.08 0.27 0 1002.4 0 2064.8 Destwater 84.6 84.633 0 0 0 0 0 0 0 0 0 Lactac9 484.1 67.97 0 0 0.87 0.079 415.1 0 0 0 0 Lactacp 416.3 0.172 0 0 0.87 0.079 415.1 0 0 0 0 3073.5 0 0 0 0.001 0.612 0 0.273 0 986.0 1.394 2085.2 Destwater 83.96 83.96 0 0 0 0 0 0 0 0 0 0 Lactac9 487.3 67.5 0 0 0.001 0.417 0 419.3 0 0 0.125 0 Lactacp 419.5 0.165 0 0 0 0.004 0 419.3 0 0 0 0
136
Mass Flow (kg/hr) Water Glycerol Formic Acid Ethanol Acetic Acid Succ Acid Lact Acid Ecoli Tri-n-amine Dichlmethan
14359.1 13779.5 578.9 0 0 0 0 0 0 0 0
14310.3 13779.5 0 0 18.1 5.64 471.7 21.35 0 0 0
14231.1 69.92 0 0 2.702 0.161 0.586 0 959.5 11073.0 2118.2 Scenario 4 Lacacid4 14243.9 70.13 0 0.03 2.535 0.17 0.605 0 901.1 11071.9
3081.2 0 0 0 2.702 0.161 0.586 0 959.5 0 2118.2 Lacacid6 3095.2 0 0 0 2.535 0.17 0.605 0 901.1 0
86.89 86.89 0 0 0 0 0 0 0 0 0 Destwater 89.87 89.87 0 0 0 0 0 0 0 0 0 Destwater 88.01 88.01 0 0 0 0 0 0 0 0 0
498.6 70.18 0 0 1.85 0.157 426.4 0 0 0 0 Lactac9 516.36 73.09 0 0 1.759 0.166 441.3 0 0 0 0 Lactac9 505.9 71.28 0.066 0 2.274 0.267 432.0 0 0 0 0
421.58 0 0 0 0 0.157 421.4 0 0 0 0 Lactacp 441.7 0.188 0 0 0.019 0.166 441.3 0 0 0 0 Lactacp 432.5 0.114 0.066 0 0.012 0.267 432.0 0 0 0 0
Dilut gly Lactac1 Mass Flow (kg/hr) Water Glycerol Formic Acid Ethanol Acetic Acid Succ Acid Lact Acid Ecoli Tri-n-amine Dichlmethan 14402.8 13821.2 580.6 0 0 0 0 0 0 0 0 14375.6 13821.2 0 1.741 17.04 5.956 487.8 28.33 0 0 0
2190.8 2190.8 Scenario 5 Lacacid4 14320.4 70.94 0.066 0 3.3 0.274 0.902 0 938.8 11155.0 2143.9 Lacacid6 3087.3 0 0.066 0 3.3 0.274 0.902 0 938.8 0 2143.9
Dilut gly Lactac1 Mass Flow (kg/hr) Water Glycerol Formic Acid Ethanol Acetic Acid Succ Acid Lact Acid Ecoli Tri-n-amine Dichlmethan 9622.8 9041.2 580.6 0 0 0 0 0 0 0 0 9640.9 9041.2 56.32 0 15.67 6.365 477.4 31.48 0 0 0
137
Otherwise, for the reactive extraction process no high differences were noticed in terms of Distribution coefficient or Loading as is shown in Table 8.19. Global recovery of D-lactic acid was around 90 % respect to its production during the fermentation process. Also, high recovery percentages were achieved for both tri-n-octylamine and dichloromethane, indicating that low requirements of fresh both extractant and diluent are required. Higher differences are noticed when the whole technological scheme is analyzed. The maximum global molar yield from glycerol to D-lactic acid was obtained for the Scenario 4, while the minimum was obtained for the Scenario 2. The relative difference between these two scenarios was 5.93 %, which was close to the relative difference for the fermentation yield obtained for the same both scenarios, 5.47 %. Thus, it can be stated that the technological performance of D-lactic acid production from raw glycerol depends mostly on the global conversion of substrate to the main product during the fermentation stage.
Table 8.19. Data representing the behavior of the downstream process for D-lactic acid production Scenario 1 Reaction-Extraction Process Extraction efficiency (%) Distribution coefficient Loading (Z) Downstream Process Global lactic acid recovery (%) Tri-n-octylamine recovery (%) Dichloromethane recovery (%) Global Process Global process yield from 71.39 70.68 71.75 75.14 73.55 90.32 99.998 99.98 90.3 99.992 99.57 89.36 99.998 99.99 90.48 99.998 99.99 90.51 99.998 99.93 92 11.59 0.63 92 11.58 0.62 92 11.68 0.64 92 11.68 0.66 92 11.78 0.65 Scenario 2 Scenario 3 Scenario 4 Scenario 5
Glycerol to Lactic acid (%)
138
8.5 Succinic acid production

In general terms, the downstream process for succinic acid purification and recovery from the fermentation broth starts with a cell separation by centrifugation or microfiltration, and in some cases an additional ultrafiltration process is used in order to separate residual cell mass, proteins, and other fermentation supernatants. For succinic acid recovery and purification different alternatives have been proposed, such as: precipitation with ammonia or calcium hydroxide, electrodialysis, reactive extraction, and sorption/ion exchange. Here, the advantages and disadvantages of these alternatives are briefly discussed.
Industrially the most used method for recovery of carboxylic acids from a fermentation broth is the precipitation process with calcium hydroxide or calcium oxide, especially for the cases of lactic and citric acids. Addition of calcium hydroxide or calcium oxide to the clarified fermentation broth leads to calcium salt formation of succinic acid, which are filtered off and treated with concentrated sulfuric acid. This last addition generates calcium sulfate (CaSO 4 ), gypsum, in an equimolar amount. Then, free succinic acid is purified by active carbon or ion exchange, and finally the product is further concentrated and crystallized by evaporation. From a commercial view of point, this purification way cannot be used because high amount of calcium sulfate are by-produced as a waste a adequate disposal is required [89-90]. Additionally, the precipitation process requires high consumption of calcium hydroxide, calcium oxide, and sulfuric acid which cannot be regenerated or recycled causing high process costs. That is why precipitation with calcium hydroxide or calcium oxide has been reported as unlikely process for large-scale production of bio-succinic acid [91]. Precipitation with ammonia has also been reported for succinic acid recovery at laboratory scale [92-93]. During this process the produced diammonium succinate must be later treated with sulfate ions, or ammonium bisulfate, or sulfuric acid at a low pH in order produce both succinic acid precipitate and ammonium sulfate. Finally succinic acid is obtained after the dissolution in methanol and recrystallisation processes while ammonium sulfate can be cracked thermally in order to produce ammonia and ammonium bisulfate. Although precipitation with ammonia reduces the amount of waste production, low selectivities have been reported for this purification alternative [93].
139
Other alternative for succinic acid purification is the electrodialysis process. Membranes are charged either with positive or negative groups and selectively allow to cations or anions passing through the membranes, thus succinate anions are able to passage through positively charged membranes while sodium cations are repelled. Glassner et al. [94] reported a total purification yield of 60% for a desalting electrodialysis combined with a water-splitting electrodialysis. This process requires both a set of chelating ionexchange columns to replace the divalent cations of the succinate salt with sodium ions and a bipolar membrane water-splitting dialysis to obtain succinic acid from succinate. Then, after electrodialysis and ion exchange process, evaporation of water and crystallization of the succinic acid are required [94-96]. Electrodialysis is known as an expensive alternative not only by the high energy consumption but also by the materials cost. Besides to the low yield, some other problems such us: low selectivity [97], handle on binary ions [95], and fouling [98] have also been reported.
On the other hand, since liquid-liquid extraction has shown low distribution coefficients for carboxylic acids recovery from the fermentation broth [99-100], reactive extraction appears as a better option to increase yield and selectivity to organic acids from an aqueous phase [101]. Mixtures of amines (reactive components) dissolved in non-water miscible organic solvents have been widely studied for carboxylic acids recovery [102103]. Amine reacts with the succinic acid thought out a proton transfer or ion pair formation mechanism depending on the type of amine and the organic solvent [104]. Long-chain aliphatic primary, secondary, and tertiary amines have been proposed for succinic acid extraction [105-111], but in these cases only the undissociated acid can be extracted. Otherwise, although quaternary amines can extract the dissociated and the undissociated succinic acid, its regeneration is difficult by back extraction. Thus, ternary amines have been the most used for carboxylic acids extraction from an aqueous solution [112] using solvents such as: octanol, xylene, heptane, kerosene, methylenchloride or nitrobenzene [105-108, 111, 113]. Mixtures of amines such us: tripropylamine/ trioctylamine or trialkylamines have also been reported [108, 114]. On the hand, different operational alternatives for the reactive extraction process of succinic acid with amines have been proposed. For instance, Huh et al. [115] studied the removal of by-products and impurities present in the fermentation broth such as: acetic acid, pyruvic acid, and salts, by mean a pre-treatment step of reactive extraction with trioctylamine. Then the aqueous succinic acid is purified up to 99.8 wt % by an evaporative crystallizer with a
140
purification yield of 71.3 %. Finally, Kurzrock and Weuster-Botz [101] stated that if the recycling of the costly amines is done efficiently, it is very likely that optimized reactive extraction processes may be applied in the near future for industrial production of biosuccinic acid.
In general terms, the final step of the succinic acid purification process is an ionic exchanger unit in which the residual cations and anions are removed. Different kinds of exchange resins have been investigated in order to produce succinic acid at high purity from either the fermentation broth or succinate. Some examples are the alkaline anionexchange resin (NERCB 04) [116] and the H-type strongly acidic cation-exchange resins [117]. On the other hand, mesoporouses silicas (SBA-15) functionalized with primary, secondary and tertiary amino-functional silanes were reported to be able for for the isolation of pyruvic and succinic acid from fermentation broth [118]. Ion-exchange must be only considered as an additional purification step for succinic acid recovery because of its low both selectivity and yield [119].
In this case, three different types of strains are here analyzed for the succinic acid production from raw glycerol. The first one is the Escherichia coli recently reported by Blankschien et al. [120], the second strain is the Mannheimia sp. Pasteurellaceae reported by Scholten and Dagele [121], and the last one is the Anaerobiospirillum succiniciproducens reported by Lee et al. [122]. In the case of Mannheimia sp. Pasteurellaceae three scenarios are studied while for Escherichia coli and
Anaerobiospirillum succiniciproducens one scenario is analyzed in each case, as shown in Table 8.20. Although it can be observed that three different qualities of glycerol are considered for the fermentation stage (i.e., 76, 90, and 98 wt %), a unique raw glycerol is considered as feedstock. A typical composition of a raw glycerol stream obtained from a biodiesel production process is: methanol 32.59 wt %, glycerol 60.05 wt %, NaOCH 3 2.62 wt %, fats 1.94 wt %, and ash 2.8 wt % [123]. This stream must be purified up to the specified concentration established in Table 1 for the fermentation stage. The purification process previously studied in the Chapter 4. As result of the purification process, a glycerol stream at 80 wt % is obtained. Thus, the glycerol concentration required for the Scenario 4 is obtained. Finally, in order to purify the glycerol stream up to 90 or 98 wt %, a distillation process is carried out.
141
Table 8.20. Base information for the glycerol fermentation to succinic acid Scenario 1 2 3 4 5 Strain Escherichia coli Mannheimia sp. Pasteurellaceae Mannheimia sp. Pasteurellaceae Mannheimia sp. Pasteurellaceae Anaerobiospirillum succiniciproducens Glycerol Glycerol Glycerol purity for consumption Conc. (%) fermentation (g/L) 20 98 wt % 96 9.6 98 wt % 55.2 8.3 9.1 6.5 90 wt % 80 wt % 98 wt % 75.9 71.4 58.5 Molar yield to S.A. 0.522 0.287 0.542 0.453 0.344
Then, the fermentation process can be performed, but due to these scenarios differ on: strain, substrate concentration, substrate purity, and fermentation time, different composition profiles are obtained in the fermentation broth according to the fermentation stoichiometry reported for each case and shown in Table 8.21. The fermentation broth is then clarified by a centrifugation process where the cell mass is withdrawn.
Table 8.21. Stoichiometry for glycerol fermentation to succinic acid by Engineered E. coli. Scenario 1 2 3 4 5 Purity of Gly (wt %) 99 99 90 76 99 Glycerol 92.09 -1 -1 -1 -1 -1 Res Gly 92.09 0.040 0.448 0.241 0.286 0.415 Succinic acid 118.09 0.544 0.520 0.714 0.634 0.588 Acetic acid 60.05 0.064 0.048 0.055 0.051 0.020 Lactic acid 90.08 0.000 0.000 0.000 0.011 x Formic acid Biomass 46.03 x 0.066 0.083 0.044 0.072 0.044 0.066 0.044 x 0.032
Recovery and purification of succinic acid from the clarified fermentation broth is based on the downstram process recently reported by Huh et al. [124]. During the so-called complex process, the by-produced acids are removed selectively from the fermentation broth by reactive extraction, followed by a vacuum distillation step where the volatile impurities are removed. Then a crystallization process is performed and succinic acid is concentrated in the order of five to six folds. Finally, another crystallization process is carrying out at pH 2.0 and 4 C in order to obtain succinic acid highly purified.
142
The pretreatment step of the reactive extraction process selectively removes contaminated organic acids from the dilute fermentation broth using tri-n-octylamine (TOA) and 1-octanol. Since TOA only extracts the undissociated form of carboxylic acids [110, 124], the selective removal of specific acids from the fermentation broth is made possible by using different degrees of dissociation of each acid with the pH. Although the distribution coefficient values decreased with the increase of pH for succinic acid, pyruvic acid, and acetic acid, the distribution coefficient of succinic acid at pH values between 4.0 and 5.0 is close to 0. Thus, the increase of the dissociated acid concentration leads to the reduction of succinic acid extraction. In this, case a multi-stage reactive extractive process was analyzed, but no significant improvements were noticed after each stage. Besides, the author recommended a one stage reactive extraction process in order to do economic the removal of succinic acid from the fermentation broth. Then, by mean a vacuum distillation process volatile impurities such as acetic acid and formic acid are effectively removed, and thus the pretreated fermentation broth is concentrated five- to six-fold by a crystallization process at low temperature (4 C) and in an acid media (pH 2.0). Table 8.22 shows the calculated removal efficiency of carboxylic acids from the fermentation broth based on the data reported by Huh et al. [124]. Thus, applying this complex process all by-produced carboxylic acids are effectively withdrawn with a succinic acid lost of 27%.
Table 8.22. Removal efficiency (%) of the carboxylic acids from the fermentation broth. Fermentation product Reactive extraction Succinic acid Maleic acid Acetic acid Pyruvic acid Fumaric acid 0.45 87.50 44.44 27.14 87.50 0.00 0.00 90.00 11.76 0.00 26.58 100.00 100.00 99.78 100.00 Vaccum distillation Cristallization
On the other hand, due to the fermentation products here considered are: succinic acid, acetic acid, and formic acid according to Table 8.21, it is required to generate the complex molecules as the corresponding reaction products of TOA with of carboxylic acids. Complex I (see Figure 8.20.a) is the reaction product of succinic acid with TOA, while the Complex II (see Figure 8.20.b) and Complex III (see Figure 8.20.c) are the reaction products of formic acid and acetic acid respectively.
143
The simplified flowsheet for succinic acid production is shown in Figure 8.21. As it was observed for the D-lactic acid production processes, the main differences among the five scenarios for the succinic acid production processes are determined by the fermentation stage. Thus different flows, equipment sizes, and operational conditions are obtained for each scenario. Glycerol is first purified up to the required quality according to Table 8.20 and then the fermentation occurs under the stoichiomentry presented in the Table 8.21. The fermentation product contains a mixture of cell mass and organic acids such as succinic acid, acetic acid, lactic acid, and formic acid. The biomass produced during the fermentation stage is withdrawn by centrifugation and the obtained clarified fermentation broth is mixed with both the fresh and the recycled streams containing TOA and octanol. Then, the resulting mixture is subjected to the reactive extraction process.
During the reactive extractive process the complex molecules are produced from each carboxylic acid according to the Figure 8.20, and then most of the succinic acid complex is contained in the aqueous phase while the rest of the impurities are extracted to the organic phase. This organic phase is subjected to a back extraction process by mean a distillation column in which some water is withdraw and most important the used TOA is recovered. This stream is again distilled and most of the carboxylic acids are discarded while a mixture of octanol and TOA containing low quantities of succinic acid and acetic acid is recycled in order to be used in the reactive extractive process. On the other hand, the aqueous phase containing most of the succinic acid is first distilled not only water but also acetic acid and and a few amount of octanol are discarded by distillated. The bottoms product is then subjected to a crystallization process and succinic acid at high purity is obtained.
The main simulation results for each scenario are shown in Table 8.23 and significant differences for the final production of succinic acid can be observed among the five analyzed scenarios. For instance based on Table 8.21, the final production of succinic acid depends mainly on the succinic acid yield (increasing order: Scenario 2 < Scenario 1 < Scenario 5 < Scenario 4 < Scenario 3) and also on the glycerol consumption (increasing order: Scenario 2 < Scenario 5 < Scenario 4 < Scenario 3 < Scenario 1). This higher dependence on the succinic acid yield can be explained because of the increasing order of succinic acid production (Scenario 2 < Scenario 1 < Scenario 5 < Scenario 3 < Scenario 4) was most similar to the succinic acid yield than the glycerol consumption.
144
Figure 8.20. Reaction complexes of succinic acid, formic acid and acetic acid with TOA.
Metanol E-1 Raw Glycerol Solids DC-4 Distillated-4 Waste Water 3 Cry-1 Distillated-3 Organic Phase DC-3 Aqueous glycerol R-1 C-1 Water Dec-1 E-2
Waste Water 1
DC-1 Waste Water 2
Glycerol (85 wt %) Adsorbate Diluted Glycerol Fresh TOA Octanol Cen-1
Distillated-2 DC-2
RE-1
Fermentation Broth
F-1
Succinic acid
Bottoms-4 Bottoms-3 Bottoms-2 M-2 Solids Aqueous Phase
Figure 8.21. Simplified flowsheet for succinic acid production from raw glycerol. E: evaporator, R: reactor, C: centrifuge, Dec: Decanter, DC: distillation column, M: Mixer, F: fermentator, RE: Reactor extractor, Cry: Crystallizator
145
The change between the Scenarios 3 and 4 occurs due to the difference in the glycerol consumption is higher than the succinic acid yield. For the reactive extractive process was noticed that most of the fermentation by-products such as formic acid and acetic acid were selectively removed from the clarified broth with extraction efficiency around 88 % for formic acid and 90 % for acetic acid, while the losses for succinic acid were lower than 1.4 % as shown in Table 8.24. The good performance of the downstream process is stated because of the high recovery levels not only for succinic acid (> 98 %) but also for both tri-n-octylamine and 1-octanol.
Table 8.23. Summary of the main simulation results for succinic acid production process Dilutgly Mass Flow kg/hr Water Glyce-01 Cell mass Succini Formi-01 Aceti-01 Tri-N-01 1-oct-01 Complex Complex3 28575.1 27995.5 578.9 0 0 0 0 X X X X Dilutgly Mass Flow kg/hr Water Glyce-01 Cell mass Succini Formi-01 Aceti-01 Tri-N-01 1-oct-01 Complex Complex2 Complex3 59382.1 58802.5 578.9 X X X X X X X X X Succaci1 28447.3 27995.5 23.16 10.25 403.8 0 24.16 X X X X Succaci1 59490.7 58802.5 259.3 6.83 386.0 24.01 18.12 X X X X X Scenario 1 Complex2 Aqueous2 29668.5 27813.1 27992.7 27659.3 23.15 23.1 X X 403.4 398.2 0 0 2.32 0.097 1.99 1.91 1368.8 2.83 1.61 149.1 X Scenario 2 Complex2 Aqueous2 62220.1 58796.6 259.3 X 385.6 3.00 1.81 3.86 2737.5 1.54 181.5 111.8 58525.4 58134.5 258.6 X 380.9 2.89 1.74 3.71 5.62 0 0 0 Extract2 1855.5 333.4 0.072 X 5.18 0 1.74 0.08 1365.9 1.61 149.1 Extract2 3694.7 662.2 0.759 X 4.7 0.115 0.069 0.147 2731.9 1.54 181.5 111.8 Concent1 3448.4 3298.1 23.1 X 398.2 0 X 1.91 trace X X Concent1 7297.5 6915.4 258.6 X 380.9 0.033 1.31 3.71 0 0 0 0 Succaci2 123.6 0.003 X X 397.6 0 0.811 X X X Succaci2 117.2 0.007 0 X 379.7 0 0 0 0 0 0 0
146
Dilutgly Mass Flow kg/hr Water Glyce-01 Cell mass Succini Formi-01 Aceti-01 Tri-N-01 1-oct-01 Complex Complex2 Complex3 68527.7 67945.3 578.9 X X X X X X X X X Dilutgly Mass Flow kg/hr Water Glyce-01 Cell mass Succini Formi-01 Aceti-01 Tri-N-01 1-oct-01 Complex Complex2 Complex3 61264.8 60675.0 579.3 0 0 0 0 X X X X X Dilutgly Mass Flow kg/hr Water Glyce-01 Cell mass Succini Aceti-01 Tri-N-01 1-oct-01 Complex Complex3 87662.1 87082.5 578.9 0 0 0 X X X X
Succaci1 68659.9 67945.3 139.5 6.83 530.0 20.83 20.76 X X X X X Succaci1 61370.3 60675.0 165.7 6.84 471.0 19.10 29.09 X X X X X Succaci1 87767.5 87082.5 240.2 4.97 436.5 7.55 X X X X
Scenario 3 Complex2 Aqueous2 61970.6 58546.1 58867.9 58248.9 120.9 120.5 X X 458.7 453.5 2.26 2.17 1.80 1.73 26.92 25.96 2555.0 5.59 1.83 0 136.4 0 111.0 0 Scenario 4 Complex2 Aqueous2 64451.9 60269.7 60689.0 59936.9 165.7 165.1 X X 470.5 464.2 2.39 2.29 2.91 2.79 13.42 12.86 3102.52 5.83 1.88 0 144.5 0 179.5 0 Scenario 5 Complex2 Aqueous2 87961.0 87073.8 240.2 X 436.0 0.755 2.43 456.2 1.74 46.58 87355.2 86965.4 240.1 X 435.4 0.752 2.42 7.82 0 0
Extract2 3424.4 619.1 0.33 X 5.23 0.081 0.064 0.961 2549.4 1.83 136.4 111.0 Extract2 4182.2 752.1 0.534 X 6.29 0.100 0.122 0.562 3096.7 1.88 144.5 179.5 Extract2 605.8 108.4 0.077 X 0.603 0.003 0.011 448.4 1.74 46.58
Concent1 7237.9 6948.8 120.5 X 453.5 0.025 1.3 25.96 0 0 0 0 Concent1 7467.8 7143.8 165.1 X 464.2 0.026 2.088 12.86 0 0 0 0 Concent1 10757.5 10375.8 240.1 X 435.4 0.563 2.42 0 0 0
Succaci2 141.3 0.007 0 X 452.9 0 0 0 0 0 0 0 Succaci2 143.9 0.007 0 X 463.5 0 0 0 0 0 0 0 Succaci2 138.6 0.01 0 X 434.8 0 0 0 0 0
147
Finally, the maximum global yield from glycerol to succinic acid was obtained for the Scenario 4, while the minimum was obtained for the Scenario 2 as shown in Table 8.24. This behavior agrees with the order of succinic acid production above described. The relative difference between these two scenarios was 18.24 %, which indicates that the fermentation process has a high influence on the final production of succinic acid.
Table 8.24. Data representing the behavior of the downstream process for succinic acid production Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5 Reaction-Extraction Process Extraction efficiency of SA (%) Extraction efficiency of FA (%) Extraction efficiency of AA (%) Distribution coefficient for SA Distribution coefficient for FA Distribution coefficient for AA Loading for SA (Z) Loading for FA (Z) Loading for AA(Z) 1.28 NA 90.40 0.168 NA 121.648 0.118 NA 0.980 1.22 87.98 90.38 0.168 99.863 128.240 0.054 0.618 0.367 1.14 87.95 90.36 0.170 107.627 138.222 0.064 0.498 0.465 1.34 88.02 90.42 0.168 91.304 117.236 0.063 0.528 0.590 0.14 NA 90.04 0.172 NA 1120.455 0.041 NA 0.915
Downstream Process Global SA recovery (%) Tri-n-octylamine recovery (%) 1-octanol recovery (%) 98.62 97.62 99.66 98.38 98.15 93.30 98.76 88.84 95.98 Global Process Global process yield from Glycerol to SA (%) S.A. succinic acid; F.A.: formic acid; A.A.: acetic acid 0.517 0.493 0.589 0.603 0.565 98.56 95.20 95.99 99.76 91.49 94.59
148
8.6 Propionic acid production

Propionic acid is a naturally occurring carboxylic acid which, in the pure state, is a colorless, corrosive liquid with an unpleasant odor. Propionic acid is used in the manufacture of herbicides, chemical intermediates, artificial fruit flavors, pharmaceuticals, cellulose acetate propionate, and preservatives for food, animal feed, and grain.
As it was performed for D-lactic acid and succinic acid production, the whole technological scheme for propionic acid production was divided in three main stages. These stages are: (i) glycerol purification, (ii) glycerol fermentation, and (iii) propionic acid recovery and purification. The glycerol purification has been widely discussed throughout this document (see Chapter 4). On the other hand, for the glycerol fermentation to propionic acid two strains were identified as the most promissory bacteria available from literature. The first one is a commercial Propionibacterium acidipropionici which consumes pure glycerol [125], and the second one is the engineered Propionibacterium acidipropionici ACK-Tet [126], which is able to consume both pure and crude glycerol as only source of carbon and energy.
Respect to the recovery and purification methods of organic acids several alternatives have been evaluated. Some examples are: liquid extraction [127], reverse osmosis [128], electrodialysis [129], liquid surfactant membrane extraction [130], anion exchange [63], precipitation and adsorption [131], and reactive liquid-liquid extraction [132]. These processes were above described for recovering and purifying of carboxylic acids. Otherwise, Keshav et al [133-144] have studied widely the reactive extraction of propionic acid from a fermentation broth, and different diluent (e.g., benzene, toluene, hexane, nheptane, n-octane, n-dodecane, ethyl acetate, butyl acetate, 1-octanol, 2-octanol, 1decanol, 1-dodecanol, petroleum ether, paraffin liquid, MIBK, oleyl alcohol, sunflower oil, ) and extractant (e.g., Tri-n-butylphosphate, tri-n-octylamine, Aliquat 336, and tri-noctylphosphine oxide)agents have been analyzed. Besides of the final extraction performance, kinetic behavior has also been studied [135]. Based on the results reported by Keshav et al [133-144] for the reactive extraction of propionic acid from the fermentation broth, it was noticed that the best configuration for this process requires the use of tri-n-octyl amine (TOA) as extractant agent while the diluent agent must be ethyl
149
acetate. The concentration that leads to the best performance is 0.686 kmol/m3 and the extraction temperature is 305 K according to Keshav et al [136].
Here and based on the reviewed literature, a technological scheme for propionic acid production is proposed, simulated, and economically assessed. First of all, five different scenarios are considered for the glycerol fermentation stage. Scenario 1 uses a commercial Propionibacterium acidipropionici strain and glycerol diluted at 20 g/L, while in the Scenario 2 the fermentation media contains the same strain and glycerol diluted at 50 g/L. For these two scenarios the fermentation process is carry out at 30 C using pure glycerol [125]. For Scenario 3, Scenario 4, and Scenario 5 the engineered strain Propionibacterium acidipropionici ACK-Tet is considered and pure glycerol at 46 g/L is required in Scenario 3. While crude glycerol at 17 g/L is used in Scenario 4, Scenario 5 considers a completely different configuration for the fermentation process. It is a fibrousbed bioreactor packed with immobilized cells and fed with pure glycerol at 41 g/L. For these three scenarios the fermentation process is carry out at 32 C [126]. In all cases, 100% of glycerol conversion was reached and the fermentation times reported in each case were: 120 h (Scenario 1), 150 h (Scenario 2), 280 h (Scenario 3), 160 h (Scenario 4), and 104 h (Scenario 5).
Since the analyzed scenarios differ mainly on the fermentation stage according to Table 8.25, which shows the stoichiometry of the fermentation process for each scenario, different operational conditions, material and energy requirements, and equipment sizes are required for each scenario. The principles of the reactive extraction process and the corresponding back extraction process for recovering carboxylic acids from an aqueous media were above described.
Table 8.25. Stoichiometry of the fermentation process for each scenario. Scenario Glycerol Propionic Ac. Acetic Ac. Succinic Ac. Biomass MW 92.09 74.08 60.05 118.09 24.73 1 0.9821 0.0925 0.0905 0.3481 1 1 0.7086 0.0785 0.0660 0.2820 2 1 0.6713 0.0368 0.0507 0.3442 3 1 0.8826 0.0537 0.0546 0.1903 4 1 0.7334 0.0414 0.0569 5
150
The simplified flowsheet for propionic acid production is shown in Figure 8.22. First, the raw glycerol is purified up to the required quality in order to perform the fermentation process. Then, fermentation takes place according to Table 8.25, and the fermentation broth is clarified throughout a centrifugation process. This clarified stream is mixed with two organic streams. These streams contain fresh and recycled mixtures of TOA and ethyl acetate and the resulting mixture is TOA in ethyl acetate at 0.686 kmol/m3. Thus, the reactive extraction process is carried out at 32 C. The aqueous phase, containing high quantities of ethyl acetate, is distilled in order to recover it. Otherwise, the organic phase is subjected to a distillation process in which the contained ethyl acetate is recovered and mixed with this coming from the aqueous phase. Then, the resulting organic phase is subjected to the back extraction process, and TOA is recovered by bottoms and recycled to the reactive extraction process. Finally, the resulting stream is distilled under vaccum conditions and propionic acid at high quality is obtained.
A summary of the main simulation results for each scenario is given in Table 8.26. In this case and due to glycerol is completely consumed during the fermentation process in all scenarios, the final production of propionic acid is only related to the fermentation yield. Thus, the decreasing order for propionic acid production is Scenario 1 > Scenario 4 > Scenario 5 > Scenario 2 > Scenario 3 (see Table 8.26). This obtained order agrees with the fermentation yield (see Table 8.25)
Waste Water 1 R-1 C-1 Water Dec-1 E-2
Metanol E-1 Raw Glycerol Solids Organic Phase Aqueous glycerol
DC-1 Waste Water 2
Glycerol (85 wt %) Adsorbate Diluted Glycerol Fresh TOA EtAc Cen-1
DC-3
DC-2
Distillated-1 DC-1
RE-1
Distillated-2 Distillated-3
Fermentation Broth
F-1
Prop Acid Bottoms-3
Bottoms-1
Aqueous Phase
M-2
Solids
151
Table 8.26. Summary of the main simulation results for prpionic acid production process Dilutgly 303.1 Temperature (K) Mass Flow (kg/hr) 27995.5 Water 578.9 Glycerol 0 Cell mass 0 Prop. acid 0 Suc. acid 0 Ac. acid 0 Tri-N 0 Complex 0 Ethyl Dilutgly 303.1 Temperature (K) Mass Flow (kg/hr) 11007.5 Water 578.9 Glycerol 0 Cell mass 0 Prop. acid 0 Suc. acid 0 Ac. acid 0 Tri-N 0 Complex 0 Ethyl Dilutgly 305.1 Temperature (K) Mass Flow (kg/hr) 11987.5 Water 578.9 Glycerol 0 Cell mass 0 Prop. acid 0 Suc. acid 0 Ac. acid 0 Tri-N 0 Complex 0 Ethyl Propaci1 303.1 27995.5 0 54.05 457.3 67.18 457.3 0 0 0 Propaci1 303.1 11007.5 0 43.78 329.96 48.99 29.63 0 0 0 Propaci1 305.1 11987.5 0 53.44 312.6 37.63 13.89 0 0 0 Scenario 1 Complex2 Aqueous2 293.1 283.1 27992.7 27763.1 0 0 0 0 96.00 88.89 67.19 62.23 34.89 32.31 357.9 26.52 2075.7 0 11117.4 2754.03 Scenario 2 Complex2 Aqueous2 293.1 283.1 11006.4 10724.5 0 0 0 0 69.26 55.48 49.01 39.20 29.61 23.72 838.21 53.76 1497.6 0 11117.4 1114.0 Scenario 3 Complex2 Aqueous2 293.1 283.1 11986.3 0 0 65.63 37.67 13.87 903.6 1418.9 11117.4 11708.6 0 0 53.56 30.70 11.35 59.06 0 1203.8 Extract2 283.4 229.6 0 0 7.11 4.96 2.58 331.4 2075.7 8363.3 Extract2 283.3 281.9 0 0 13.78 9.80 5.88 784.4 1497.6 10003.4 Extract2 283.3 277.8 0 12.07 6.97 2.58 844.6 1418.9 9913.6 Organic5 293.1 8.79 0 0 368.4 4.96 2.58 2056.3 0 0 Organic5 293.1 6.34 0 0 274.5 9.80 5.88 2029.0 0 0 Organic5 293.1 6.00 0 0 259.1 6.97 2.58 2023.7 0 0 Propacp 379.8 0.018 0 0 367.4 4.96 0.54 0 0 0 Propacp 380 0 0 0 274.0 9.80 1.20 0 0 0 Propacp 380 0 0 0 258.3 6.97 0.540 0 0 0
152
Dilutgly 305.1 Temperature (K) Mass Flow (kg/hr) 12036.6 Water 578.9 Glycerol 0 Cell mass 0 Prop. acid 0 Suc. acid 0 Ac. acid 0 Tri-N 0 Complex 0 Ethyl Dilutgly 305.1 Temperature (K) Mass Flow (kg/hr) 13492.5 Water 578.9 Glycerol 0 Cell mass 0 Prop. acid 0 Suc. acid 0 Ac. acid 0 Tri-N 0 Complex 0 Ethyl
Propaci1 305.1 12036.6 0 29.55 411.0 40.5 20.27 0 0 0 Propaci1 305.1 13492.5 0 53.44 341.5 42.24 15.63 0 0 0
Scenario 4 Complex2 Aqueous2 293.1 283.1 12035.4 11757.2 0 0 0 0 86.30 70.45 40.50 33.06 20.30 16.57 532.6 34.66 1865.4 0 11117.4 1213.7 Scenario 5 Complex2 Aqueous2 293.1 283.1 13491.2 0 0 71.71 42.27 15.61 794.7 1550.4 11117.4 13218.4 0 0 59.93 35.31 13.03 53.05 0 1351.4
Extract2 283.3 278.2 0 0 15.85 7.44 3.72 498.0 1865.4 9903.7 Extract2 283.4 272.8 0 0 11.78 6.97 2.58 741.66 1550.4 9766.0
Organic5 293.1 7.89 0 0 340.47 7.44 3.72 2047.8 0 0 Organic5 293.1 6.56 0 0 281.6 6.97 2.58 2029.4 0 0
Propacp 379.9 0.018 0 0 339.5 7.44 0.781 0 0 0 Propacp 379.9 0 0 0 280.8 6.97 0.54 0 0 0
Finally, the behavior of the reactive extraction process was analyzed based on its extraction efficiency, the obtained distribution coefficient and the used loading, but no high differences were found for the extraction efficiency while the distribution coefficient varied from 3.8 to 15. Also, it was observed that the global recovery of propionic acid from the fermentation broth is directly related to the extraction efficiency in the reactive extractive process as shown in Table 8.27.
Besides of the high recovery levels for propionic acid, the extractan and diluent agents were almost completely recovered during the downstream process. Thus, low quantities of both agents must be fed as a fresh stream. As it could be expected, the global yield for the downstream process agreed with the order found for the fermentation yield and the
153
propionic acid production (i.e, Scenario 1 > Scenario 4 > Scenario 5 > Scenario 2 > Scenario 3) with a relative difference between Scenario 1 to Scenario 3 of 29.69%.
Table 8.27. Data representing the behavior of the downstream process for propionic acid production Scenario 1 Reaction-Extraction Process Extraction efficiency (%) Distribution coefficient Loading (Z) Downstream Process Global P.A. recovery (%) Tri-n-octylamine recovery (%) EtAc recovery (%) Global Process Global process yield from Glycerol to PA (%) PA: Propionic acid. EtAc: Ethyl acetate. 0.7605 0.5672 0.5347 0.7027 0.5814 80.34 98.73 99.53 83.05 97.42 99.45 82.64 97.17 98.51 82.61 98.33 98.49 82.24 97.45 98.34 80.56 8.761 0.844 83.19 3.790 0.629 82.87 14.156 0.594 82.85 12.965 0.781 82.46 15.032 0.645 Scenario 2 Scenario 3 Scenario 4 Scenario 5
8.7 Economic assessment

8.7.1 1,3-propanediol production
For most industrial processes the cost of raw material represents near to 50 % of the total production costs, while if raw glycerol is used for the production of 1,3-propanediol by mean of engineered K. pneumoniae strains this value is lower between 8.2 to 9.2 % of the total production cost as shown in Table 8.28. The values here obtained do not consider the transportation costs, since the 1,3-propanediol production process was assumed to be a biorefinery adjacent the biodiesel production process. Both utilities and capital costs represent the highest production cost which combined vary between 64.46 % for the Scenario 3 to 66.99 % for the Scenario 1. Also, it was noticed that capital costs increased when final concentration of 1,3-propanediol increased.
154
Table 8.28. Economic results for raw glycerol conversion to 1,3-propanediol: Cost (USD$/kg) and Share (%) Scenario 1 Raw materials Cost Share Utilities Cost Share Operating labor Cost Share Maintenance Cost Share Operating charges Cost Share Plant Overhead Cost Share G and A cost Cost Share Depreciation of capital Cost Share Co-products sales Cost Share Gly. Purify. + ferm. Cost Share 1,3-PD purification Cost Share Total Sale price/production cost Cost 0.0889 8.61 0.2151 20.83 0.0698 6.76 0.0479 4.64 0.0172 1.66 0.0577 5.59 0.0650 6.29 0.4767 46.16 -0.0055 -0.53 0.304 29.40 0.729 70.60 1.033 1.710 Scenario 2 Scenario 3 0.0892 8.19 0.2245 20.61 0.0746 6.85 0.0503 4.62 0.0187 1.71 0.0625 5.73 0.0712 6.54 0.5042 46.28 -0.0057 -0.52 0.331 30.38 0.758 69.62 1.089 1.622 0.0887 9.20 0.1458 15.12 0.0714 7.40 0.0486 5.04 0.0178 1.85 0.0600 6.22 0.0615 6.38 0.4757 49.34 -0.0054 -0.56 0.313 32.42 0.652 67.58 0.964 1.832
In order to obtain more detailed information about the 1,3-propanediol production from raw glycerol, the total production cost was divided in two processing sections: (i) Glycerol purification plus glycerol fermentation and (ii) 1,3-propanediol recovery and purification, as shown in Table 8.28. The first section (i.e., glycerol purification and fermentation) represents in all cases between 29.4 to 32.4 %, while the second section (i.e., 1,3-
155
propanediol purification) is between 67.6 to 70.6 %. The 1,3-propanediol purification cost was estimated to be between 0.652 to 0.758 USD$/kg of 1,3-propanediol. The increasing order for the total production is: Scenario 3 < Scenario 1 < Scenario 2, which is not related to the order found to the 1,3-propanediol production, neither its yield. Also, the commercial sale price for 1,3-propanediol was compared to the production cost and it was found that this ratio (i.e., sale price/production cost) is higher than the unity for all scenarios, and for the Scenario 3 this ratio was 1.832. These results indicate that 1,3propanediol production from raw glycerol by mean of engineered K. pneumoniae could be a profitable alternative for glycerol usage.
8.7.2
Ethanol production
The economic assessment for the purification process considers two scenarios. In the first one, the retired methanol from raw glycerin stream is considered as a waste, and in the second scenario the methanol is recycled to the transesterification process, which contains 99 wt % of methanol and 1 wt % of glycerol. Therefore, in the second case methanol is obtained as a co-product as it was shown in Chapter 4.
Raw glycerol in bioethanol production represented only 30% of the total costs as shown in Table 8.29. Transportation costs were not considered in the economic assessment since the purification step was assumed to be adjacent to the biodiesel production process. On the other hand, utilities and capital costs represent the highest cost on the purification process (i.e., between 20% and 30%). Also, the final quality of glycerol increases mainly the utility costs. Purification costs of raw glycerol to pure glycerol at 88 wt % are 0.1574 US$/L (scenario I) and 0.0984 US$/L (scenario II) when the methanol price is considered. Moreover, when glycerol at 98 wt % is used the Purification costs are 0.1782 US$/L (scenario I) and 0.1124 US$/L (scenario II). Approximate costs for refining raw glycerol have been reported to about 0.15 US$/lb or 0.26 US$/L [37], which are higher than the purification costs obtained here, but near to the obtained purification costs in the scenario I. Also, purification costs obtained are lower than the sale price of each product, which are 0.28 US$/L for glycerol at 88 wt %, 1.39 US$/L for glycerol at 98 wt % from vegetable oil and 1.11 US$/L for glycerol at 98 wt % from tallow. Then a decrease in the whole fuel ethanol production costs from glycerol can be expected due to the purification process.
156
The economic assessment carried out for the glycerol bioconversion process to fuel ethanol does not consider the raw material cost because it is involved in the purification costs. Table 8.29 shows the bioconversion costs (BCCs) obtained using Aspen Icarus. The lowest BCC was obtained for crude glycerol (88 wt %) when it was diluted at 20 g/L, since it uses a lower quantity of water than the other two processes. In this way, equipment size and utilities are modified in each case. On the other hand, when pure glycerol (98 wt %) is used a higher water quantity is necessary then increasing sizing and utilities.
Table 8.29. Bioconversion costs (BCCs) for fuel ethanol production form raw glycerol Crude Item (US$/L) glycerol (10 g/l) Share (%) Crude glycerol (20 g/l) Pure Share glycerol Share (%) (10 g/l) (%)
Utilities Operating labor Maintenance and operating charges Plant overhead and general and administrative costs Capital depreciation Product (US$/L) production cost
0.0599 0.0188
31.82 9.97
0.0503 0.0188
29.41 10.99
0.0975 0.0188
41.28 7.96
0.0205
10.89
0.0154
9.02
0.0193
8.17
0.0266 0.0625
14.14 33.18
0.0309 0.0556
18.05 32.54
0.0410 0.0596
17.37 25.23
0.1883
100.00
0.1710
100.00
0.2361
100.00
Finally, global production costs (GPCs) for raw glycerol bioconversion to fuel ethanol are obtained by adding the purification costs and the BBCs, like shown in Table 8.30. In all cases the purification costs are near 35 % and the BCCs are near 65 %. Furthermore, these obtained GPCs from crude glycerol are lower than those reported by Quintero et al [17] for fuel ethanol production from corn (0.3381 US$/L), where using crude glycerol at 10 g/L and 20 g/L could represent a saving of 15 % and 20 %, respectively. The obtained GPCs are higher than those reported by Quintero et al [17] for fuel ethanol production from sugar cane (0.2153 US$/L). Nevertheless, these obtained GPCs are lower than the international prices for fuel ethanol ranging from (0.4552 US$/L [145] to 0.6057 US$/L [146]). The
157
Although a rigorous analysis of ethanol fuel market and its prices should be carried out, the obtained results indicate that the production process of ethanol fuel from raw glycerol using E. coli can be as profitable as those using sugar cane or corn as feedstocks.
Table 8.30. Global production costs (GPCs) for fuel ethanol production from raw glycerol. Crude Costs glycerol (10 g/l) Purification Costs Bioconversion Costs Global Costs 0.1883 0.2867 65.68 0.1710 63.47 0.2361 67.74 100.00 0.0984 Share (%) 34.32 Crude glycerol (20 g/l) 0.0984 Share (%) 36.53 Pure glycerol Share (10 g/l) 0.1124 (%) 32.26
100.00 0.2694
100.00 0.3485
On the other hand, for the sustainable production of biodiesel from oil palm, total production cost of ethanol from crude glycerol was estimated in 0.2694 US$/L, but this result considers the buying of crude glycerol. This cost is lower than these reported for fuel ethanol production from corn (0.3381 US$/L) and higher that these reported from sugar cane (0.2153 US$/L) as showed Quintero et al [17].
Total production cost of ethanol from lignocellulosic biomass has been calculated by McAloon et al. [147] was 0,396 US$/L, which is lower than the current international price of fuel ethanol, reported by ICISpricing [145] as 0.4552 US$/L. On the other hand, total production cost of biodiesel from oil palm under above described operation conditions was estimated in 0,7971 US$/L (results are not shown), where its sale price ranged between 0,76 0,86 US$/L (ICISpricing, 2010) [145].
Total production cost of biodiesel from oil palm in this integrated production process was estimated in 0,8803 US$/L, which is higher than biodiesel production cost from palm oil. This cost has been discriminated by raw material, services, operative and administrative costs, depreciation and co-products credit, as shows the Table 8.31. Also, column 2 shows the share for each item.
158
Table 8.31. Discriminated costs for integrated biodiesel and raw-ethanol production from oil palm. Item Raw materials Service fluids Labour Maintenance Operating charges General and administrative costs Capital depreciation Co-products credit Total production cost 0,2225 -0,0657 0,8803 23,52 -6,95 100,00 Cost (US$/L) 0,3376 0,2105 0,0083 0,0621 0,0301 0,0750 Share (%) 35,69 22,25 0,87 6,57 3,18 7,92
This technology is promising from an environmental viewpoint because of petrochemical methanol is not required and also in the case of oil palm industry an independence from feedstock can be achieved. On the other hand, regarding crude glycerin, traditional biodiesel industries have had only two options, the first one is its sale as low-quality glycerin with a rather low price; the second is its refining until either, industrial or USP grades, where the involved purification cost have been previously estimated as 0.20 and 0.27 US$/L, respectively Posada and Cardona [37]. These values agree with the purification cost reported by Johnson and Taconi [146], which is 0.26 US$/L. Although, biodiesel production cost by this technology indicates that the process is not economically viable, a further analysis of process design which considers a higher detail level as: energy cogeneration, heat integration, and the use of new technologies, among others, could help to improve the process efficiency which would be reflected directly in the production cost of biodiesel.
Described technology is a perfect representation of the biorefinery concept applied to the biofuels production, in which oil palm is used as a sole feedstock for producing mainly biodiesel, and ethanol as an important co-product which is widely used by the transport sector. Therefore, it is expected that this integrated technology for biodiesel production which uses efficiently the lignocellulosic residues and crude glycerol to produce the
159
required ethanol as feedstock, can be regarded as a valuable opportunity for biofuels production. Moreover, the two proposed ways for ethanol production should eliminate the need of using other feedstock which is important as foods, such as: sugar cane, sugar beet, corn, wheat, sorghum, etc.
8.7.3 PHB production

The total production costs of PHB at 99.9 wt % from raw glycerol are shown in Table 8.32. This table was obtained using glycerol at 88 or 98 wt % in the fermentation stage. Also, the three downstream processes were compared. In all cases the cost for raw material represents the glycerol purification cost. In this way, the glycerol purification process represents only between 4.8 to 5.6 % of the total PHB production cost when glycerol at 88 wt % is used and this values increased between 6.3 to 7.7 % when glycerol at 98 wt % is used.
Table 8.32. Total PHB production costs from crude glycerol through raw glycerol (88 wt %) and pure glycerol (98 wt %) Item Cost (US$/kg) *DSP I *DSP II *DSP III
and Share (%) 88 wt % 98 wt % 88 wt % 98 wt % 88 wt % 98 wt % Raw material Cost Share Utilities Cost Share Operating labor Cost Share Maintenance and operating charges Plant overhead, general Cost Share Cost 0.118 5.6 0.7787 36.94 0.0752 3.57 0.2296 10.89 0.2128 10.09 0.6938 32.91 2.108 100 0.149 7.71 0.658 33.96 0.066 3.41 0.21 10.84 0.186 9.6 0.668 34.48 1.937 100 0.118 4.84 0.9809 40.23 0.0752 3.08 0.2402 9.85 0.2287 9.38 0.7951 32.61 2.438 100 0.149 6.27 0.953 39.97 0.068 2.85 0.236 9.9 0.218 9.14 0.76 31.87 2.384 100 0.118 5.42 0.8678 39.87 0.0752 3.45 0.2451 11.26 0.2091 9.61 0.6616 30.39 2.1767 100 0.149 7.07 0.841 39.77 0.064 3.03 0.223 10.55 0.191 9.03 0.646 30.55 2.114 100
and administrative costs Share Depreciation of capital Cost Share Total Cost Share *DSP: Downstream process
160
In general terms the lower production costs are obtained when glycerol at 98 wt % is used in the fermentation process, due to the higher PHB yield in the fermentation stage and the lower energy requirements in the spray drying process. On the other hand, the lower glycerol purification cost obtained for glycerol at 88 wt % generates a lower share in the total PHB production costs compared with glycerol at 98 wt %. In both cases (88 and 98 wt %) the higher value in the total production cost was obtained for the Downstream Process II, which uses a solvent extraction stage. This extraction requires heating the expensive solvent DES up to 110 C which increase the utility costs.
The total PHB production costs obtained are between 2.11 and 2.44 US$/Kg when glycerol at 88 wt % is used in the fermentation process, and between 1.94 and 2.38 US$/Kg when glycerol at 98 wt % is used. These production costs are near to the lower sale prices reported in the literature (involving profits) and were obtained from other substrates (see Table 8.33). In economical terms the best technological scheme for PHB production from crude glycerol includes three important features as follows: i) purification of crude glycerol up to 98 wt %, ii) a two continuous fermentation stages and finally iii) the PHB recovering performed with the Downstream Process I, which is similar to the BIOPOL flow sheet [45-46].
Table 8.33. Main producers of PHA in the world PHB: Product Company name Biomer: P(3HB) Biotechnoly Co. Germany Biocycle: P(3HB) PHB industrial S/A company, Brazil Biogreen: P(3HB) Mitsubishi GAS Chemical, Japan P(3HB) Metabolix, USA (BASF, ADM) Substrate Small scale production [148] Sugar cane[149] Methanol [148] Price (US$/kg) 25 (2003) 3.75-6.25 (2010) 12.5-15 (2003) 3.12-3.75 (2010) 2.75 (2010) Production (t/y) 50 (2003) 1400(2003) 30-60000(2010) -
Corn, [149]
sugar -
Pure glycerol has a higher cost than raw glycerol, but the yield of PHB is higher when pure glycerol is used as feedstock. Moreover, each downstream process should be adjusted to the conditions of the fermentation broth. Thus, using the blocks process
161
analysis (glycerol purification, fermentation, and PHB isolation and purification) the best technological scheme for the production of PHB was found.
In all the cases assessed in this study the total production costs (see Table 8.32) were lower than those reported in the literature (see Table 8.33) using feedstocks different to glycerol. Generally, it has been suggested that the higher share to the total PHB production cost is the substrate cost which is up to 45%. Meanwhile, if glycerol is used as feedstock this share is below 8 %. These results indicate that using crude glycerol as feedstock to produce PHB could be a profitable alternative to develop biorefineries in the biodiesel industry.
8.7.4 D-Lactic acid production

The five scenarios were economically compared as shown in Table 8.34. The total production cost was discriminated, in general terms, by raw materials, services, operatives, depreciation, and products and co-product sales.
If raw glycerol is used for the D-lactic acid production by mean of engineered E. coli strains this value is lower than 8 % of the total production cost. The values here obtained do not consider the transportation costs, since the D-lactic acid production process was assumed to be a biorefinery adjacent the biodiesel production process. Both utilities and capital costs represent the highest production cost which combined vary between 69.04 % for the Scenario 3 to 73.12 % for the Scenario 1. Also, it was noticed that utilities costs increased when glycerol concentration decreased, which makes sense due to the higher requirement of both water and heat for the fermentation stage. The opposite behavior occurred for the capital costs.
In order to obtain more detailed information about the D-lactic acid production from raw glycerol, the total production cost was divided in two processing sections: (i) Glycerol purification plus glycerol fermentation and (ii) D-lactic acid recovery and purification, as shown in Table 8.32. The first section (i.e., glycerol purification and fermentation) represents in all cases between 21 to 26 %, while the second section (i.e., D-lactic acid purification) is between 74 to 79 %. The D-lactic acid purification cost was estimated to be between 0.789 to 0.915 USD$/kg of L.A. Although these values are lower than that ones
162
reported by Joglekar et al. [55] for a similar downstream process (see: Table 8.15, Route 1), the esterification and hydrolysis by reactive distillation processes were not here considered. Also, a rigorous economic analysis was carried out meanwhile the purification costs obtained by Joglekar et al. [55] were based only on reported literature.
Table 8.34. Economic results for raw glycerol conversion to D-lactic acid: Cost (USD$/kg) and Share (%) Scenario 1 Raw materials Cost Share Utilities Cost Share Operating labor Cost Share Maintenance Cost Share Operating charges Cost Share Plant Overhead Cost Share G and A cost Cost Share Depreciation of capital Cost Share Co-products sales Cost Share Gly. Purify. + ferm. Cost Share L.A. purification Cost Share Total Sale price/production cost Cost 0.081 6.93 0.4556 38.95 0.0542 4.63 0.0488 4.17 0.0136 1.16 0.0515 4.40 0.0707 6.04 0.3997 34.17 -0.0055 -0.47 0.304 25.98 0.866 74.02 1.1696 1.327 Scenario 2 Scenario 3 Scenario 4 Scenario 5 0.0812 6.78 0.3932 32.85 0.0583 4.87 0.0628 5.25 0.0146 1.22 0.0606 5.06 0.0788 6.58 0.453 37.85 -0.0057 -0.48 0.282 23.56 0.915 76.44 1.1968 1.297 0.0789 7.16 0.2956 26.83 0.0565 5.13 0.0659 5.98 0.0141 1.28 0.0612 5.56 0.0699 6.34 0.465 42.21 -0.0054 -0.49 0.274 24.86 0.828 75.14 1.1017 1.409 0.0791 7.79 0.2844 28.02 0.0565 5.57 0.0579 5.70 0.0141 1.39 0.0572 5.64 0.0681 6.71 0.4468 44.02 -0.0492 -4.85 0.226 22.27 0.789 77.73 1.0149 1.530 0.0252 2.43 0.2736 26.35 0.0581 5.60 0.069 6.65 0.0146 1.41 0.0635 6.12 0.0703 6.77 0.4696 45.23 -0.0056 -0.54 0.219 21.10 0.819 78.90 1.0383 1.495
163
High differences can be observed for the total production cost of D-lactic acid from raw glycerol, which range from 1.0149 to 1.1968 USD$/Kg of L.A. The increasing order for the total production cost is: Scenario 4 < Scenario 5 < Scenario 3 < Scenario 1 < Scenario 2, which is completely related to the order found to the D-Lactic acid production. Also, the commercial sale price for D-lactic acid was compared to the production cost and it was found that this ratio (i.e., sale price/production cost) is higher than the unity for all scenarios, and for the Scenario 4 this ratio was 1.53. These results indicate that D-lactic acid production from raw glycerol by mean of engineered E. coli could be a profitable alternative for glycerol usage.
The lowest total production cost of D-lactic acid from raw glycerol was obtained for the Scenario 4 in which three fermentative advantages are joined simultaneously: (i) use of crude glycerol (85 wt %), (ii) high glycerol concentration in the fermentation media (40 g/l), and (iii) total consumption of glycerol. Thus, the economical successful of a biotechnological scheme depends highly on the fermentation stage and the above described three characteristics could lead to high process performance. In this order of ideas, one of the main purposes of metabolic engineering should be to develop specific strains able to consume completely raw substrates at high concentrations.
8.7.5 Succinic acid production

The five scenarios were economically assessed and the results are shown in Table 8.34. These production costs were discriminated by raw materials, services, operatives, depreciation, and products and co-product sales. The values here obtained do not consider the transportation costs, since the succinic acid production process was assumed to be a biorefinery adjacent the biodiesel production process.
In all cases the raw material costs represent only between 2.8 to 3.91 %, while the utilities are the highest share of the total production costs with a weight between 65.9 to 70.12 % followed by the capital costs with a share among 11.23 to 14.55 %.
164
Table 8.35. Economic results for raw glycerol conversion to succinic acid: Cost (USD$/kg) and Share (%) Scenario 1 Raw materials Cost Share Utilities Cost Share Operating labor Cost Share Maintenance Cost Share Operating charges Cost Share Plant Overhead Cost Share G and A cost Cost Share Depreciation of capital Cost Share Co-products sales Cost Share Gly. Purify. + ferm. Cost Share S.A. purification Cost Share Total Sale price/production cost Cost 0.0906 3.91 1.5673 67.60 0.0583 2.51 0.0366 1.58 0.0146 0.63 0.0474 2.05 0.1720 7.42 0.3373 14.55 -0.0055 -0.24 0.3104 13.39 2.008 86.61 2.319 1.0747 Scenario 2 Scenario 3 Scenario 4 Scenario 5 0.0908 3.00 2.0850 68.95 0.0587 1.94 0.0549 1.82 0.0147 0.49 0.0568 1.88 0.2882 9.53 0.3804 12.58 -0.0055 -0.18 0.3205 10.60 2.703 89.40 3.024 0.8241 0.0885 3.41 1.7877 68.79 0.0556 2.14 0.0446 1.72 0.0139 0.53 0.0501 1.93 0.2547 9.80 0.3097 11.92 -0.0060 -0.23 0.2837 10.92 2.315 89.08 2.599 0.9589 0.0885 3.33 1.8644 70.12 0.0503 1.89 0.0458 1.72 0.0126 0.47 0.0481 1.81 0.2561 9.63 0.2986 11.23 -0.0055 -0.21 0.2516 9.46 2.407 90.54 2.659 0.9373 0.0908 2.79 2.1486 65.93 0.0584 1.79 0.0669 2.05 0.0146 0.45 0.0626 1.92 0.4072 12.50 0.4152 12.74 -0.0055 -0.17 0.3104 9.52 2.949 90.48 3.259 0.7647
Also, it was noticed that utilities costs increased with both the global fermentation yield (see Table 8.20) and the low concentration of glycerol in the fermentation media. This behavior is due to both the inefficient usage of glycerol during the fermentation process and the higher requirement of serves.
165
The total production costs were divided in two processing sections: (i) Glycerol purification plus glycerol fermentation and (ii) succinic acid recovery and purification, as shown in Table 8.32. Thus, more detailed information about the succincic acid production from raw glycerol can be obtained. In this way, glycerol purification and fermentation represent, in all cases, lower than 13.4 % indicating that most of the total production cost of succinic acid from raw glycerol are in the recovery and purification processes (> 86.5 %).
Regarding to the succinic acid purification costs, high differences were found among the five scenarios since this value varies from 2.008 to 2.949 USD$/Kg of S.A, being its relative difference almost 50%. These differences found in the purification costs are mainly related to the complex behavior of the fermentation process, due to variables such as glycerol concentration, glycerol purity, glycerol consumption, and yield to succinic acid affects the performance of the global process. Thus, the increasing order for succinic acid recovery and purification costs and for total production costs of succinic acid from raw glycerol are the same, it is Scenario 1 < Scenario 3 < Scenario 4 < Scenario 2 < Scenario 5.
On the other hand, the commercial sale price for succinc acid was compared to its production cost and it was found that this ratio (i.e., sale price/production cost) was only higher than the unity for the Scenario 1, but this value is still too close to the unity. Thus, it can be stated that the succinic acid production from glycerol still requires performing high effort in order to make this process profitable since the its production costs is high yet.
8.7.6 Propionic acid production

The five analyzed scenarios were economically assessed as shown in Table 8.36 and the production costs results were discriminated by raw materials, services, operatives, depreciation, and products and co-product sales. The values here obtained do not consider the transportation costs, since the propionic acid production process was assumed to be a biorefinery adjacent the biodiesel production process.
Low raw material costs were obtained in all cases when raw glycerol is used, representing only around 5 % of the total production costs. But both utilities (between 45.8 to 58.9 %) and capital (between 20.8 to 29.9 %) costs represent the highest production cost which
166
combined vary between 75.76 % for the Scenario 3 to 79.69 % for the Scenario 1 and the Scenario 5. Also, it was found that the utilities costs depend mainly on the glycerol concentration in the fermentation media.
Table 8.36. Economic results for raw glycerol conversion to propionic acid: Cost (USD$/kg) and Share (%). Scenario 1 Raw materials Cost Share Utilities Cost Share Operating labor Cost Share Maintenance Cost Share Operating charges Cost Share Plant Overhead Cost Share G and A cost Cost Share Depreciation of capital Cost Share Co-products sales Cost Share Gly. Purify. + ferm. Cost Share L.A. purification Cost Share Total Sale price/production cost Cost 0.0851 4.41 1.1355 58.89 0.0561 2.91 0.0342 1.78 0.0140 0.73 0.0452 2.34 0.1625 8.43 0.4011 20.80 -0.0055 -0.28 0.3100 16.07 1.618 83.93 1.928 0.985 Scenario 2 Scenario 3 Scenario 4 Scenario 5 0.0851 5.05 0.8192 48.67 0.0624 3.71 0.0651 3.87 0.0156 0.93 0.0638 3.79 0.1161 6.90 0.4614 27.41 -0.0055 -0.33 0.3108 18.46 1.372 81.54 1.683 1.129 0.0851 4.71 0.8275 45.85 0.0706 3.91 0.0783 4.34 0.0176 0.98 0.0744 4.12 0.1171 6.49 0.5398 29.91 -0.0055 -0.30 0.3104 17.20 1.495 82.80 1.805 1.053 0.0851 4.38 0.9709 49.99 0.0745 3.83 0.0841 4.33 0.0186 0.96 0.0793 4.08 0.0636 3.27 0.5717 29.43 -0.0055 -0.28 0.2959 15.23 1.647 84.77 1.942 0.978 0.0851 4.59 0.9497 51.22 0.0691 3.73 0.0762 4.11 0.0173 0.93 0.0727 3.92 0.0618 3.33 0.5279 28.47 -0.0055 -0.30 0.3104 16.74 1.544 83.26 1.854 1.025
167
In addition, the total production cost was divided in two processing sections: (i) glycerol purification plus glycerol fermentation and (ii) propionic acid recovery and purification, as shown in Table 8.36. The glycerol purification and glycerol fermentation stages only represented between 15.23 to 18.46 %, which indicates that most efforts are required for the downstream process since a high share of the total production costs (between 81.54 to 84.77 %) is consumed for the propionic acid purification.
Total production costs of propionic acid from raw glycerol range from 1.683 to 1.942 USD$/Kg of P.A., which represents a relative difference of 15.4 %. Being the increasing order for the total production costs: Scenario 2 < Scenario 3 < Scenario 5 < Scenario 1 < Scenario 4.
The ratio between the sale price to the production cost was calculated for the five analyzed scenarios, and this ratio was higher than the unity only for the Scenario 2, Scenario 3, and Scenario 5. Thus, it is stated that this process could be profitable only when high concentration of glycerol is used in the fermentation media and high yield to propionic acid is obtained.
8.8 Conclusions
In the past, important efforts have been made in order to introduce the biotechnological production of 1,3-propanediol from glycerol to the industry. However, research tendencies were focused on microorganism development and on the analysis of some specific process conditions. On the other hand, the drastic increment in the use of biodiesel has caused an oversupply of glycerol in the market. Thus, in order to suitably use the glycerol obtained from the biodiesel industry, the mass production of 1,3-propanediol from glycerol needs additional analysis. Here, a complete technological scheme for its production has been not only proposed by also assessed. Even more, the fermentation stage was optimized by three different ways and as result it was possible to assess economically three scenarios where the third one had the lowest production cost. The relative differences respect to the first and the second scenarios were 7.2% and 13.0 %. The obtained results provide enough information to understand the different possibilities for process intensification using this technology and also to compare it with other new industrial alternatives for the utilization of glycerol as a raw material.
168
Due to the low cost of raw glycerol, methanol recovery from glycerol implies low purification costs. Meanwhile, the three possibilities assessed for glycerol bioconversion to ethanol showed that the global production costs of fuel ethanol from raw glycerol are lower than the commercial price of fuel ethanol. These facts show the potential for raw glycerol bioconversion to fuel ethanol using E. coli. Also, the comparison carried out with a previous paper (which considers the fuel ethanol production from sugarcane and corn in the Colombian case) shows that the global production costs of fuel ethanol from raw glycerol can be as profitable as the production of fuel ethanol from conventional raw materials as sugarcane. The latter is a completely developed industry in Colombia.
Production of ethanol from glycerol is presented as an alternative which uses a renewable resource that does not generate direct or indirect competition with the food industry, property that do have the raw materials used in commercial (maize, cassava Beets, sugar cane, molasses, etc.). The techno-economic feasibility of this process was demonstrated through simulation and evaluation of processes and found that the cost of production is similar to the process established commercially in Colombia from sugar cane. The environmental performance of this process was not the best of those obtained for the different raw materials, but significant higher compared with the process benefits from corn.
Biodiesel and ethanol can be jointly produced using oil palm as sole source by mean of processes integration, such as the biodiesel production with the ethanol production from two feedstocks: lignocellulosic residues (empty fruit bunches and palm press fiber produced during) and crude glycerol. Thus, alcohol is completely self-supplied by the integrated process and low quantities of wastes are produced without any global production of glycerol. Economical evaluation showed a higher biodiesel production cost for the integrated process than the traditional biodiesel production process which uses ethanol and palm oil as feedstocks. But the first one is a promising technology available to build an autonomous biodiesel production plant with low waste levels. This process must be economically improved by further analysis from a process design view point, based on process simulation which showed be a powerful tool for performing processes integration. Three technological schemes to produce PHB from crude glycerol were analyzed under two fermentation conditions (i.e., using glycerol at 88 wt % and 98 wt %). In this work it was found that it is better to use pure glycerol as feedstock for the production of PHB than
169
raw glycerol. This phenomenon is explained by the fact that the higher PHB yield reduces the utility costs in the downstream process. The results shown here are important for the industrial production of PHB using glycerol as a raw material. Currently, in the biodiesel production the total profitability of any new project could be determined by the right use of glycerol as a massive by-product. The proposed strategy to use pure glycerol as substrate can be understood as a very interesting alternative since the final composition in the glycerol streams depends on the source of the feedstock used for biodiesel production. Also, most of the biotechnological alternatives to produce added value compounds from glycerol are sensitive to contaminants in the raw material. Thus, several technical and economical advantages as well as a more stable production of PHB are obtained when a standardized raw material as pure glycerol is used.
Usage of raw glycerol, engineered Escherichia coli strains, and processes integration for the production of optically pure D-lactic acid is an important alternative to transform the by-produced glycerol during the biodiesel synthesis. Although five different configurations for the fermentation stage were considered, in all cases the total production costs were lower than its sale price. Thus, the whole process scheme for D-lactic acid production could be considered as potentially profitable design. Also, it was found that the combined effect of both high glycerol concentration and use of low quality glycerol in the fermentation media, lead to the best economic performance. The results shown here are important for the industrial production of D-lactic acid using glycerol as a raw material. On the other hand, in the biodiesel production the total profitability of any new project could be determined by the right use of glycerol as a massive raw material.
Otherwise, for succinic acid and propionic acid production results showed that most efforts are required in order to reduce the downstream process since the total production costs were too close to the sale price of these carboxylic acids.
References
[1] Gonzalez-Pajuelo, M., Meynial-Salles, I., Mendes, F., Andrade, J.C., Vasconcelos, I., and Soucaille, P.: Metabolic engineering of Clostridium acetobutylicum for the industrial production of 1,3-Propanediol from glycerol. Metab. Eng. 7, 329-336 (2005).
170
[2] Ito, T., Nakashimada, Y., Senba, K., Matsui, T., and Nishio N.: Hydrogen and ethanol production from glycerol-containing wastes discharged after biodiesel manufacturing process. J. Biosci. Bioeng.. 100, 260265 (2005). [3] Zhao, Y.N., Chen, G., and Yao, S.J.: Microbial production of 1,3-propanediol from glycerol by encapsulated Klebsiella pneumoniae. Biochem. Eng. J. 32, 9399 (2006). [4] Menzel, K., Zeng, A.-P., and Deckwer, W.-D.: High concentration and productivity of 1,3-propanediol from continuous fermentation of glycerol by Klebsiella pneumoniae. Enzyme Microb. Technol. 20, 8286 (1997). [5] Xiu, Z.L., Song, B.H., Sun L.H., and Zeng, P.: Theoretical analysis of effects of metabolic overflow and time delay on the performance and dynamic behavior of a twostage fermentation process. Biochem. Eng. J. 11, 101109 (2002). [6] Xiu, Z.L., Song, B.H., Wang, Z.T., Sun, L. H., Feng, E. M., and Zeng, A.P.: Optimization of dissimilation of glycerol to 1,3-Propanediol by Klebsiella pneumoniae in one- and two-stage anaerobic cultures. Biochem. Eng. J. 19, 189-197 (2004). [7] Edgar, T.F., Himmelblau, D.M., and Lasdon L.S.: Optimization of chemical processes. Second Edition, McGraw Hill, New York (2001). [8] Raval, K.N., Hellwig, S., Prakash, G., Ramos-Plasencia, A., Srivastava, A., and Biichs. J.: Necessity of a two-stage process for the production of azadirachtin-related limonoids in suspension cultures of Azadirachta indica. J. Biosci. Bioeng.. 96, 16-22 (2003). [9] Janusz J. Malinowski. Reactive Extraction for Downstream Separation of 1,3Propanediol. Biotechnol. Prog. 2000, 16, 76-79. [10] Jian Hao, Hongjuan Liu, and Dehua Liu. Novel Route of Reactive Extraction To Recover 1,3-Propanediol from a Dilute Aqueous Solution. Ind. Eng. Chem. Res. 2005, 44, 4380-4385. [11] Jian Hao, Feng Xu, Hongjuan Liu and Dehua Liu. Downstream Processing of 1,3Propanediol Fermentation Broth. J. Chem. Technol. Biotechnol. 81:102108 (2006). [12] Serafimov L. A., Zharov V. T., Timofeyev V. S., Rectification of multicomponent mixtures. I. Topological analysis of liquid vapor phase equilibrium diagrams, Acta Chim. Acad. Sci. Hung. 1971, Tomus 69, 4, 383.
171
[13] Pisarenko Y. A., Serafimov L. A., Cardona, C. A., Efremov, D.L., Shuwalov, A.S., Reactive distillation design: analysis of the process statics, Reviews in Chemical Engineering 17 (4), 2001. [14] Serafimov L. A., Pisarenko Yu. A., Kulov N. N., Coupling Chemical Reaction with Distillation: Thermodynamic Analysis and Practical Applications, Chemical Engineering Science 54 (1999) 1383-1388. [15] Giessler S., Danilov R. Y., Pisarenko R. Y., Serafimov L. A., Hasebe S., and Hashimoto I., Feasible Separation Modes for Various Reactive Distillation Systems. Ind. Eng. Chem. Res. 1999, 38, 4060-4067 [16] Giessler S., Danilov R. Y., Pisarenko R. Y., Serafimov L. A., Hasebe S., Hashimoto I.. Systematic structure generation for reactive distillation processes. Computers and Chemical Engineering 25 (2001) 4960 [17] Quintero JA, Montoya MI, Snchez OJ, Giraldo OH, Cardona CA. Fuel ethanol production from sugarcane and corn: Comparative analysis for a Colombian case. Energy 2008;33:385-399. [18] Cardona CA, Snchez OJ. Fuel ethanol production: Process design trends and integration opportunities. Bioresour. Technol 2007;98:2415-2457. [19] Raskovic P, Anastasovski A, Markovska Lj, Mesko V, Process integration in bioprocess indystry: waste heat recovery in yeast and ethyl alcohol plant. Energy 2010;35:704717. [20] Yu S, Tao J. Energy efficiency assessment by life cycle simulation of cassava-based fuel ethanol for automotive use in Chinese Guangxi context. Energy 2009;34:22-31. [21] Yu S, Tao J. Simulation-based life cycle assessment of energy efficiency of biomassbased ethanol fuel from different feedstocks in China. Energy 2009;34:476484. [22] Cardona CA, Snchez OJ. Energy consumption analysis of integrated flowsheets or production of fuel ethanol from lignocellulosic biomass. Energy 2006;31:2447-2459. [23] Cardona CA, Quintero JA, Paz IC. Production of bioethanol from sugarcane bagasse: Status and perspectives. Bioresour Technol 2009;101:4754-4766.
172
[24] Gutirrez LF, Snchez J, Cardona CA. Process integration possibilities for biodiesel production from palm oil using ethanol obtained from lignocellulosic residues of oil palm industry. Bioresour Technol 2009;100:1227-1237. [25] Felix E, Tilley DR. Integrated energy, environmental and financial analysis of ethanol production from cellulosic switchgrass. Energy 2009;34:410436. [26] Dharmadi Y, Murarka A, Gonzalez R. Anaerobic fermentation of glycerol by Escherichia coli: a new platform for metabolic engineering. Biotechnol Bioeng 2006;94:821829. [27] Yazdani SS, Gonzalez R. Engineering Escherichia coli for the efficient conversion of glycerol to ethanol and co-products. Metab. Eng. 2008;10:340351 [28] Nielsen J, Villadsen J, Liden G. Bioreaction engineering principles. Kluwer Academic/Plenum Publishers, New York, 2003. [29] J. M. Hernndez-Salas, M. S. Villa-Ramrez, J. S. Veloz-Rendn, K. N. RiveraHernndez, R. A. Gonzlez-Csar, M. A. Plascencia-Espinosa, S. R. Trejo-Estrada. (2009) Bioresource Technology 100 1238-1245. [30] C. A. Cardona Alzate, O. J. Snchez Toro. (2006) Energy 31 2447-2459. [31] M. Rivera, C.A. Cardona. (2004) Ingeniera y Competitividad 6 [32] L. F. Gutirrez (2008) Ph.D. thesis National University of Colombia at Manizales Manizales, Colombia [33] A. Abdul Aziz, M. Husin, A. Mokhtar. (2002) Journal of Oil Palm Research 14 9-14. [34] A. Abdul Aziz, K. Das, M. Husin, A. Mokhtar. (2002) Journal of Oil Palm Research 14 10-17. [35] M. W. Zahari, A.R. Alimon. (2004) Palm Oil Developments 40 5-9. [36] S. Prasertsan, P. Prasertsan. (1996) Biomass and Bioenergy 11 387-395. [37] Posada JA, Cardona CA. 2010. Anlisis de la refinacin de glicerina obtenida como co-producto en la produccin de biodiesel (Validation of glycerin refining obtained as a by-Product of biodiesel production). Ingeniera y Universidad 14:2-27.
173
[38] Posada JA, Cardona CA, Rincn LE. Sustainable biodiesel production from palm using in situ produced glycerol and biomass for raw bioethanol. In 32nd symposium on biotechnology for fuels and chemicals. Clearwater Beach, Florida. April 19-22. 2010. [39] Mothes G, Schnorpfeil C, Ackermann CJU. Production of PHB from crude glycerol. Eng Life Sci 2007;7(5);475479. [40] Cavalheiro JMBT, de Almeida MCMD, Grandfilis C, da Fonseca MMR. Poly(3hydroxybutyrate) production by Cupriavidus necator using waste glycerol. Process Biochem 2009;44(5);509-515. [41] Jacquel N, Lo CW, Wei YH, Wu HS, Wang SS. Isolation and purification of bacterial poly(3-hydroxyalkanoates). Biochem Eng J 2008;39;1527. [42] Tamer M, Moo-Young M., Chisti Y. Disruption of Alcaligenes latus for recovery of poly (-hydroxybutyric acid): comparison of high-pressure homogenization, beadmilling and chemically induced lysis, Ind Eng Chem Res 1998;37;18071814. [43] Chen Y, Chen Y, Chen J, Yang H. Kinetics of PHB-containing biomass disruption in surfactantchelate aqueous solution. Process Biochem 2003:38;1173-1182. [44] Hejazi P, Vasheghani-Farahani E, Yamini Y. Supercritical uid disruption of Ralstonia eutropha for poly ( -hydroxybutyrate) recovery. Biotechnol Progr 2003;19;15191523. [45] Asrar J, Grys KJ. Biopolymers Vol. 4: Polyesters III - Applications and commercial products. Wiley-VCH, Munster; 2002. p 398. [46] Harding KG, Dennis JS, von Blottnitz H, Harrison STL. Environmental analysis of plastic production processes: Comparing petroleum-based polypropylene and
polyethylene with biologically-based poly-B-hydroxybutyric acid using life cycle analysis. J. Biotechnol. 2007;130;5766. [47] Kapritchkoff FM, Viotti AP, Alli RCP, Zuccolo M, Pradella JGC, Maiorano AE, Miranda EA, Bonomi A. Enzymatic recovery and purification of polyhydroxybutyrate produced by Ralstonia eutropha, J Biotechnol 2006;122;453462.
174
[48] Dien, B. S., N. N. Nichols, and R. J. Bothast. 2001. Recombinant Escherichia coli engineered for production of L-lactic acid from hexose and pentose sugars. J. Ind. Microbiol. Biotechnol. 27:259264. [49] Okano, K., T. Tanaka, C. Ogino, H. Fukuda, and A. Kondo. 2010. Biotechnological production of enantiomeric pure lactic acid from renewable resources: recent achievements, perspectives, and limits. Appl. Microbiol. Biotechnol. 85:413423. [50] Zhou, S., K. T. Shanmugam, and L. O. Ingram. 2003. Functional replacement of the Escherichia coli D()-lactate dehydrogenase gene (ldhA) with the L(+)-lactate
dehydrogenase gene (ldhL) from Pediococcus acidilactici. Appl. Environ. Microbiol. 69:22372244. [51] Zhou, S., T. B. Causey, A. Hasona, K. T. Shanmugam, and L. O. Ingram. 2003. Production of optically pure D-lactic acid in mineral salts medium by metabolically engineered Escherichia coli W3110. Appl. Environ. Microbiol. 69:399407. [52] Zhu, Y., M. A. Eiteman, K. DeWitt, and E. Altman. 2007. Homolactate fermentation by metabolically engineered Escherichia coli strains. Appl. Environ. Microbiol. 73:456464. [53] Mazumdar, S., Clomburg, J.M., and Gonzalez, R. 2010. Escherichia coli Strains Engineered for Homofermentative Production of D-Lactic Acid from Glycerol. Applied and Environmental Microbiology. 76: 43274336. [54] Hong, A.A.; Cheng, K.K.; Peng, F.; Zhou, S.; Sun, Y.; Liu, C.M.; Liu, D.H. Strain isolation and optimization of precess parameters for bioconversion of glycerol to lactic acid. J. Chem. Technol. Biotechnol., 2009, 84, 1576-1581. [55] Joglekar, H.G., Rahman, I., Babu, S., Kulkarni, B.D., Joshi, A. 2006.Comparative assessment of downstream processing options for lactic acid. Separation and Purification Technology. 52:117 [56] Tung L.A., and C.J. King, Ind. Eng. Chem. Res. 33 (1994) 32173223. [57] Kaufman E.N., S.P. Cooper, S.L. Clement, M.H. Little, Appl. Biochem. Biotechnol. 5152 (1995) 605620. [58] Evangelista L.R., Z.L. Nikolov, Appl. Biochem. Biotechnol. 5758 (1996) 471480.
175
[59] Kulprathipanja S., A.R. Oroshar, US Patent 5,068,418, 1991 [60] Zheng Y.J., X.H. Ding, P.L. Cen, C.W. Yang, G.T. Tsao, Appl. Biochem. Biotechnol. 5758 (1996) 627632. [61] Antonio G.R., G. Vaccari, E. Dosi, A. Trilli, M. Rossi, D. Matteuzzi, Biotechnol. Bioeng. 67 (2000) 147156. [62] Srivastava A., P.K. Roychoudhury, V. Sahai, Biotechnol. Bioeng. 39 (1992) 607613 [63] Cao X., H.S. Yun, Y.M. Koo, Recovery of lactic acid by anion-exchange resin Amberlite IRA-400. Biochem. Eng. J. 11 (23) (2002) 189196. [64] Tong W.-Y., X.-Y. Fu, S.-M. Lee, J. Yu, J.-W. Liu, D.-Z.Wei, Y.-M. Koo, Biochem. Eng. J. 18 (2) (2004) 8996. [65] Raya-Tonetti G., P. Cordoba, J. Bruno-barcena, F. Sineriz, N. Perotti, Biotechnol. Tech. 13 (3) (1999) 201205. [66] Vaccari G., A. Gonzalez-Vara, A.L. Campi, E. Dosi, P. Brigidi, Appl. Microbiol. Biotechnol. 40 (1993) 2327. [67] Senthuran A., V. Senthuran, B. Mattiasson, R. Kaul, Biotechnol. Bioeng. 55 (1996) 841853. [68] Kaufman E.N., S.P. Cooper, S.L. Clement, M.H. Little, Appl. Biochem. Biotechnol. 5152 (1995) 605620. [69] Lee H.-J., Y. Xie, Y.-M. Koo, N.-H.L.Wang, Biotechnol. Prog. 20 (2004) 179192. [70] Chen C.-C., L.-K. Ju, Sep. Sci. Technol. 33 (10) (1998) 14231437. [71] Min-tian G., M. Hirata, M. Koide, H. Takanashi, T. Hano, Process Biochem. 39 (2004) 19031907. [72] Bailly M., H. Roux-de Balmann, P. Aimar, F. Lutin, M. Cheryan, J. Membr. Sci. 191 (2001) 129142. [73] Bailly M., Desalination 144 (13) (2002) 157162.
176
[74] Habova V., K. Melzoch, M. Rychtera, B. Sekavova, Desalination 163 (2004) 361 372. [75] Li H., R. Mustacchi, C.J. Knowles,W. Skibar, G. Sunderland, I. Dalrymple, S.A. Jackman, Tetrahedron 60 (2004) 655661. [76] Hong, Y. K. and W. H. Hong (2000) Reactive extraction of succinic acid with tripropylamine (TPA) in various diluents. Biopro. Eng. 22: 282-284 [77] Hong Y.K., Hong W.H., and Han D.H., Application of Reactive Extraction to Recovery of Carboxylic Acids. Biotechnol. Bioprocess Eng. 2001, 6: 386-394. [78] Tamada J.A., A.S. Kertes, and C.J. King. Extraction of Carboxylic Acids with Amine Extractants. 1. Equilibria and Law of Mass Action Modeling. Ind. Eng. Chem. Res. 1990, 29, 1319-1326. [79] Yabannavar V.M., D.I.C.Wang, Biotechnol. Bioeng. 37 (1991a) 716722. [80] Yabannavar V.M., D.I.C. Wang, Biotechnol. Bioeng. 37 (1991b) 10951100. [81] Seo Y., W.H. Hong, T.H. Hong, Korean J. Chem. Eng. 16 (5) (1999) 556561 [82] Kim J.Y., Y.J. Kim, W.H. Hong, G. Wozny, Biotechnol. Bioprocess Eng. 5 (2000) 196201. [83] Kim Y.J., W.H. Hong, G. Wozny, Korean J. Chem. Eng. 19 (5) (2002) 808814. [84] Choi I., W.H. Hong, J. Chem. Eng. Jpn. 32 (1999) 184189. [85] Nielsen J, Villadsen J, Liden G. Bioreaction engineering principles. New York: Kluwer Academic/Plenum Publishers; 2003. [86] Posada JA, Cardona CA. 2010. Design and analysis of fuel ethanol production from raw glycerol. Energy. 35(12):5286-5293. [87] Posada JA, Cardona CA. 2010. Anlisis de la refinacin de glicerina obtenida como co-producto en la produccin de biodiesel (Validation of glycerin refining obtained as a by-Product of biodiesel production). Ingeniera y Universidad 14:2-27.
177
[88] Posada JA, Naranjo JM, Lpez JA, Higuita JC, Cardona CA. 2011a. Design and Analysis of PHB Production Processes from Crude Glycerol. Process Biochemistry. 46:310-317. [89] Datta R (1992) Process for the production of succinic acid by anaerobic fermentation. US patent 5,143,833 [90] Datta R, Glassner DA, Jain MK, Vicky Roy JR (1992) Fermentation and purification process for succinic acid. US patent 5,168,055 [91] Hermann BG, Patel M (2007) Todays and tomorrows bio-based bulk chemicals from white biotechnologya techno-economic analysis. Appl Biochem Biotechnol 136: 361388 [92] Berglund KA, Yedur S, Dunuwila D (1999) Succinic acid production and purification. US patent 5,958,744 [93] Yedur S, Berglung KS, Dunuwila DD (2001) Succinic acid production and purification. US patent 6,265,190 [94] Glassner DA, Satinwood RD, Datta R (1992) Process for the production and purification of succinic acid. US patent 5,143,834 [95] Zeikus JG, Jain MK, Elankovan P (1999) Biotechnology of succinic acid production and markets for derived industrial products. Appl Microbiol Biotechnol 51:545552 [96] Berglund KA, Elankovan P, Glassner DA (1991) Carboxylic acid purification and crystallization process. US patent 5,034,105 [97] McKinlay JB, Vieille C, Zeikus JG (2007) Prospects for a biobased succinate industry. Appl Microbiol Biotechnol. doi: 10.1007/s00253-007-1057-y [98] Jaquet A, Quan R, Marison W, vain Stockar U (1999) Factors influenceing in potentioluse of aliquat 336 for the in situ extraction of carboxylic acids from cultures of Pseudomonas putida. J Biotechnol 68:185 [99] King CJ, Dtarr J (1990) Recovery of carboxylic acids from water by precipitation from organic solutions. US patent 5,104,492
178
[100] King CJ, Poole LJ (1993) Craboxylic acid sorption regeneration process. US patent 5,412,126 [101] Kurzrock T. Weuster-Botz D. Recovery of succinic acid from fermentation broth. Biotechnol Lett (2010) 32:331339. [102] Tamada JA, King CJ (1990a) Extraction of carboxylic acids with amine extractants: 2. Chemical interactions and interpretation of data. Ind Eng Chem Res 29:13271333 [103] Tamada JA, King CJ (1990b) Extraction of carboxylic acids with amine extractants: 3. Effects of temperature, water coextraction and process considerations. Ind Eng Chem Res 29:13331338 [104] Hong YK, Hong WH, Han DH (2001) Application of reactive extraction to recovery of carboxylic acids. Biotechnol Bioproc Eng 6:386394 [105] Tung LA, King CJ (1994a) Sorption and extraction of lactic and succinic acids at pH[pKa1. 1. Factors governing equilibria. Ind Eng Chem Res 33:32173223 [106] Hong YK, Hong WH (2000a) Equilibrium studies on reactive extraction of succinic acid from aqueous solutions with tertiary amines. Bioproc Eng 22:477481 [107] Hong YK, Hong WH (2000b) Reactive extraction of succinic acid with tripropylamine (TPA) in various diluents. Bioproc Eng 22:281284 [108] Hong YK, Hong WH (2000c) Extraction of succinic acid with 1-octanol/n-heptane solutions of mixed tertiary amine. Bioproc Eng 23:535538 [109] Hong YK, Han DH, Hong WH (2002) Water enhanced solubilities of succinic acid in reactive extraction using tertiary amines/alcohol systems. Korean J Chem Eng 19:8386 [110] Hong YK, Hong WH (2004) Influence of chain length of tertiary amines on extractability and chemical interactions in reactive extraction of succinic acid. Korean J Chem Eng 21:488493. [111] Song H, Huh YS, Lee SY, Hong WH, Hong YK (2007) Recovery of succinic acid produced by fermentation of a metabolomically engineered Mannheimia
succiniciproducens strain. J Biotechnnol 132:445452
179
[112] Lee SY, Kim JM, Song H, Lee JW, Kim TY, Yang YS (2008) From genome sequence to integrated bioprocess for succinic acid production by Mannheimia succiniciproducens. Appl Microbiol Biotechnol 79:1122 [113] Juang RS, Huang RH (1997) Equilibrium studies on reactive extraction of lactic acid with an amine extractant. Chem Eng J 65:4753 [114] Heyberger A, Prochazka J, Volaufova E (1997) Extraction of citric acid with tertiary aminethird phase formation. Chem Eng Sci 97:0032100327 [115] Huh YS, Jun Y-S, Kong YK, Song H, Lee SY, Hong WH (2006) Effective purification of succinic acid from fermentation broth produced by Mannheimia succiniciproducens. Proc Biochem 41:14611465 [116] Li Q, Li WL, Wang D, Liu BB, Tang H, Yang MH, Liu Q, Xing JM, Su ZG (2008) pH neutralization while succinic acid adsorption onto anion-exchange resins. Appl Biochem Biotechnol. doi:10.1007/s12010-008-8355-4 [117] Kushiku T, Fujiwara K, Satou T, Sano (2006) Method for purifying succinic acid from fermentation broth. US patent 2006/0276674A1. [118] Jun YS, Huh YS, Park HS, Thomas A, Jeon SJ, Lee EZ, Won HJ, Hong WH, Lee SY, Hong YK (2007) Adsorption of pyruvic and succinic acid by amine-functionalized SBA-15 for the purification of succinic acid from fermentation broth. J Phys Chem C 200(111):1307613086. [119] Pai RA, Doherty MF, Malone MF (2002) Design of reactive extraction systems for bioproduct recovery. AIChE J 48:514526 [120] Blankschien M.D., Clomburg J.M., Gonzalez R. Metabolic engineering of Escherichia coli for the production of succinate from glycerol. Metabolic
Engineering12(2010)409419 [121] Scholten E. and Dirk Dagele. Succinic acid production by a newly isolated
bacterium. Biotechnol Lett (2008) 30:21432146.
180
[122] Lee P.C., Lee W.G., Lee S.Y., Chang H.N. Succinic Acid Production with Reduced By-Product Formation in the Fermentation of Anaerobiospirillum succiniciproducens Using Glycerol as a Carbon Source. Biotechnology and Bioengineering, 72:2143-2146, 2001. [123] Thompson, J.C., He, B.B., Characterization of crude glycerol from biodiesel production for multiple feedstocks. Appl. Eng. Agric. 2006, 22(2):261-265. [124] Huh YS, Hong YK, Hong WH, Chang HN (2004) Selective extraction of acetic acid from the fermentation broth produced by Mannheimia succiniciproducens. Biotechnol Lett 26:15811584. [125] Zhu Y., Li J., Tan M., Liu L., Jiang L., Sun J., Lee P., Du G., Chen J., Optimization and scale-up of propionic acid production by propionic acid-tolerant Propionibacterium acidipropionici with glycerol as the carbon source. Bioresource Technology 101 (2010) 89028906. [126] Zhang A., Yang S.T., Propionic acid production from glycerol by metabolically engineered Propionibacterium acidipropionici. Process Biochemistry 44 (2009) 1346 1351 [127] Hauer, E. and Marr, R. (1994). Liquid Extraction in Biotechnology, International Chemical Engineering, 34(2), pp. 178-187. [128] Timmer, J. K. M., Kromkamp, J. and Robbertsen, T. (1994). Lactic acid separation from fermentation broth by reverse osmosis and nanofiltration, Journal of Membrane Science, 92, pp. 185-197 [129] Nomura, Y., Iwahara, M. and Hongo, M. (1987). Lactic acid production by electrodialysis fermentation using immobilized growing cells, Biotechnology and Bioengineering, 30, pp. 788-793 [130] Sirman, T., Pyle, D. L. and Grandison, A. S. (1991) Extraction of organic acids using a supported liquid membrane, Biochemical Society Transactions, 19 (3), pp. 274-279 [131] Dai, Y. and King, J. (1996). Selectivity between lactic acid and glucose during recovery of lactic acid with basic extractants and polymeric sorbents, Industrial Engineering and Chemistry Research, 35, pp. 1215-1224.
181
[132] Kertes, A. S. and King, C. (1986). Extraction Chemistry of Fermentation product Carboxylic Acids, Biotechnology and Bioengineering, 28, 269-282. [133] Keshav A., Wasewar K., Chand S., Extraction of Propionic Acid Using Different Extractants (Tri-n-butylphosphate, tri-n-octylamine and Aliquat 336). Industrial & Engineering Chemistry Research 2008, 47, 61926196. [134] Keshav A., Chand S., Wasewar K.L., Equilibrium studies for extraction of propionic acid using tri-n-butyl phosphate in different solvents. Journal of Chemical and Engineering Data, 2008, 53, (7), 14241430 [135] Keshav A., Chand S., Wasewar K., Equilibrium and kinetics of extraction of propionic acid using tri-n-octylphosphineoxide. Chemical Engineering and Technology 2008, 31 (9), 1290-1295. [136] Keshav A., Wasewar K., Chand S., Reactive extraction of propionic acid with trinoctylamine in different diluents. Separation and Purification Technology 2008, 63, 179183. [137] Keshav A., Wasewar K.and Chand S., Study of binary extractants and modifier diluents systems for reactive extraction of propionic acid, Fluid Phase Equilibria 2008, 275, 2126. [138] Keshav A., Chand S., Wasewar K., Recovery of propionic acid from aqueous stream by reactive extraction: Effect of diluents. Desalination 2009, 244, 12-23. [139] Keshav A., Chand S., Wasewar K., Recovery of propionic acid by reactive extraction using tri-n-butyl phosphate in petroleum ether: Equilibrium study, Chemical and Biochemical Engineering Quarterly 2008, 22(4), 433-437. [140] Keshav A., Wasewar K.L., Chand S., Extraction of acrylic, propionic and butyric acid using Aliquat 336 in oleyl alcohol: Equilibria and effect of temperature, Industrial & Engineering Chemistry Research 2009, 48 (2), 888893. [141] Keshav A., Wasewar K.L., Chand S., Reactive extraction of propionic acid using trinoctylamine, tri-n-butyl phosphate and Aliquat 336 in sunflower oil as diluent, Journal of Chemical technology and Biotechnology. Published Online: 14 Oct (2008).
182
[142] Keshav A., Chand S., Wasewar K., Recovery of propionic acid by reactive extraction using quaternary amine (Aliquat 336) in various diluents. Chemical Engineering Journal. 2009, 152, 95-102. [143] Keshav A., Wasewar K.L., Chand S., Uslu H., Inci I., Thermodynamics of reactive extraction of propionic acid, i-managers Journal on Future Engineering and Technology. 2009, 4(2), 41- 49. [144] Keshav A., Wasewar K., Chand S., Extraction of propionic acid from model fermentation broth: Effect of pH, salts, substrate, and temperature, AIChE Journal, 2009, 55(7), 1705-1711. [145] ICISpricing report. See also: http://www.icispricing.com [146] Johnson DT, Taconi KA. The glycerin glut: options for the value-added conversion of crude glycerol resulting from biodiesel production. Environ Prog 2007;26(4):338-348. [147] McAloon, A., et al., Technical Report NREL/TP-580-28893. 2000, National Renewable Energy Laboratory: Golden, CO (USA). p. 35 [148] Mitsubishi Gas Chemical Company, Inc; Tokyo, Japan. Annual Report 1999. [149] Shen L, Haufe J, Patel MK. Product overview and market projection of emerging bio-based plastics PRO-BIP 2009: Final report. Utrecht University. Commissioned by European Polysaccharide Network of Excellence (EPNOE) and European Bioplastics; 2009 p. 243.
9. Experimental Setup for Glycerol Fermentation to

PHB
Polyhydroxyalkanoates (PHAs) have been recognized as good substitutes for the nonbiodegradable petrochemically produced polymers. However, their high current production cost limits their industrial applications. Carbon source can be represent between 25% to 45% of the total PHB production costs. Glycerol, a by-product from the biodiesel industry, can be used as primary carbon source for cell growth and PHB synthesis and it is an interesting alternative for to increase the PHB feasibility economic process. Currently, PHB is produced in an industrial scale using Gram negative bacteria. Nevertheless, Gram-negative organisms contain lipopolysaccharides (LPS) which copurify with the PHAs and induces a strong immunogenic reaction. Gram-positive bacteria lack LPS and are hence potentially better sources of PHAs when used for biomdical purposes. In this work, the conditions and capability of poly (hydroxybutyrate) (PHB) production by a Bacillus megaterium (Gram Positive bacteria isolated from superficial sediments of Baha Blanca Estuary (Buenos Aires, Argentina)) using glycerol as only carbon source were studied. This microorganism was adapted and tested at different initial glycerol concentrations and compared with other substrates as glucose. Aspen Plus and Aspen Icarus were used for the processes simulation and for the economic assessment, respectively.
9.1 Generalities
Polyhydroxyalcanoates are attractive substitute biopolymers for conventional petrochemical plastics which have similar physical properties to thermoplastics and elastomers. PHAs are homo or heteropolyesters and can be synthesized and stored intracellularly by many bacterias in the form of granules and can account for up to 80% of the total bacterial dry weight [1] [2]. They can be produced from renewable resources through a fermentation process under restricted growth conditions for nitrogen,
184
phosphorus, sulfurs and/or oxygen in the presence of an excess carbon source [3]; PHAs are also completely biodegraded and biocompatible [1], [2], [3], [4].
Polyhydorxybutyrate (PHB) was the first type of PHAs discovered and the most widely studied. PHB has similar mechanical properties to conventional plastics like
polypropylene or polyethylene, but its production cost are higher than the petrochemical plastic. The PHB production cost depends on the microorganism (yield and productivity), carbon and nitrogen sources (substrates), fermentation conditions (temperature, aeration), recovery and purification of the PHB. Carbon source could represent between 25 to 45 % of the total production costs [4][5].
Many researches have been developed in order to find cheaper carbon sources. Agroindustrial wastes are attractive candidates because they have desired characteristics: such as: low prices and high availability. Moreover, when these wastes are used an environmental problem is avoided.
There are different kinds of microorganisms able to produce PHB from diverse agroindustrial wastes. Substrates such as whey, lignocellulosic materials and glycerol from biodiesel have been studied (Table 9.1).
Glycerol is by-generated during the biodiesel production. With every 100 lbs of biodiesel produced by the transesterification of vegetable oils or animal fats, 10 lbs of crude glycerol are generated (10 wt %) [16]. Although pure glycerol is an important industrial feedstock used in foods, drugs, cosmetics, pharmaceuticals, pulp and paper, leather, textile and tobacco industries, the growth on biodiesel industry has carried out a glycerol surplus and its consequent price has decreased. Thus, the economy of biodiesel industry has been directly affected.
Chemical and biological transformations have been analyzed in order to convert glycerol to added-valuable products. Biological conversion offers the opportunity to synthesize a large array of products and functionalities. Glycerol can be used such carbon source in microbiological process substituting sugars owing to the highly reduced nature of carbon atoms. Different works have shown that some strains of microorganisms can produce PHB using crude glycerol as carbon and energy source. This microorganisms can be wild
9. Experimental Setup
185
strains such as Cupriavidus necator [12], Methylobacterium rhodesianum [13] or recombinant microorganism such as E. coli recombinant [15-16]. In the group of PHB producer bacteria there are both Gram positive and Gram negative strains. Currently, PHB is produced in an industrial scale using Gram negative bacteria. Nevertheless, Gram-negative organisms contain lipopolysaccharides (LPS) which copurify with the PHAs [17] and induces a strong immunogenic reaction. Therefore it is undesirable for biomedical applications. Those purification processes increase the PHB production costs. Gram-positive bacteria lack LPS and are hence potentially better sources of PHAs when used for biomedical purposes [18].
Table 9.1. Some microorganisms PHB producer from different agroindustrial wastes. Agroindustrial Waste Microorganism Methylobacterium sp. ZP24 Pseudomonas hydrogenovora Thermus thermophilus HB8 Recombinant Escherichia coli Burkholderia sacchari IPT 101 Burkholderia cepacia IPT 048 Ralstonia eutropha B. cepacia ATCC 17759 Cupriavidus necator DSM 545 Methylobacterium rhodesienum MB 126 Osmophilic organism Escherichia coli CT1061 Productivity (g PHB l-1h-1) 1.18 0.18 Reference [5] [6] [7] 0.90 0.11 0.09 [8] [9] [9] [10] 0.47 1.51 0.26 0.05 0.18 [11] [12] [13] [14] [15]
Whey
Lignocellusic
Cane bagasse
Xilosa + Glucose
Glycerol
Bacillus megaterium is Gram positive, strict aerobic, non motile, rod shaped, spore forming, citrate positive, bacteria hydrolyzing gelatin and casein [19]. We continue the study of B. megaterium isolated from superficial sediments of Bahia Blanca Estuary
186
(Buenos Aires, Argentina) after to check in other work [20] that this microorganism is a PHB producer from glucose. In this paper, we report the total adaptation of B. megaterium to glycerol as carbon source.
9.2 Materials and Methods

9.2.1 Bacterial strain and its maintenance
The Bacillus megaterium was isolated from superficial sediments of Bahia Blanca Estuary (Buenos Aires, Argentina) and characterized as a PHB producer in the presence of excess carbon source with the restriction of nitrogen. This strain was used for the current study. Stock cultures were grown at 33 C in nutrient broth and maintained at 4C after growth on nutrient agar during the activation level. After adaptation to glycerol as sole carbon source, the stock culture was maintained at 4C after growth on formulated agar with glycerol. B. megaterium cells were stored at -80 C in 2 ml cryovials containing 300 l of glycerol and 700 l of a previously prepared growth liquid culture.
9.2.2 Culture medium

The seeding medium was prepared with the following concentrations: (NH 4 ) 2 SO 4 , 1g/l; KH 2 PO 4 , 1.5 g/l; Na 2 HPO 4 , 9 g/l; MgSO 4 7H 2 O, 0.2g/l; and 1 ml of trace element solutions composed by: FeSO 4 7H 2 O, 10 g/l; ZnSO 4 7H 2 O, 2.25 g/l; CuSO 4 5 H 2 O, 1g/l; MnSO 4 4H 2 O, 0.5; CaCl 2 2H 2 O, 2 g/l; H 3 BO 4 , 0.23; (NH 4 ) 2 Mo 7 O 24 , 0.2 g/l; and HCl, 10ml. The carbon sources analyzed are both glycerol and glucose. The carbon source and MgSO 4 .7H 2 O were autoclaved separately and added aseptically to the medium after cooling.
9.2.3 Microorganisms adaptation and culture conditions

B. megaterium cells were used to inoculate nutrient agar plates supplemented with 20 g/l of glycerol previously incubated at 30C until growth and then stored at 4C. Single colonies from the grown nutrient agar plates were inoculated in 100ml erlenmeyer flasks containing 10 ml of sterile nutrient broth medium supplemented with 20 g/l of glycerol and incubated at 30C in an orbital shaker at 200 rpm during a period of 2-3 days. 1 ml of these cultures was again transferred to 10 ml of seeding medium with glycerol. Successive subculturing was performed 2-3 times, to assure a good adaptation of cell
187
growth on glycerol as carbon source. The next step was the seed of the culture in the formulated medium. 1 ml of culture growth in nutrient broth supplemented with 20 g/l of glycerol was transferred to 10 ml of formulated medium with 20 g/l of glycerol. Successive subculturing was performed 2-3 times, to assure a good adaptation of cell growth formulated medium and glycerol. Those tests were incubated at 30C in an orbital shaker at 200 rpm during a period of 2-3 days.
9.2.4 Batch cultivations

The PHB fermentation was carried out for 36 h in a 3.7 liters Lab Fermenter (Bioengineering, Switzerland) by using a grown formulated medium. Two different carbon sources are studied: glucose and glycerol. For both, the formulated medium and fermenter are the same. The fermentations with glycerol were carried out at 30C and 33C and 200rpm. Two different initial glycerol concentrations are used 20 and 50 g/l. The fermentation with glucose was carried out at 33C and 200 rpm. The initial glucose concentration was of 20g/l. The culture volume was of 1.5 l for all batch fermentations.
9.3 Analytical Methods

9.3.1 Biomass
The biomass of the culture was determined using the optical density measurement at 600nm (Spectronicspectrophotometer, Thermo SCIENTIFIC gravimetric. GENESYS 20) and by
9.3.2 PHB extraction

After fermentation, the cells were harvested by centrifugation at 18C and 6.000 rpm for 20 min and then the intracellular PHB was extracted by using the Chloroform hypochlorite dispersion extraction. The dispersion media contains 50ml of chloroform and 50ml of a diluted (30 wt %) sodium hypochlorite solution in water, in an orbital shaker at 100 rpm. The cell powder was treated at 38C for 1 h. The mixture obtained was then centrifuged at 4000 rpm for 10 min, which resulted in three separate phases. PHB was recovered from the bottom phase that contains PHB dissolved in chloroform. PHB is precipitated using 10 volumes of ice-cold methanol [21].
188
9.3.3 PHB quantification

Dried biomass is used for methanolysis of monomers according to the method described by Braunegg et al. [22] and modified by Lageveen al. [23]. Approximately 10 mg of cells mass was reacted in a small screw-cap test tube with a solution containing 1 ml of chloroform, 0.85 mL of methanol, and 0.15 mL of sulfuric acid for 140 min at 100 C. After reaction, 0.5 mL of distilled water was added and the test tube was shaken vigorously for 1 min. After phase separation, the organic phase (bottom layer) was removed and transferred to a small screw-cap glass vial. 50 l from this organic phase were taken and added to a test tube and injected in the GC-MS. And an Agilent Technologies 6850 series II gas chromatograph was used. The gas chromatograph was equipped with a HP-5MS capillary column (25 m length, 0.32 mm internal diameter). Helium (5cm/min) was used as the carrier gas. Injector and detector were operated at 230 C and 275 C, respectively. A temperature program was used for efficient separation of the esters (120 C for 5min, temperature ramp of 8 C per min, 180 C during 12 min). An Agilent Technologies 5975B mass spectrometer was used.
9.3.4 Glycerol quantification

Glycerol concentration was determined off-line by HPLC (Hitachi LaChrom Elite) equipped with an auto sampler (Hitachi LaChrom Elite L-2200), a Bio-Rad Aminex Fermentation Monitoring Column (150 mm x 7.8 mm), a column oven (Hitachi LaChrom Elite L-2300), a HPLC pump (Hitachi LaChrom Elite L-2130) and a Hitachi LaChrom Elite L-2490 refraction index detector. Injection volume was 20 ml and elution was achieved using a 50 mM solution of H 2 SO 4 . The column was kept at 65C and the pump was operated at a flow rate of 0.8 ml min-1.
9.3.5 PHB characterization

Fourier transform infrared spectroscopy (FTIR): Freeze-dried, precipitated PHB from B. cereus SPV was used to prepare KBr discs (sample:KBr, 1:100). An FTIR spectrum 1720X spectrometer (Perkin Elmer, USA) was used under the following conditions: spectral range, 4000400 cm1; window material, CsI; 16 scans; resolution 4 cm1; the detector was a temperature-stabilized, coated FRDTGS detector [24].
189
9.4 Results and discussion

Glycerol was used as only carbon source and we studied the productivity and yield of the microorganism. The fermentation profile of glycerol to PHB and biomass, in formulated medium by B. megaterium are shown in Figures 9.1 and Figure 9.2. Each one of fermentation was carried out with different initial glycerol concentrations (20g/l and 50 g/l, respectively). Both fermentations were carried out at 30C and 200 rpm, in the 3.7 liters Lab Fermenter (Bioengineering, Switzerland). The culture volume was of 1.5 l. Final concentration of PHB was 1.116 g/l with a productivity of 0.0248 g/l*h when glycerol at 20 g/l was used, while the final concentration of PHB was 2.356 g/l with a productivity of 0.7365 g/l*h when glycerol at 50 g/l was used. concentration is the 50 g/l. Therefore, the best initial glycerol
5 4.5 4 3.5 3
g/l
2.5 2 1.5 1 0.5 0 0 5 10 15 20 25 30 35 40 45 50
Time (hours)
Figure 9.1. Accumulation profile I of Bacillus megaterium cultivated in the 3.7 liters Lab Fermenter (Bioengineering, Switzerland) with a culture volume of 1.5 l, initial glycerol concentration of 20 g/l at 30C and 200 rpm. Biomass, PHB.
Then, the influence of temperature on the fermentation process was analyzed. The temperature range for the growth of Bacillus was found to be from 25 to 45C [17]. The temperatures analyzed in this work were 30C and 33C. Both fermentations were carried out at 200 rpm, in the 3.7 liters Lab Fermenter (Bioengineering, Switzerland). The culture volume was of 1.5 l, at the same initial glycerol concentration of 20g/l (Figures 9.1 and Figures 9.3). Final PHB concentration was 1.116 g/l with a productivity of 0.0248 g/l*h at 30 C, while the final concentration of PHB was 3.4 g/l with a productivity of 0.0771 g/l*h at 33 C. Therefore, the best temperature operation is the 33C.
190
10 9 8 7 6
g/l
5 4 3 2 1 0 0 5 10 15 20 25 30 35 40
Time (hours)
Figure 9.2. Accumulation profile II of Bacillus megaterium cultivated in the 3.7 liters Lab Fermenter (Bioengineering, Switzerland) with a culture volume of 1.5 l, initial glycerol concentration of 50 g/l at 30C and 200 rpm. Biomass, PHB.
6 5 4
g/l
3 2 1 0 0 5 10 15 20 25 30 35 40 45 50
Time (hours))
Figure 9.3. Accumulation profile III of Bacillus megaterium cultivated in the 3.7 liters Lab Fermenter (Bioengineering, Switzerland) with a culture volume of 1.5 l, initial glycerol concentration of 20 g/l at 33C and 200 rpm. Biomass, PHB.
In the same form, glucose was used as only carbon source in order to compare the use of glycerol as raw material. The initial glucose concentration used was of 20g/l at 33C (Figure 9.4). Final PHB concentration was 2.83 g/l with a productivity of 0.0783 g/l*h. A comparison between obtained results for PHB production by mean of B. megaterium using glucose and glycerol as substrates was made as shown in Table 9.2.
191
25 20 15
g/l
10 5 0 0 5 10 15 20 25 30 35 40
Time (hours)
Figure 9.4. Accumulation profile IV of Bacillus megaterium cultivated in the 3.7 liters Lab Fermenter (Bioengineering, Switzerland) with a culture volume of 1.5 l, initial glucose concentration of 20 g/l at 33C and 200 rpm. Biomass, Glucose, PHB
Table 9.2. Comparison between experimental results for glucose and glycerol fermentation to PHB GLUCOSE Variable 20g/lT=33C DCW(g/l) 5.08g/l 20g/lT=33C 5.7g/l GLYCEROL 50g/lT=30C 7.8 g/l 20g/lT=30C 4.7 g/l
PHB(g/l) % Accumulation
2.83 g/l 55.7%
3.4g/l 62.3%
2.356 g/l 30.2%.
1.116g/l 23.74%.
Time of Max. PHB production (hours) Productivity gPHB/l*h Total substrate consumption (g /l) Yield P/S
36
42
32
45
0.0786
0.0771
0.07365
0.0248
10.5
11.80
15.50
0.30
0.29
0.07
The produced PHB was characterized. The peak got form the FT-IR spectrum (1728 cm1
) corroborates with the peak reported, indicated the presence of PHB using glycerol as
sole carbon resource.
192
References
[1] Khanna S, Srivastava AK. Recent advances in microbial polyhydroxyalkanoates. Process Biochemistry 2005;40:607619. [2] Mahishi LH, Tripathi G, Rawal SK. Poly(3-hydroxybutyrate) (PHB) synthesis by recombinant Escherichia coli harbouring Streptomyces aureofaciens PHB biosynthesis genes: Effect of various carbon and nitrogen sources. Microbiol Res 2003;158:1927. [3] Steinbchel A. Perspectives for Biotechnological Production and Utilization of Biopolymers: Metabolic Engineering of Polyhydroxyalkanoate Biosynthesis Pathways as a Successful Example. Macromolecular Bioscience 2001;1:1-24. [4] Lee SY. Plastic bacteria? Progress and prospects for polyhydroxyalkanoate production in bacteria. Tibtech 1996; (VOL 14). [5]. A. Nath, M. Dixit, A. Bandiya, S. Chavda, A.J. Desai. Enhanced PHB production and scale up studies using cheese whey in fed batch culture of Methylobacterium sp. ZP24. Bioresource Technology 99 (2008); 57495755 [6]. Martin Koller, Rodolfo Bona, Emo Chiellini, Elizabeth Grillo Fernandes, Predrag Horvat, Christoph Kutschera, Paula Hesse, Gerhart Braunegg. Polyhydroxyalkanoate production from whey by Pseudomonas hydrogenovora. Bioresource Technology 99 (2008); 48544863 [7]. Pantazaki AA, et al. Production of polyhydroxyalkanoates from whey by Thermus thermophilus HB8.Process Biochem (2009), doi:10.1016/j.procbio.2009.04.002 [8]. Beom Soo Kim. Production of poly(3-hydroxybutyrate) from inexpensive substrates. Enzyme and Microbial Technology 27 (2000) 774777. [9]. L. F. Silva, M. K. Taciro, M. E. Michelin Ramos, J. M. Carter, J. G. C. Pradella, J. G. C. Gomez. Poly-3-hydroxybutyrate (P3HB) production by bacteria from xylose, glucose and sugarcane bagasse hydrolysate. J Ind Microbiol Biotechnol (2004) 31: 245254 [10]. Jian Yu, Heiko Stahl. Microbial utilization and biopolyester synthesis of bagasse hydrolysates. Bioresource Technology 99 (2008) 80428048.
193
[11]. L. F. Silva, M. K. Taciro, M. E. Michelin Ramos, J. M. Carter, J. G. C. Pradella, J. G. C. Gomez. Poly-3-hydroxybutyrate (P3HB) production by bacteria from xylose, glucose and sugarcane bagasse hydrolysate. J Ind Microbiol Biotechnol (2004) 31: 245254 [12]. Cavalheiro JMBT, et al. Poly(3-hydroxybutyrate) production by Cupriavidus necator using waste glycerol. Process Biochem (2009), [13]. Bormann EJ, Roth M. Production of polyhydroxybutyrate by Methylobacterium rhodesianum and Ralstonia eutropha in media containing glycerol and casein hydrolysates. Biotechnol Lett 1999;21:105963. [14]. Koller M, Bona R, Braunegg G, Hermann C, Horvat P, Kroutil M, et al. Production of polyhydroxyalkanoates from agricultural waste and surplus materials. Biomacromolecules 2005;6:5615 [15]. Pablo I. Nikel, M. Julia Pettinari, Miguel A. Galvagno, Beatriz S. Mndez. Poly(3hydroxybutyrate) synthesis from glycerol by a recombinant Escherichia coli arcA mutant in fed-batch microaerobic cultures. Appl Microbiol Biotechnol (2008) 77:13371343. [16] Syed Shams Yazdani and Ramon Gonzalez. Anaerobic fermentation of glycerol: a path to economic viability for the biofuels industry. Current Opinion in Biotechnology 2007, 18:213219 [17]. S.P. Valappil, S.K. Misra, A.R. Boccaccini, T. Keshavarz, C. Bucke, I. Roya. Largescale production and efficient recovery of PHB with desirable material properties, from the newly characterized Bacillus cereus SPV. Journal of Biotechnology 132 (2007), 251258 [18] Chen, G.Q., Wu, Q. The application of polyhydroxyalkanoates as tissue engineering materials. Biomaterials 26 (2005), 65656578. [19]. Sureshbabu K. P., Venkatraman D., Kalimuthu., Neelamegam R., Muniyandi J., Sangiliyandi G. Optimization and fed-batch production of PHB utilizing dairy waste and sea water as nutrient sources by Bacillus megaterium SRKP-3. Bioresource Technology 101 (2010), 705711. [20]. Naranjo, J. M., Vasquez, J. A., Higuita, J.C., Cardona, C.A. PHB production and characterization from a Bacillus megaterium strain isolated from Marine Sediments.
194
[21]. Hahn, S.K., Chang, Y.K., Lee, S.Y. Recovery and characterization of poly(3hydroxybutyric acid) synthesized in Alcaligens eutrophus and recombinant Escherichia coli. Appl. Environ. Microbiol. 61 (1995), 3439. [22] Braunegg, G. Sonnleitner, B., Lafferty, R.M. A rapid gas chromatographic method for the determination of poly-B-hydroxybutyric acid in microbial biomass. Eur. J. Appl. Microbiol. Biotechnol. 6 (1978), 29-37. [23]. Lageveen, R. G., Huisman, G. W., Preusting, H., Ketelaar, P., Eggink, G. & Witholt, B. Formation of polyesters by Pseudomonas oleovorans : eect of substrates on formation and composition of poly-(R)-3-hydroxyalkanoates and poly-(R)-3-
hydroxyalkenoates. Appl Environ Microbiol 54 (1988), 29242932. [24]. S.P. Valappil, S.K. Misra, A.R. Boccaccini, T. Keshavarz, C. Bucke, I. Roy. Largescale production and efficient recovery of PHB with desirable material properties, from the newly characterized Bacillus cereus SPV. Journal of Biotechnology 132 (2007) 251258
10. Conclusions
Commercially three qualities of glycerol were identified as the most important ones. Crude glycerol with a purity ranging from 80-88 wt %, technical glycerol mainly found at 97 wt %, and refined glycerol (USP or FCC grades) at 99.7 wt %. These three types of glycerol differ significantly in the content of water, fatty acid residues, esters, and other organic wastes. Also, some differences were found for the use of diverse feedstocks for biodiesel production on the composition of the glycerol layer. Although, most of the first use oils lead to not big differences in the glycerol layer, a completely different behavior was observed for the glycerol obtained from WVO represented by low concentration of glycerol and methanol with a high content of fats. On the other hand, based on the traditional purification of glycerol, a flowsheet able to purify raw glycerol up to the three commercial qualities above described was designed, simulated and economically assessed. Results showed that not only quality requirements were successfully obtained but also for the analyzed purification scale all the processes were profitable. Thus, a homogenized raw material and a purification process were obtained in order to continue the analysis of different possibilities of glycerol transformation to added-value products.
Acrolein, hydrogen, and 1,2-propanediol, are three of the most commercially important products obtained from glycerol, due to their applications, established market, and sale prices. Here the technological schemes to produce these compounds were designed, simulated, and economically assessed. Thus, simulation results showed that all the processes are technologically feasible reaching high purity of product. Also, acrolein production was found to be viable at a purity of 92 wt %, but do not at a purity of 98.5 wt %. Finally, both hydrogen and 1,2-propanediol production processes are also economically viable, where the last one generates the highest profit margin.
In the past, important efforts have been made to introduce the biotechnological production of 1,3-propanediol from glycerol to the industry. However, research tendencies were
196
focused on microorganism development and some process conditions analysis. The drastic increment in the use of biodiesel caused an oversupply of glycerol in the market. For this reason and in order to optimize its productivity, mass production of 1,3propanediol from glycerol needs additional analysis. Here, fermentation of glycerol with K. pneumoniae was optimized using three different models and considering two simultaneous goals: i) high volumetric productivity and ii) high 1,3-propanediol concentration. In conclusion, the obtained results provide enough information to understand the different possibilities for process intensification using this technology and also to compare it with other new industrial alternatives for the utilization of glycerol as a raw material.
Due to the low cost of raw glycerol, methanol recovery from glycerol implies low PCs. Meanwhile, the three possibilities assessed for glycerol bioconversion showed that the GPCs of fuel ethanol from raw glycerol are lower than the commercial price of fuel ethanol. These facts show the potential for raw glycerol bioconversion to fuel ethanol using E. coli. Also, the comparison carried out with a previous paper (which considers the fuel ethanol production from sugarcane and corn in the Colombian case, [13]) shows that the GPCs of fuel ethanol from raw glycerol can be as profitable as the production of fuel ethanol from conventional raw materials as sugarcane. The latter is a completely developed industry in Colombia.
Biodiesel and ethanol can be jointly produced using oil palm as sole source by mean of processes integration, such as the biodiesel production with the ethanol production from two feedstocks: lignocellulosic residues (empty fruit bunches and palm press fiber produced during) and crude glycerol. Thus, alcohol is completely self-supplied by the integrated process and low quantities of wastes are produced without any global production of glycerol. Economical evaluation showed a higher biodiesel production cost for the integrated process than the traditional biodiesel production process which uses ethanol and palm oil as feedstocks. But the first one is a promising technology available to build an autonomous biodiesel production plant with low waste levels. This process must be economically improved by further analysis from a process design view point, based on process simulation which showed be a powerful tool for performing processes integration.
10. Conclusions
197
Three technological schemes to produce PHB from crude glycerol were analyzed under two fermentation conditions (i.e., using glycerol at 88 wt % and 98 wt %). In this work it was found that it is better to use pure glycerol as feedstock for the production of PHB than raw glycerol. This phenomenon is explained by the fact that the higher PHB yield reduces the utility costs in the downstream process. The results shown here are important for the industrial production of PHB using glycerol as a raw material. Currently, in the biodiesel production the total profitability of any new project could be determined by the right use of glycerol as a massive by-product. The proposed strategy to use pure glycerol as substrate can be understood as a very interesting alternative since the final composition in the glycerol streams depends on the source of the feedstock used for biodiesel production. Also, most of the biotechnological alternatives to produce added value compounds from glycerol are sensitive to contaminants in the raw material. Thus, several technical and economical advantages as well as a more stable production of PHB are obtained when a standardized raw material as pure glycerol is used.
Usage of raw glycerol, engineered Escherichia coli strains, and processes integration for the production of optically pure D-lactic acid is an important alternative to transform the by-produced glycerol during the biodiesel synthesis. Although five different configurations for the fermentation stage were considered, in all cases the total production costs were lower than its sale price. Thus, the whole process scheme for D-lactic acid production could be considered as potentially profitable design. Also, it was found that the combined effect of both high glycerol concentration and use of low quality glycerol in the fermentation media, lead to the best economic performance. The results shown here are important for the industrial production of D-lactic acid using glycerol as a raw material. On the other hand, in the biodiesel production the total profitability of any new project could be determined by the right use of glycerol as a massive raw material.
11. List of Publications and Submitted Papers

This chapter shows the published results throughout scientific meeting, papers, book chapters, invited book chapters, and books. Also, a list containing the submitted papers was made.
11.1 Published
Scientific Meetings
1. Posada JA, Higuita JC, Cardona CA. 2011. Optimal crude glycerol biorefinery from biodiesel production to produce poly-3-hydroxybutyrate. World
Renewable Energy Congress 2011 Sweden. Linkping University.
2. Posada JA, Quintero JA, Cardona CA. 2010. Energy and environmental comparison among the production of fuel ethanol from crude glycerol, sugar cane, and crop. IV International congress of science and technology of the biofuels. Bucaramanga, Colombia. November 30th - December 3rd.
3. Posada JA, Cardona CA, Rincn LE. 2010. Sustainable biodiesel production from palm using in situ produced glycerol and biomass for raw bioethanol. In: Society for Industrial Microbiology. 32nd symposium on biotechnology for fuels and chemicals. Clearwater Beach, Florida. Abril 19th-22nd.
200 Papers
4. Posada JA, Naranjo JM, Lpez JA, Higuita JC, Cardona CA. 2011. Design and Analysis of PHB Production Processes from Crude Glycerol. Process Biochemistry. 46:310-317.
5. Posada JA, Cardona CA. 2010. Design and analysis of fuel ethanol production from raw glycerol. Energy. 35(12):5286-5293.
6. Posada JA, Cardona CA. 2010. Anlisis de la refinacin de glicerina obtenida como co-producto en la produccin de biodiesel (Validation of glycerin refining obtained as a by-Product of biodiesel production). Ingeniera y Universidad 14:2-27.
7. Posada
JA,
Orrego
CE;
Cardona
CA.
2009.
Biodiesel
production:
Biotechnological approach. International Review of Chemical Engineering (I.Re.Che.), 1(6):571-580.
Book Chapters
8. Posada JA, Cardona CA, Cetina DM, Orrego CE. 2009. Bioglicerol como materia prima para la obtencin de productos de valor agregado (Bioglycerol as raw material to obtain added value products). En: CARDONA CA (ed). Avances investigativos en la produccin de Biocombustibles (Reasearching advances for biofuels production). Manizales: Artes Graficas Tizan. p. 103127. ISBN: 978-958-44-5261-0
11. List of Publicatiions
201
Books
9. Cardona CA, Posada JA, Quintero JA. 2010. Aprovechamiento de subproductos y residuos agroindustriales: Glicerina y Lignocelulsicos (Use of agroindustrial wastes and by-products: Glycerin and Lignocellulosics). Manizales: Artes Graficas Tizan. p. 218. ISBN: 978-958-44-7611-1
Invited Book Chapters
10. Posada JA, Rincn LE, Cardona CA. Integral Use of Palm Oil: Production of Biodiesel and Added Value Compounds from Glycerin. In: Oil Palm: Cultivation, Production and Dietary Components. Editor: Susan A. Penna. Book Chapter Ed. Nova Publisher. Series: Agriculture Issues and Policies. ISBN: 978-1-61122-201-2.
11.2 Submitted
Papers
1. Posada JA, Jaramillo JJ, Cardona CA. Glycerol fermentation to 1,3propanediol: Comparison among four culture configurations. Submitted to Biochemical Engineering Journal.
2. Posada JA, Higuita JC, Pisarenko YA, Cardona CA. Design and economical analysis of the technological scheme for 1,3-propanediol production from raw glycerol. Submitted to Theoretical Foundations of Chemical Engineering.
202
Glycerol Conversion to Added Value Products 3. Posada JA, Cardona CA. Comparison among three chemical processes for glycerol conversion: Acroleine, Hydrogen and 1,2-Propanediol. Submitted to Bioresource Technology.
4. Posada JA, Rincn LE, Cardona CA. Sustainable biodiesel production from oil palm: crude glycerol and biomass conversion to raw-bioethanol. Submitted to Applied Biochemistry and Biotechnology.
5. Posada JA, Quintero JA, Cardona CA. Comparacin econmica y ambiental entre la produccin de etanol a partir caa de azcar, maz y glicerol crudo. Submitted to Revista de Ingeniera Qumica-UIS.
6. Posada JA, Quintero JA, Cardona CA. Comparison among three technologies for biodiesel production from Jatropha seeds. Under review by the advisor.
7. Posada JA, Cardona CA. Possibilities of glycerol conversion as a sole raw material: a review. Under review by the advisor.
8. Posada JA, Cardona CA, Gonzalez R. Design, simulation, and economic assessment of D-lactic acid production process from raw glycerol using engineered Escherichia coli strains. Under review by the advisor.
9. Posada JA, Naranjo JM, Lpez JA, Higuita JC, Cardona CA. Poly(3hydroxybutyrate) production by Bacillus megaterium using glycerol as substrate: experimentation and process simulation. Under review by the advisor.
10. Posada JA, Cardona CA. Design, simulation, and economic assessment of succinic acid production process from raw glycerol. Under review by the advisor.
11. List of Publicatiions
203
11. Posada JA, Cardona CA. Design, simulation, and economic assessment of propionic acid production process from raw glycerol. Under review by the advisor.

Analisis Comparativo de Usos de Glicerol

Încărcat de

Informații document

Descriere originală:

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Analisis Comparativo de Usos de Glicerol

Încărcat de

Drepturi de autor:

Formate disponibile

Design and Analysis of Technological Schemes for Glycerol Conversion to Added Value Products

John Alexander Posada Duque

John Alexander Posada Duque

John Alexander Posada Duque

Advisor: Ph.D., M.Sc, Chemical Engineer Carlos Ariel Cardona Alzate

______________________ A mi madre y mi hermana por su apoyo, a Patricia por su valiosa presencia.

Glycerol Conversion to Added Value Products

assessment; glycerol-based biorefineries.

85 85 104 115 124 138 148 153 167 169

10. Chapter 10: Conclusions

11. Chapter 11: List of Publications

104 106 113

Glycerol Conversion to Added Value Products

137 141 141

Glycerol Conversion to Added Value Products

1.1 Application field and motivation

Figure 1.1. World biodiesel production and capacity.

Glycerol Conversion to Added Value Products

Figure 1.2. Global biodiesel production by feedstock

1.2 Thesis objectives

1.3 Thesis structure

Chapter 2: The glycerols world

Chapter 3: Methodology to design and analyze processes based on simulation tools.

Glycerol Conversion to Added Value Products

Chapter 4: Separation and purification of glycerol

Chapter 5: Chemical conversion of glycerol

Chapter 6: Biochemical conversion of glycerol

Additionally, conversion levels, yields, selectivities, and productivities are presented.

Chapter 7: Cases of study for chemical conversion of glycerol

Chapter 8: Cases of study for biochemical conversion of glycerol

Chapter 9: Experimental setup

Chapter 10: Conclusions

Chapter 11: List of publications and submitted papers

2. The Glycerols World

Glycerol Conversion to Added Value Products

2.2 Biodiesel industry

2. The Glycerols World

Glycerol Conversion to Added Value Products

2.3 Glycerol market and its oversupply problem

2. The Glycerols World

2.4 Glycerol as raw material

Glycerol Conversion to Added Value Products

2. The Glycerols World

Glycerol Conversion to Added Value Products

3. Methodology for Processes Design and Analysis

3.1 Processes design

Glycerol Conversion to Added Value Products

Figure 3.1. Hierarchical decomposition according to the "onion diagram"

Glycerol Conversion to Added Value Products

3.2 Processes simulation

Glycerol Conversion to Added Value Products

3.3 Processes assessment

Glycerol Conversion to Added Value Products

4. Separation and Purification of Glycerol

4.1 Commercial qualities of glycerol

Glycerol Conversion to Added Value Products

5 ppm Mx. 30 ppm Mx.

100 ppm Mx. 1000 Mx

12.0% Max 4.0 - 9.0 2.0% Mx

5.0% Mx 4.0 - 9.1 2.0% Mx

0.5% Mx N/A N/A

4.2 Effect of the feedstock for biodiesel production on glycerol composition