ADDIS ABABA UNIVERSITY

SCHOOL OF GRADUATE STUDIES

Addis Ababa Institute of Technology School of
Chemical and Bio-Engineering
Optimization of Molasses Fermentation to Ethanol to Reduce
Effluent Generation: The Case of Metahara Sugar Factory

By
Bekele Mirkena
A thesis submitted to Addis Ababa University AAiT School of Chemical
and Bio Engineering in Partial Fulfillment of the Requirements for the
Degree of Master of Science in Chemical Engineering (Food Engineering)

Advisor : Dr.Ing. Belay Woldeyes (Associate Professor)

December 2014
 Addis Ababa University
School of Graduate Studies

Addis Ababa Institute of Technology

School of Chemical and Bio engineering

Process Optimization of Cane molasses Fermentation to Reduce Effluent Generation

The Case of Metahara Sugar Factory Distillery Plant

By
Bekele Mirkena

Approved by the Examining Board :

_________________________________ _____________________
Chairman of School Graduate Committee
Dr.Ing. Belay W/yes (Advisor) _____________________
Dr.Ing. Nurelegn Tefera _____________________
( External Examiner)
Ato Teshome Worku _____________________
(Internal Examiner)

i
Acknowledgment
Above all, I praise the LORD Jesus Christ and next, I want to express my sincere gratitude to my
advisor Assoc. Prof. Dr.Ing, Belay Woldeyes for excellent scientific guidance, support in fruitful
ideas and editing the thesis material. I also appreciate all staff of Chemical and Bio Engineering
School of AAiT for their unreserved assistance, sharing of knowledge and for extensive scientific
discussions. All colleagues and friends of the members of food engineering, environmental and
process engineering streams particularly Ato Sendeku Takele and W/t Yemisrach Yisehak are
greatly thanked for the enthusiastic studying atmosphere during the entire stay throughout the
program duration. It is my greatest pleasure to thank Dr.Eng.shimelis Admassu for his unreserved
assistance and sharing of knowledge during my post graduate study.
I would like to thank the management of MSF for giving me the chance to attend my MSc. Program.
It is my greatest pleasure to thank Ato Kassahun Mirkena, MSF Sugar Production Team Leader.
Without his effort, the chance couldn’t have been successful. The Cooperation of MSF distillery
plant Quality control laboratory, for my laboratory experiment, is gratefully acknowledged. Here,
my special thanks go to Ato Mulugeta Eshete and Ato Tesfaye Abdissa MSF Quality Control
supervisors, for their assistance in facilitating all necessary conditions for the Laboratory analysis
and all Lab analyzers and their assistants for their great material and moral support during all the
research time.
My deepest gratefulness goes to my beloved wife W/O Genet Assefa, for her understanding love,
patience and shouldering all the responsibilities. It is her greatest encouragement and sacrifices that
brought this work to a successful end. Finally, I would like to thank my parents specially my mother
W/O Birkea Sakuma, My brother Ermiyas Mirkena and my Sisters Rahel and Elsa Mirkena for their
required support and moral given to me while I followed my post graduate education.

ii
 Table of Content
Chapter Title page
Title Page i
Acknowledgement ii
Table of content iii
Lists of Tables vii
Lists of Figures ix
Lists of Appendices xi
Lists of Acronyms xii
Abstract xiii
1. Introduction 1
1.1. Background 1
1.2. Statements of the Problem 2
1.3. Objectives 3
1.3.1. General Objectives 3
1.3.2. Specific Objectives 3
1.4. Significance of the Study 3
2. Literature Review 5
2.1. Fermentation Process 5
2.1.1. Raw Material Preparation 5
2.1.2. Yeast Culturing and Recycling System 6
2.1.2.1. Media Preparation 6
2.1.2.2. Inoculation 7
2.1.2.3. Yeast Recycling 7
2.1.3. Final Fermentation 7
2.1.3.1. Batch Fermentation 8
2.1.3.2. Continuous Fermentation 9
2.1.4. Chemical Reaction During Fermentation 10
2.1.5. Factor affecting Fermentation of Molasses 10
2.1.5.1. Effect of Sugar Concentration 10
2.1.5.2. Effect of pH 11
2.1.5.3. Effect of Temperature 11
iii
2.1.5.4. Effect of Oxygen 12
2.1.5.5. Effect of Ethanol and Secondary Component Inhibition 12
2.1.6. Effects of Nutrient on Molasses Fermentation 13
2.2. Distillation 13
2.2.1. Primary Column 14
2.2.2. De-aldehyde Column 15
2.2.3. Rectifier Column 15
2.3. Alcohol Dehydration 15
2.4. Over all Water Balance 16
2.5. Distillery Effluent Treatment 17
2.5.1. Recirculation of Distillery Effluent to Fermentor 19
2.5.2. Effects of Re- circulating Effluent on Food Quality and Safety of Sprit 19
3. Materials and Methods 22
3.1. Characterization of Input Materials 22
3.1.1. Materials 22
3.1.2. Methods 23
3.1.2.1. Compositional analysis of molasses 23
3.1.2.2. Physico chemical analysis of process water 24
3.2. Optimization of Nutrients 24
3.2.1. Materials 24
3.2.2. Design for Optimization of Nutrients 24
3.2.3. Methods 25
3.3. Optimization of effluent Recirculation to fermenters 26
3.3.1. Materials 26
3.3.2. Design for Optimizing Recirculation of Distillery Effluent to Fermentors 26
3.3.3. Methods 29
3.4. Effect of Yeast Recycling on Bulk Fermentation Process 29
3.4.1. Materials 29
3.4.2. Methods 30
3.5. Determining the Effects of re circulating distillery effluent on Fermentation
to food Safety and Quality of Rectified Sprit Alcohol 30
3.5.1. Materials 30
iv
3.5.2.Methods 30
4. Results and Discussion 31
4.1. Characterization of Input Material 31
4.1.1.Characterization of Molasses Input 31
4.1.2.Characterization of Process water 32
4.2. Optimization of Nutrients 33
4.2.1.Optimization of Nutrient at 15% Brix (8.67% fermentable Sugar) 33
4.2.1.1.Data Analysis for Optimization of Nutrient at 8.67 % Sugar concentration 34
4.2.1.2. Correlation Matrices for Nutrient Optimization at 8.67 % fermentable
Sugar concentration 36
4.2.1.3.Equation for Nutrient Optimization at 8.67 % sugar concentration 37
4.2.1.4.Response Surface Plot Analysis for Nutrient Optimization
at 8.67 % Fermentable sugar concentration 39
4.2.1.5. Optimization Solution for nutrient Application for Sugar Concentration
of 8.67% fermented wash 41
4.2.2. Optimization of Nutrient at 20% Brix (11.5% fermentable Sugar) 42
4.2.2.1.Data Analysis for Optimization of Nutrient at 11.5 % Sugar concentration 42
4.2.2.2. Correlation Matrices for Nutrient Optimization at 11.5 % fermentable
Sugar concentration 44
4.2.2.3.Regression equation for nutrient optimization at 11.5 % sugar concentration 45
4.2.2.4.Response Surface Plot Analysis for Nutrient Optimization
at 11.5 % Fermentable sugar concentration 46
4.2.2.5. Optimization Solution for nutrient Application for
Sugar Concentration of 11.5 % fermented wash 48
4.2.3.Optimization of Nutrient at 25% Brix (14.3% fermentable Sugar) 49
4.2.3.1.Data Analysis for Optimization of Nutrient at 14.3 % Sugar concentration 49
4.2.3.2.Correlation Matrices for Nutrient Optimization at 14.3 % fermentable
Sugar concentration 51
4.2.3.3.Equation for Nutrient Optimization at 14.3 % sugar concentration 52
4.2.3.4.Response Surface Plot Analysis for Nutrient Optimization
at 14.3 % Fermentable sugar concentration 52

v
4.2.3.5. Optimization Solution for nutrient Application for Sugar Concentration
of 14.3% fermented wash 55
4.3. Optimization of Effluent Recirculation to Fermentors 55
4.3.1.Characterization of Effluent 55
4.3.1.1.Compositional analysis of Spent Wash 56
4.3.1.2.Compositional analysis of process condensate and Spent lees 56
4.3.2.Optimization at 15˚ Brix( 8.67% Sugar Concentration) 57
4.3.2.1.Data Analysis for Optimization of Effluent at 8.67% sugar Concentration 57
4.3.2.2.Correlation Matrices for Effluent Optimization at 8.67% sugar Concentration 60
4.3.2.3.Equation for Effluent Optimization at 8.67% Sugar concentration 60
4.3.2.4.Response Surface Plot Analysis for Effluent Optimization
at 8.67% fermentable sugar concentration 62
4.3.2.5.Optimization Solution for Effluent Recirculation for Sugar concentration
of 8.67% fermented wash 67
4.3.3. Optimization of Effluent at different Fermentable sugar concentration 67
4.3.3.1.Data Analysis for Optimization of Effluent Recirculation at
different sugar Concentration 69
4.3.3.2.Correlation Matrices for Effluent Optimization at different sugar Concentration 71
4.3.3.3.Equation for Effluent Optimization at different Sugar Concentration 72
4.3.3.4.Response Surface Plot Analysis for Effluent Optimization
at different fermentable sugar concentration 74
4.3.3.5.Optimization Solution for Effluent Recirculation at different Sugar concentration
of fermented wash 78
4.4. Effects of Yeast Cream Separator on Bulk Production Process 79
4.5. Effect of Re circulating Effluent to Fermenter on Food Safety 79
4.6. Material and Energy Balance 80
4.6.1.Material and Energy Balance for 15˚Brix Fermented wash 80
4.6.1.1.Solid Balance of the Effluent 81
4.6.1.2.Energy Balance 81
4.6.2.Equipment Design 83
5. Financial Analysis of Effluent Recirculation 85
5.1. Investment Cost Estimation 85
vi
5.2. Operating Cost Estimation 86
5.3. Cost to treat Effluents 86
5.4. Projected Cash flow 87
5.5. Profitability 88
6. Conclusion and Recommendation 89
6.1. Conclusion 90
6.2. Recommendation 90
6.2.1.Recommended future study 91
References 92

vii
 List of Tables
Table 2.1 – type and total quantity of effluent generated 17
Table 2.2- General Characteristics of Distillery Effluents 18
Table 2.3 – Standard Administrative orders (SOS) for different types of distilled Sprits 20
Table 3.1 – Central Composite design for nutrient optimization 25
Table 3.2 – Box Behnken Experimental Design to optimize distillery effluent recirculation
to fermenters at 15ºbrix fermented wash 27
Table 3.3- Box Behnken Experimental Design to optimize distillery effluent recirculation
to fermenters at 15ºbrix fermented wash 28
Table 4.1 – Molasses analysis results 32
Table 4.2 – Analysis results of Process water 33
Table 4.3 - Optimization results of nutrient at 8.67% fermentable sugar (15 º brix)
Fermented wash 33
Table 4.4- ANOVA test for fermented wash Alcohol at 8.67 % fermented wash sugar
Concentration 34
Table 4.5 – Post ANOVA statistics for fermented wash alcohol % to optimize Nutrient 35
Table 4.6- Post ANOVA statistics for fermented wash Cell count to optimize Nutrient 35
Table 4.7 - Post ANOVA statistics for fermented wash Residual sugar to optimize Nutrient 36
Table 4.8- Correlation Matrix of factors[ Pearsons’s R] for nutrient optimization at
8.67% fermentable sugar (15- brix) fermented wash 37
Table 4.9 - Optimization results of nutrient at 11.55% fermentable sugar (20 º brix)
Fermented wash 42
Table 4.10 – Post ANOVA statistics for fermented wash alcohol % to optimize Nutrient 43
Table 4.11- Post ANOVA statistics for fermented wash Cell count to optimize Nutrient 43
Table 4.12-Post ANOVA statistics for fermented wash Residual sugar to optimize Nutrient 44
Table 4.13- Correlation Matrix of factors[ Pearsons’s R] for nutrient optimization at
11.55% fermentable sugar (20- brix) fermented wash 44
Table 4.14 - Optimization results of nutrient at 14.33% fermentable sugar (25º brix)
Fermented wash 49
Table 4.15 – Post ANOVA statistics for fermented wash alcohol % to optimize Nutrient 50
Table 4.16- Post ANOVA statistics for fermented wash Cell count to optimize Nutrient 50
Table 4.17-Post ANOVA statistics for fermented wash Residual sugar to optimize Nutrient 51
viii
Table 4.18- Correlation Matrix of factors[ Pearsons’s R] for nutrient optimization at
14.33% fermentable sugar (25- brix) fermented wash 51
Table 4.19 – Spent wash analysis results 56
Table 4.20 – Process condensate and Spent lees analysis results 56
Table 4.21 - Optimization results of Effluent Re circulation at 8.67% fermentable sugar (15 º
brix) Fermented wash 57
Table 4.22- ANOVA test for fermented wash Alcohol at 8.67 % fermented wash sugar
Concentration 58
Table 4.23 – Post ANOVA statistics for fermented wash alcohol % to optimize effluent
Recirculation to fermentors 59
Table 4.24- Post ANOVA statistics for fermented wash Cell count to optimize effluent
recirculation 59
Table 4.25 - Post ANOVA statistics for fermented wash Residual sugar to optimize effluent
recirculation 60
Table 4.26- Correlation Matrix of factors[ Pearsons’s R] for optimization of effluent
Recirculation at 8.67% fermentable sugar (15- brix) fermented wash 60
Table 4.27 - Optimization results of Effluent Re circulation at different fermentable sugar
concentration of fermented wash 68
Table 4.28- ANOVA test for optimization of effluent re circulation of fermented wash
Alcohol at different fermented wash sugar Concentration 69
Table 4.29 – Post ANOVA statistics for fermented wash alcohol % to optimize effluent
Recirculation to fermentors at different fermentable sugar concentration 70
Table 4.30- Post ANOVA statistics for fermented wash Cell count to optimize effluent
Recirculation at different fermentable sugar concentration 71
Table 4.31 - Post ANOVA statistics for fermented wash Residual sugar to optimize effluent
Recirculation at different fermentable sugar concentration 71
Table 4.32- Correlation Matrix of factors[ Pearsons’s R] for optimization of effluent
Recirculations at different fermentable sugar fermented wash 72
Table 4.33- Comparative analysis test results of alcohol produced with and without
Effluent recirculation 80
Table 5.1- Investment cost to install Effluent re circulation system 85
Table 5.2- Operating cost to run effluent re circulation system 86
ix
Table 5.3- Treatment Cost of Effluent 86
Table 5.4- Projected cash flow of the proposed Effluent re circulation system 87
Table 5.5 – Payback period on investment for proposed Effluent re circulation system 88
Table 5.6- NPV analysis at 12% Discounted cash flow rate 88
Table 7.1- Mass balance on each unit process 95
Table 7.2-7.4-analysis results of effluent recirculation at 15 brix FW run 1-3 96
Table 7.5-7.7- analysis results of effluent recirculation at different FW sugar
Concentration Run 1-3 99
Table 7.8 – Compositional analysis results of molasses 104
Table 7.9 – Compositional analysis results of process water 105
Table 7.10- Compositional analysis results of effluents
Table 7.11 – Color ranges of fusel oil in Rectified Sprit 110

x
List of Figures
Figure 2.1- Total Material Balance of Metahara Sugar Factory Distillery Plant 17
Figure 4.1- Normal plot of Residual for Alcohol % fermented wash 38
Figure 4.2- Response surface plot for fermented wash Alcohol content at 8.67% sugar
Concentration(15º brix) at optimized nutrient application 39
Figure 4.3- Response surface plot for nutrient optimization of yeast cell count at 8.67%
(15º Brix) Fermentable sugar concentration fermented wash 40
Figure 4.4- Response surface plot for nutrient optimization of Residual sugar at 8.67%
(15º Brix) Fermentable sugar concentration fermented wash 40
Figure 4.5- Response surface plot for nutrient optimization of fermented wash Alcohol content
at 11.5% sugar Concentration(20º brix) at optimized nutrient application 46
Figure 4.6 - Response surface plot for nutrient optimization of yeast cell count at 11.5%
(20º Brix) Fermentable sugar concentration fermented wash 47
Figure 4.7- Response surface plot for nutrient optimization of Residual sugar at 11.5%
(20º Brix) Fermentable sugar concentration fermented wash 47
Figure 4.8- Response surface plot for nutrient optimization of fermented wash Alcohol content
at 14.3% sugar Concentration(25º brix) at optimized nutrient application 53
Figure 4.9 - Response surface plot for nutrient optimization of yeast cell count at 14.3%
(25º Brix) Fermentable sugar concentration fermented wash 53
Figure 4.10- Response surface plot for nutrient optimization of Residual sugar at 14.3%
(25º Brix) Fermentable sugar concentration fermented wash 54
Figure 4.11- Response surface plot of Alcohol% fermented wash for effluent treatment
At fermentable sugar concentration of 8.67%( 15º brix ) fermented wash at
Constant process condensate re circulation of 50 % to fermenter 62
Figure 4.12- Response surface plot of Alcohol% fermented wash for effluent treatment
At fermentable sugar concentration of 8.67%( 15º brix ) fermented wash at
Constant Spent lees re circulation of 100 % to fermenter 63
Figure 4.13- Cubic graph of Alcohol% fermented wash at 8.67% fermentable sugar
(15º brix) and effluent recirculation 63
Figure 4.14- Response Surface plot for cell mass count at constant process condensate
Recirculation and fermented wash fermentable sugar content of 8.67% (15ºbrix) 64
Figure 4.15-Response surface plot for cell mass count at constant spent lees re circulation 65
xi
Figure 4.16- Cubic graph of yeast cell mass count to show interaction of factors 65
Figure 4.17- Response surface plot for residual sugar concentration 66
Figure 4.18- cubic graph for residual sugar concentration of fermented wash at 15º brix sugar
concentration 66
Figure 4.19- Response surface plot for alcohol % fermented wash at different sugar
Concentration 74
Figure 4.20- Cubic graph for alcohol% fermented wash at different sugar concentration and
50% process condensate recirculation 74
Figure 4.21- Cubic graph of alcohol% fw at 100%spentlees recirculation rate 75
Figure 4.22- Response surface plot for cell mass count at different sugar concentration and
effluent recirculation rate 76
Figure 4.23- Cubic graph for cell mass count at different sugar concentration and effluent
recirculation rate 76
Figure 4.24- Response surface plot for residual sugar of fermented wash 77
Figure 4.25-Cubic graph for residual sugar concentration of fermented wash at different
sugar concentration 78
Figure 4.26-Solid balance around buffer tank 81
Figure 4.27-Energy balance on buffer tank 81
Figure 4.28-Energy balance on cooler 82
Figure A1- yeast cell counting by Hemocytometer 103
Figure A2- Hemocytometer counting chamber and fill point 103

xii
Lists of Appendices
Appendices Title page
Appendix –A : Preparation of fermented wash sample 94
Appendix – B: Modeling Experimentals samples from Bulk fermentation Process 94
Appendix – C: Analysis Results and Experimental runs for Effluent Re circulation 96
Appendix –D : Yeast Counting by HeamoCytometer 102
Appendix –E: Compositional analysis results of input materials and effluents 104
Appendix –F: Test procedures for alcohol quality analysis 109

xiii
Lists of Symbols and Acronyms
Acronym Nomenclature
ANOVA Analysis of variance
ATP Adenosine triphosphate
BOD Biological oxygen demand
Brix /BX Percent by weight of soluble solid in sugar solution
COD Chemical oxygen demand
DAP Di ammonium phosphate
EDTA Ethylene diamine tetra acetic acid
FS Fermentable sugar
FW Fermented wash
GL Gaye Lukas
ICUMSA International Commission for Uniform Method of Sugar Analysis
MSF Metahara Sugar Factory
PC Process Condensate
SAO Standard Adminstrative order
SD Standard Deviation
SL Spent Lees
SW Spent Wash
t/d Tons/ day
TRS Total reducing sugar
UFS Un fermentable sugar
YSM Yeast separator machine
SMRI Sugar Milling Research Institute

xiv
Abstract
Distilleries are one of Agro based industries which produces several types of liquors to serve
communities as a drink, feedstock for other industries and power alcohol. While producing these
valuable products for consumption more than 6 times effluents with high COD and BOD values has
been produced. This effluent will cause huge disaster on environment if it is discharged without
treatment. On the other hand treating such huge quantity of effluent will incur additional cost to the
production processes of the distilleries. Therefore, this study is conducted with the view of reducing
the volume of effluent generated through proper nutrient application and optimum recirculation of
effluent to fermenters.
This study was conducted first by making an optimization of macronutrients for the existing
fermented wash sugar concentration of 15 brix and application of 62.51 and 14.51 mg/L of wash
for Urea and DAP respectively has been found an optimum. Using this optimum nutrient dosage test
for effluent recirculation were conducted for the existing fermentable sugar concentration of 8.67%
(15 brix). Data of alcohol% fermented wash, yeast cell mass count and residual sugar were generated
and analyzed using response surface methodology and application of 1% spent wash, 92% spent
lees and 99% of the generated effluents were found to be safely re-circulated to fermenters.
Anticipating future change in operating conditions of fermentation process the study has been
conducted at different fermentable sugar concentration of 8.67%(15 brix), 11.5%(20 brix) and 14.3%
(25 brix). Data are generated and analyzed using response surface methodology and it was found
that fermentable sugar concentration of 14.3%(25-brix), macro nutrients of 62.51mg/L Urea and
31.26 mg/L DAP, recirculation of the generated 50% of spent wash, 68.5% of spent lees and 85.83%
of process condensate taken as an optimum input and alcohol % fermented wash of 7.43, yeast cell
mass of 2.40674E+008 and residual sugar concentration of 0.417% are found as a response variable.
Determination of different attribute of rectified sprit produced by using an optimum recirculation
rate of effluent to fermenters and it is found in safe range from food safety and quality points of
view. Besides, sizing of equipments such as cooler, pump and piping required to handling effluent
recirculation were done. Moreover, financial analysis was carried out and it was found viable. The
study shows that the initial investment will be fully returned within eleven months.

xv
 Optimization of Molasses Fermentation to Reduce Effluent

1. Introduction
1.1. Background
Today, water resources have been the most exploited of the natural systems, most of our water
bodies are seriously polluted due to rapid population growth, industrial proliferation,
urbanizations, increasing living standards and wide spheres of human activities. Many rivers of
the world receive heavy flux due to industrial effluents. The wastewater consisting of
substances varying from simple nutrients to highly toxic hazardous chemicals, which when
used for irrigation caused both beneficial and damaging effects to various crops including
vegetables (Madhavi, A, & Rao, A. P.,2003).
Many industries are playing the crucial role in water pollution such as textile industries, dairy
industries and distillery etc. Distilleries are the major agro-based industries, which utilize
molasses as raw material for the production of rectified spirit. In addition to rectified spirit,
distilleries also produce ethanol, which can be mixed with diesel and used as biofuel and helps
in reducing import of crude oil thereby saving foreign exchange.
Ethyl alcohol can be produced synthetically by indirect or direct hydration of ethylene,
fermentation of sacchariferous (sugar containing) or amylaceous (starch containing) raw
materials, by hydrolysis and fermentation of cellulose from sulphite liquors of paper factories.
The proportion of ethyl alcohol production in U.S.A., Western Europe, Japan and Brazil are
given below.
 Fermentation process --------------- 94.6%
 Synthetic process -------------------- 5.0%
 Sulphite liquors ----------------------- 0.4%
 Total --------------------------------- 100%
It can be seen from the above data that fermentation process is the primary source of ethyl
alcohol in the above regions.
Ethanol production from cane or beet molasses has been done for longer time by using
fermentation processes as a drinking liquor and bio fuel. For longer time fermentation is done
by using batch process in the world till 1985. But currently utilization of continuous bio reactor
has been started to produce alcoholic distilled liquors and power alcohols and brought an
enormous benefits to this sector with regard to reduction in man power requirements as well as
in eliminating batch to batch variation.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 1
 Optimization of Molasses Fermentation to Reduce Effluent

Production of ethyl alcohol as a beverage or drinking liquors in Ethiopia was started in the
house hold level and practiced for long time in the high land parts of the country. This liquor is
manufactured on fermenting maize, barley and other cereals. In industrial scale production of
fermented liquors in Ethiopia had been started at small scale level in different industries like
Mollamaru, Balezaf, National Alcohol and liquor Factories and now more than 10 Factories are
available in the country. Besides, 2 of the existing sugar factories Metahara and Finchaa were
equipped with ethanol plant to utilize their by product molasses to value adding product such
as impure sprit, rectified sprit and power alcohol. The country has huge potential with respect
to production of alcohol for beverage as well as bio fuel and Ethiopian sugar corporation has
planned all sugar factory projects to have ethanol plant.
Recent advancement in fermentation technology creates good opportunity in the development
of continuous types of Bio- fermentor. In Ethiopia the only distillery which utilizes this
technology has been the Ethanol plant which was installed at Metahara Sugar Factory in 2011
by the Indian company called KBK- Chemical Engineering PLC. It has the capacity of
producing 28 m3/hr fermented wash on using 4 continuous stirred tank reactors arranged in
series. In addition it comprises pre fermenter , culture propagating vats, air and molasses
supply system as well as sludge settling units, water and recirculated vinase supply systems.
The main goal of Ethanol plant is to have an efficient and profitable as well as Environmental
friendly operation with the required Rectified, technical as well as Power alcohol quality and
maximum recovery. One of the biggest problems in Ethanol plant is the disposal of acidic and
concentrated effluent generated from distillation section. Damping this effluent directly to river
stream will harm aquatic animals. In other hand treating the effluent will incur additional non
value adding cost to the product.
1.2. Statements of the problem
Fermentation can be done in a more efficient and cost effective ways through continuous bio
reactor, but as this technology is recently adopted to ethanol production particularly in our
country the process parameters to obtain optimum output is not set up to now. Because of this
variation in the performance from time to time as well as from shift to shift is observed. In
addition, the effluent spent wash sent to biocompost plant is very high as compared to the
design of the distillery and the plant also releases higher quantity of spent lees and process
condensate water from vinasse concentrating unit (Falling Film Quadruple effect evaporators).

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 2
 Optimization of Molasses Fermentation to Reduce Effluent

About 350 tons/day concentrated vinase, 236 tons/day process condensate and 87.4 tons/day
spent lees are produced. Bio compost plant has the capacity of 245 tons vinasse to handle daily
the remaining will be accumulated in the pond and latter during off season it is directed to river
and other areas. This condition contributes to pollution of the environment and penalizes the
organization as it violates environmental regulation. In order to safeguard the environment the
surplus vinasse has to be utilized on devising different methods like Bio-methanation, adding
to the sugar cane plantation with irrigation water and recycling some portion of the vinasse to
fermentation process(N.k.Saha etal,2004).
1.3. Objectives
1.3.1. General Objectives
The main objective of this research is to study the effect of reusing distillery effluent to
fermentation process in order to maximize alcohol production and reduce effluent generation.
1.3.2. Specific Objectives
• Conduct compositional analysis of input materials.
• Study the effect of varying different nutrients and concentration of fermentable sugar
on the overall performance of fermentation process.
• Study the effect of using spent lees, spent wash and process condensate water
obtained from vinase multiple effect falling film evaporators in order to reduce
effluent.
• Study the effect of yeast cream separator machine on bulk fermentation process.
• Carrying out the impact of re-circulating effluent on the food safety and quality of
rectified sprit as drinking liquor.
• To carry out material and energy balance for fermentation process and suggest
suitable vinasse, spent lees and process condensate cooler and piping system for
recycling.
1.4. Significance of the Study
In these days the competitiveness of industries mainly depends on their own performance of
utilizing all available resources and minimizing waste. In addition, due to government policies
concern for environment create high pressure on organization. As a result, in order to assure

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 3
 Optimization of Molasses Fermentation to Reduce Effluent

their sustainability, organizations start to focus on the reduction of effluent streams as well as
improve their productivity (product output per input).
The main aim of this thesis is to provide vital information to the power alcohol and liquor
producing factories with the view of preventing environmental pollution, maximizing
profitability and reducing cost of production through utilizing the available resource
efficiently. Moreover, this thesis work is expected to provide vital information for liquor
factories, sugar and other food industries which mainly faces high effluent problem. In
addition, it can also serve as a reference in the area of distillery technology.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 4
 Optimization of Molasses Fermentation to Reduce Effluent

2. Literature Review
Production of ethyl alcohol from cereals and molasses for drinking and power alcohol
comprises different unit processes and operations such as Fermentation, Distillation, Water
treatment, ingredient preparation and Effluent handling process. In order to obtain good quality
and maximum conversion of fermentable sugar in to ethanol it is necessary to focus on
fermentation process (DR.Nalini B.2009)
2.1. Fermentation Process
Fermentation is a bio-chemical process in which six carbon sugars such as fructose, glucose,
levulose etc are converted to ethyl alcohol and carbon dioxide which is catalyzed by enzymes
generated from yeast. Any polymeric hexose sugar present in any materials in the first instance
must be converted to hexose sugar units and then fermented in to ethyl alcohol. Conversion of
polyhexose to hexose is called hydrolysis and it is carried out either enzymatic ally or by
chemical means in batch and continuous fermentors. The major processing steps in alcoholic
fermentation are:
• Raw material (substrate) preparation
• Yeast propagation( inoculums preparation)
• Final fermentation
2.1.1. Raw Material Preparation/Molasses treatment
Objective of molasses treatment (preparation):-
 To reduce level of impurities
 To obtain better performance & yield
 To reduce scale formation
 Lower steam consumption
Any sugar containing substances can be used as a raw material for alcohol production. The
raw material preparation process differs from substance to substance. Here we will consider the
raw material (molasses) which is used for ethanol production in distillery plants integrated to
sugar factories. The molasses supplied to the distillery must be prepared for alcoholic
fermentation. Raw material preparation includes dilution molasses to adjust the concentration
of the molasses media suitable for yeast, heating the media to increase the reaction of acid with
calcium in side molasses to avoid scale formation within distillation unit and decantation to
remove out precipitate (Ethiopian sugar corporation Training Manual).
AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 5
 Optimization of Molasses Fermentation to Reduce Effluent

2.1.2. Yeast Culturing and Recycling System

Yeasts are eukaryotic micro-organisms classified in the kingdom Fungi, with about 1,500
species currently described. These unicellular organisms reproduce asexually by budding.
Yeast species grow aerobically. Unlike bacteria, there are no known yeast species that grow
only an aerobically. Yeasts grow best in neutral or slightly acidic pH environment. Most yeast
cells die above 50ºC. There is little activity in the range of 0ºC to 10ºC. The cells can survive
freezing under certain conditions with viability decreasing over time. The most common mode
of vegetative growth in yeast is asexual reproduction or fission.
Propagation of yeast are done in two main steps such as first in the lab scale followed by
culturing in the bulk fermentation step on using culture vessels to increase the number of yeasts
suitable for bulk fermentation. In the lab about 10 liter molasses solution 14.0 brix in water and
maintain its pH 4.0-4.2 by sulfuric acid /H 3 PO 4 with following supplements- 10 gram urea,
DAP 5 gram, 1.5 gram magnesium sulfate, 5.1 gram zinc sulphate, 30 gram yeast extract.
About 20 ml of this media is taken and poured to test tube which contains slanted yeast and
after propagation for 10 hours the media will be transferred to 250 ml flask having 125 ml
prepared media. Here after retention time of 10 hours it is transferred to 1 liter flask which
contains 800 ml media and finally after retention time of 10 hours the yeast mass is transferred
to bigger stainless vessel having 15 liter gross volume and 8 liter media here after passing for
10 hours, the propagated media will be transferred to fermentation station for further
propagation.
In bulk fermentation process culturing of yeast is done by using four culture vessels(1-4) which
having 125, 600, 6000 and 40,000 liters respectively (KBK Chemical Engineering PLC
manual).
2.1.2.1. Media Preparation
Media of molasses have 12-14 brix is taken nutrient such as Urea, DAP, Zinc Sulphate,
Magnesium Sulphate are added. In addition, to bring pH of media between 4.3- 4.6 and the
media is pasteurized by heating with steam to temperature of 95 to 100°C for one hour. Then
close the steam and steam condensate valve and cool the vessels by opening cold and hot
water valves till 32 to 35 degree centigrade temperature is attained ( Sharma KK, KBK
Chemical Engineering Manual).

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 6
 Optimization of Molasses Fermentation to Reduce Effluent

2.1.2.2. Inoculation
The prepared media in culture vessel number 1 is inoculated by yeast culture propagated in
laboratory and finally the culture media after retention time of 10 hours it will be transferred to
relatively bigger size of culture vessel number 2 for multiplying the yeast cells. This
propagation process will continue in similar fashion in culture vessel -2 and transferring will be
done to culture vessel-3. From number-3 the propagated yeast will be passed to culture vessel-
4 (pre-fermenter) and propagation of yeast culture will be carried out till the yeast
concentration exceeds 250 million per ml media. The yeast mass from culture vessel number-4
will be pumped to fermenter-1and the mass growth as well as alcohol production will continue
in fermenter-1and 2 aerobically. And production of alcohol will resumed in fermenter-3 and 4
an aerobically (KBK, KaurU1995 and Sharma KK).

2.1.2.3. Yeast Recycling

In order to get maximum amount of alcohol from the same quantity of molasses matured yeast
will be re circulated. This process will be done by using either Separating the yeast after the
fermentation is completed in a fermentation tank by using a high speed centrifugal machine
and recycling the yeast so separated or separating the yeast from the fermented wash by
sedimentation of the yeast by adopting suitable technology. In the case of Metahara Sugar
factory recycling of yeast is done by directing fermented wash to centrifugal yeast cream
separator revolving at 6500rpm. The yeast cream will be directed to continuous fermenter
number-2(KBK, Peter Rein,2007).

2.1.3. Final Fermentation Process

Fermentation is a process by which a chemical changes are brought about in an organic
substrate through the action of biochemical catalysts, called enzymes, elaborated by specific
types of living microorganisms. It is a metabolic process characterized by: incomplete
oxidation, and the transformation of large amounts of substances by comparatively small
amounts of organisms (paturau, 1989). There are different types of fermentations among which
the batch and continuous fermentation processes are the most common.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 7
 Optimization of Molasses Fermentation to Reduce Effluent

2.1.3.1. Batch Fermentation

Fermentation of molasses can be carried out in batch as well as in continuous bio reactors. In
batch fermenter, molasses having brix of 90 and 50% fermentable sugar is diluted with water
to give a solution of about 25 bx and 15 % fermentable sugars. This solution is called Wort.
The wort is pumped in to fermentation tanks, which are made of mild steel. Yeast which is
usually a strain of saccharomyces cervisiae is added to the wort contained in the fermentation
tanks. In some distilleries the yeast is developed every day from a test tube slant to about
20,000 liters in four stages of propagation by aerobic fermentation. The wort used for the
development of yeast is sterilized with steam heating, so that the yeast is not contaminated with
any kind of microorganisms which may lead to the production of undesirable products, other
than alcohol during fermentation. As the fermentation starts and the yeast multiply, a part of
the fermentable sugars content in the wort is consumed by the yeast for its own survival and
multiplication. Then anaerobic fermentation takes place, when the enzyme invertase contained
in the yeast converts the disaccharides like sugar in the molasses in to mono saccharides like
glucose and fructose. Subsequently the Enzyme ‘Zymase’ contained in the yeast converts the
monosaccharide’s in to ethyl alcohol and carbon dioxide. When CO 2 tries to escape from the
fermented wort it forms bubble and foam and needs addition of anti foaming agents like turkey
red oil. Since yeast needs nutrients for their survival, nutrients like ammonium sulphate, urea,
Phosphatic salts, etc are added to the wort in small quantities. In some distilleries in order to
arrest the growth of undesirable micro organisms, antibiotics like Benzyl Pencillin is added to
the wort in small doses. To maintain fermentation pH in the range of 4.3-5 sulfuric acid has
been usually added to the wort.(AgrawlPk etal,1998)
Anaerobic fermentation of feed stock in the distillery is an exothermic process. Therefore, the
temperature of wort in the fermentation tank goes up. The ideal temperature during
fermentation is between 30 to 33ºC, and beyond 37ºC, the yeast becomes in active and the
yield of alcohol will be reduced. Therefore, cooling of the fermented wort has been done by
spraying water on the surface of the fermenter to keep temperature of the fermenter in the
range of 30-33ºC. After the fermentation is completed in about 36 to 48 hours, which is
indicated by chemical analysis, the fermented wort is now known as wash containing about 6
to 8% alcohol in the wash. The activity of yeast will slow down as the alcohol % of wash goes
up. The fermentation efficiency, which is the ratio of actual production of ethyl alcohol in the

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 8
 Optimization of Molasses Fermentation to Reduce Effluent

wash to the theoretical yield of alcohol calculated on the basis of the total fermentable sugar is
usually 80-85% in the batch process. The efficiency is relatively low as most of the batch
fermenters are open to atmospheres and liable for attack by micro organisms as well as for the
escape of alcohol with CO 2 as the fermentation was carried out briskly. In some distilleries
CO 2 the fermentation tanks are covered with mild steel hoods to collect and scrub it with
water and use this water for diluting the molasses and there by recovering the alcohol escaping
with the CO 2 ( Bansal R.,SinghR,2003) .
Control of microbial contamination as well as ethanol quality per batch are considered as a
merit for batch fermentation where as requirements of high capital investment for large scale
production, decrease in volumetric efficiency of fermentors and possible variation in
fermented wash quality from batch to batch on the other hand are taken as a demerit.

2.1.3.2. Continuous Fermenter

In continuous fermentation process, additional mash is added continuously or more commonly,

at short intervals to the fermenters. Continuous fermentations are simply a number of separate
fermentation tanks in series. The mash should enter at the bottom of the first fermenter, exit the
top, enter the bottom of the second fermenter, and so on. Total fermentation time is determined
by dividing the total capacity of the fermenter by the flow rate. There are different types of
continuous fermenters designs like Biostil, Hiferm, Encillium, Hoechst-UHDE, STRACOSA”
Membrane” all are working under continuous operation and with greater fermentation
efficiency it is greater than 90%. However, fermented wash brix was varied for different
fermenter design and consequently the alcohol % fermented wash is to the higher sides.
Continuous fermentation has also advantage and disadvantages. In Metahra Sugar factory 4-
continuously mixed fermenters are connected in series with all the accessories like PHE for
cooling, air spargers, broth mixers, agitators and air blowers, yeast recycling system, etc.
High volumetric efficiency, Yeast recycling, a more consistent product than can be obtained
from batch fermentation, lower capital and lower labor cost and reduced time requirement are
considered to be an advantage for continuous fermentation processes while high potential for
serous microbial contamination is taken as a disadvantage.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 9
 Optimization of Molasses Fermentation to Reduce Effluent

2.1.4. Chemical Reaction During Fermentation

There are different chemical reactions occurred during fermentation processes such as the
production of alcohol from glucose by the action of Zymes-enzyme, inversion of sucrose to
glucose is done by invertase and different other types of biochemical reactions will occur to
form Acetic acid, lactic acid, diacetyl, acetone, Iso propyl alcohol, n- propyl alcohol, n-butyl
alcohol, etc. The biochemical reaction rather than formation of ethanol which occurs during
fermentation of molasses are not important as it resulted in reduction of fermentation
efficiency.
C 6 H 12 O 6 yeast 2C 2 H 5 OH + CO 2 + Heat
Glucose zymase Ethanol carbon dioxide
100 kg 51.1kg 48.9 kg 17935 KJ
C 12 H 22 O 11 + H2O Yeast C 6 H 12 O 6 + C 6 H 12 O 6
Sucrose Invertase Glucose Fructose

2.1.5.Factors Affecting Fermentation of Molasses

In the anaerobic pathway every mole of glucose is converted into 2mole of ethanol, 2mole of
carbon dioxide, and 2mole of ATP along with 56kcals of heat. The ATP produced is used in
biosynthesis or maintenance. In this pathway every gram of glucose converted will yield
0.511gram of ethanol. However, in practice since both biomass and some secondary products
are produced by the yeasts, the yield is reduced to 95% of the theoretical yield. In certain cases,
the yield is further reduced to typically to only 90% of the theoretical yield when complex
substrates are used. On the other hand, under aerated metabolism, sugar is converted
completely to CO 2 , water, cell mass and by-products with no ethanol formed and is typical
baker’s yeast fermentation as it yields more ATP, (38moles) and more biomass(0.52g/g of
sugar). ATP is a nucleoside triphosphate used in cells as co-enzymes which is capable of
transporting chemical energy within the cell for metabolism.

2.1.5.1. Effect of sugar concentration

Hexoses sugars such as glucose are the primary reactants in the yeast metabolism. Under
fermentative conditions, the rate of ethanol production is related to the available sugar
concentration by a Monad type equation

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 10
 Optimization of Molasses Fermentation to Reduce Effluent

V = V max C s 2.1
(K s + C s )
Where, V is the specific ethanol productivity (g of ethanol/g of cell/h)
C s is the sugar substrate concentration (g/l) and
K s is the saturation constant having a very low value typically 0.2 – 0.4 g/l (Maiorella,
1985). At very low sugar concentration(less than 0.3%), the yeast is starved and the
productivity decreases. Sugar concentration up to 15%, the rate of ethanol production per cell
is essentially at its maximum, beyond this concentration, catabolite inhibition of enzyme in the
fermentative pathway takes place and the conversion rate is slowed. At concentration above 0.3
to3% the production of oxidative enzyme is inhibited, thus forcing fermentative metabolism.
This catabolite suppression is a desirable characteristic in industrial strain.

2.1.5.2. Effect of pH
The pH is the measure of acidity or alkalinity of aqueous solution expressed on scale of 1 – 14.
Neutral is pH 7, pH 1-7 is acid, and pH 7-14 is alkaline. The pH is most conveniently measured
with test papers that change color according to the pH of the solution being tested.
Control of pH during the mashing and fermentation process is important for two reasons; the
growth of harmful bacteria is retarded by acid solutions, and yeast will grow only in an
(slightly) acid solution.
Generally Distiller’s yeast shows a broad pH optimum from 4 to 5. The development of
bacteria is severely repressed at pH value under 5. The acid most commonly used is sulfuric
acid, although any mineral acid is perfectly suitable. Further, yeast can tolerate as low pH as 2
without permanent damage.

2.1.5.3. Effect of Temperature

As heat energy is liberated during fermentation of sugar by yeasts there is always an increase
in temperature and cooling of fermenters is required, and, therefore, it is desirable to use
temperature tolerant strains. Most strains have a temperature growth optimum of 30 – 35ºC.
However, the optimum fermentation temperature at a low alcohol concentration is often
slightly higher (38ºC), but alcohol tolerance is improved at reduced temperature. S. Cerevisiae
strains having 37ºC as optimum temperature for ethanol production have been selected.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 11
 Optimization of Molasses Fermentation to Reduce Effluent

Exposure to higher temperature results in excessive enzyme degradation and loss of yeast
viability. Yeast can be stored inactive at low temperatures (above 0ºC) and are readily revived.

2.1.5.4. Effect of Oxygen

It has been found that trace amount of oxygen may greatly stimulate yeast fermentation, which
is used as building block for synthesis of poly-unsaturated fats and lipids required in
mitochondria and plasma membrane . High sugar concentration is adequate to repress aerobic
sugar consumption in yeasts which show the Crabtree effect. For other yeasts or at low sugar
concentration, the oxygen supply should be limited. Trace amounts (0.7 mm Hg oxygen
tension) of oxygen are adequate and do not promote aerobic metabolism.
Conversion of sucrose to glucose is taking place outside cell wall, where as conversion of
glucose to ethanol takes place inside cell wall in the mitochondria. If there is no sufficient air
the yeasts will be weak.

2.1.5.5. Effect of Ethanol and Secondary Component Inhibition

Ethanol tolerance is a desirable trait in industrial yeast strains. Ethanol concentration above 2
%( w/v) show toxic effect and cell growth is halted (stopped) completely at about 11% (w/v).
However, slow fermenting sake yeast (Saccharomyces sake) can tolerate ethanol concentration
up to 16 %( w/v) at low temperatures as it contains lipoproteins. Ethanol inhibition is directly
related to the inhibition and denaturation of important glycolytic enzymes as well as to
modification of the cell membrane. S. Cerevisiae that can tolerate high ethanol (12%v/v) at
400C is being isolated by chemo stat selection.
The by-products of fermentation such as acetate and lactate may inhibit yeast growth and
ethanol production at high concentration. Some non metabolized feed components may
become concentrated when back-setting is used and may inhibit the yeast growth.
Blackstrap molasses may contain high concentration of calcium salts which are inhibiter to
yeast growth. Sterilization of sugar at high temperature in presence of salts (phosphates) and
proteins can produce yeast toxins. High salt concentration may encourage glycerol formation.
Yeasts may tolerate 16 to 20% non fermentable dissolved solids in industrial fermentation.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 12
 Optimization of Molasses Fermentation to Reduce Effluent

2.1.6. Nutrient Requirements of Fermentation

In addition to sugar the fermentation mash must also be provided with additional nutrients for
cell growth and maintenance. Glucose medium supplemented with Ammonium Chloride,
Magnesium Sulphate & yeast extract supports rapid cell growth and ethanol productions. The
yeast extract provides all necessary growth factors such as amino acid, purines, vitamins as
well as minerals. Phosphorus, Potassium Magnesium & Calcium are incorporated in to cell
mass and are also Co- factors (activators) of several enzymes. Ammonium ion provides
nitrogen for protein nucleic acid synthesis. Vitamins such as biotin & pantothanate are
stimulatory to growth, while thiamine may also increase ethanol tolerance. In Methara Sugar
Factory distillery plant DAP and Urea are used as nutrient and there admission to fermentors
are monitored arbitrarily based on the output of alcohol as 1 kg/kl of ethanol for DAP and 3.5
kg/kl of ethanol for urea requirements. In order to improve productivity of the plant and reduce
the effluent generation it is essential to obtain the optimum nutrient requirement.
2.2. Distillation

Distillation of fermented mash is the next important step in production of alcohol after
fermentation. This step consumes a considerable amount of energy and is also a deciding factor
in the quality of ethanol produced. Hence, in line with the demand of the industry, efforts have
always been to minimize requirement of energy and to improve the basic quality of alcohol
produced. Ease of operation, reliability, lower down time and flexibility of operations are other
parameters considered during the design.
Distillation is a method of separating liquid mixtures based on differences in their volatilities in
a boiling liquid mixture. Distillation is a unit operation, or a physical separation process, and
not a chemical reaction. It is broadly defined as the separation of more volatile components
from less volatile components by a process of vaporization and condensation.
The term distillation is properly applied only to those operations where vaporization of a liquid
mixture yields a vapor phase containing more than one constituent and desired to recover one
or more of these constituents in a nearly pure state.
The process of distillation is affected based on the relative volatilities of the liquids in the
mixture and taking advantage of their different boiling point. Distillation is the most widely
used method of separating liquid mixture and is at the heart of separation process in many

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 13
 Optimization of Molasses Fermentation to Reduce Effluent

chemical and petroleum plants. The basic requirement of separation of components by

distillation is that the composition of the vapor be different from the composition of the liquid
with which it is in equilibrium. Theoretically, distillation can never yield a component in
absolutely pure form, although practically the product may be made of any purity that is
economically warranted. In distilleries there is certain limit of alcohol production during
fermentation. Usually the alcohol obtained by fermentation is within the range of 6.0 – 12% by
volume. The alcohol obtained from fermentation process should be strengthened to the desired
quality to be used for specific purpose. With the view of reducing energy consumption, scale
formation in trays and reduced by-product formation in the mash column recently distillation
column is designed to operate at multi-pressure in similar principles to that of multiple effect
evaporators which consists of distillation columns operating under different pressures.
Thermal energy from columns operating under pressure is recycled to columns under vacuum
to conserve energy.
Multi-pressure vacuum distillation system consists of three distillation columns namely
• Primary distillation column which consists of degasify column and analyzer column
which operate under vacuum.
• Rectifier cum exhaust column – operated under pressure.
• De - aldehyde column may be added to improve quality of alcohol further.
2.2.1. Primary Column
Fermented wash is preheated in fermented wash pre-heater by using thermal energy of spent
wash and fed at the top of the primary column(degasifying section), the working pressure and
temperature are maintained in the range of -400 to -450 cmHg and 68-72˚C respectively. Here
lower boiling point impurities are separated from fermented wash. The vapour withdrawn
from this section has the concentration of 40-45 ˚GL and it will be condensed in the condenser
to be directed to de- aldehyde column feed tank. The remaining wash is gravitated down to
analyzer section and primary alcohol vapors having 45-50˚GL separated out at the top of this
section and directed to the steam chest of falling film evaporator to provide thermal energy for
concentrating spent wash(vinasse). The alcohol vapour after giving its latent heat to the vinasse
it will be directed to rectifier column. Rest of the fermented wash flows down and is taken as
spent wash from analyzer column bottom. Apportion of spent wash is directed to primary
column re-boiler and it is vaporized by getting heat from alcohol vapor coming from Rectifier

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 14
 Optimization of Molasses Fermentation to Reduce Effluent

column. The condensed alcohol coming out from primary column re boiler is transferred to
rectifier reflux tank and supplied to rectifier column as a reflux.
2.2.2. De-aldehyde Column
The alcohol vapour which is separated from fermented wash in degasser section of primary
column at the temperature range of 68-72ºC contains high aldehyde and ketone mixture. This
mixture is condensed in degasser condenser and gravitated to aldehyde feed tank to be pumped
to the de-aldehyde column feed tray. Low boiling impurities are concentrated in the de-
aldehyde column. A top draw is taken out as impure alcohol from the top of the de-aldehyde
column. The bottom of de-aldehyde column is sent to rectifier feed tank.
2.2.3. Rectifier Column
Rectifier exhaust is operated under pressure and heats analyzer column through re boiler.
Alcohol is enriched towards the top and is drawn out as rectified spirit. Fusel oil build-up is
avoided in the rectifier column by withdrawing out side streams of fusel oil from the middle of
the rectifier column. And is sent to decanter for further separation. The fusel oil wash water is
recycled back to the column. Bottom temperature of rectifier column is operated in the range of
120-125ºC and top temperature is generally kept in the range of 95-97ºC. Spent lees which
mostly contains negligible alcohol and environmentally pollutant liquid is drawn as a bottom
product and apportion of which is vapourized on passing through rectifier re-boiler using steam
as a thermal energy source.
2.3. Alcohol Dehydration
The purpose of dehydrating Rectified Sprit Alcohol is to produce power alcohol which is
suitable for blending with petrol. It can be done by using Azeotropic distillation, by utilizing
molecular sieve beds or by using extractive distillation. In Metahara Sugar Factory molecular
sieve beds are utilized for the purpose of dehydration. It comprises the following steps : the
rectified spirit will be vapourized in the Recovery Column by getting thermal energy from
steam supplied to recovery reboiler. The vapour leaving from top tray of Recovery column
having temprature of 107-112ºCis Superheated to temprature of 130-140ºC in Shell and tube
Super heater. Lower concentrated liquor or liquid having nil alcohol(Spent lees) is drawn from
the bottom of the Recovery Column. The superheated alcohol vapour is admitted to Packed
Molecular Sieve bed and water vapour is adsorbed by the molecular beads and dehydrated
alcohol vapour is extracted from the top of the bed. The dehydrated alcohol vapour is
AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 15
 Optimization of Molasses Fermentation to Reduce Effluent

condensed in the product condenser cooled in the product cooler and sent to alcohol storage
tanks. For this purpose two packed bed of molecular sieve beds are employed when one is in
operation the other will be in regeneration mode. Regeneration will be carried out by
employing vacuum pumps. The adsorbed water vapour is sucked by vacuum from the beads
and collected as avapour, then it is condensed in the Regeneration condenser and returned to
Recovery Column(Peter Rein,2007).
2.4. Overall Water Balance
Distillery plants of Metahara Sugar Factory is designed to produce about 42.5 tons per day of
Rectified Sprit alcohol and 40 tons per day of Ethanol power alcohol. For this purpose the plant
consumes about 206 tons of molasses having 90 % brix , 571 tons per day process water and
400 tons per day soft water. About 713 tons per day fermented wash with 7.62% alcohol
content and 20 tons per day sludge will be produced from fermentation section and pumped to
distillation section. In the distillation section about 585 tons per day spent wash and 85 tons per
day spent lees with pH of 3.3-3.8 as an effluent and 42.5 tons per day rectified spirit. The
rectified sprit is dehydrated in dehydration section and about 40 tons per day Ethanol of
99.99% strength and 2.4 tons spent lees. The voluminous dark brown spent wash is
concentrated in multiple effect Evaporators and produces 350 tons per day concentrated
vinasse and 236 tons/day process condensate.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 16
 Optimization of Molasses Fermentation to Reduce Effluent

Process water 571tons/day 44t/d

CO 2
RS alcohol
Molasses Fermentation Vacuum Distillation Storage 42.5
Storage tons/day
206t/d Yeast rec.
100t/d Spent lees RS to market

85 T/day
Yeast Separator
713t/d
Sludge 20 t/d Anhydrous
Vinasse (Spent
Section
Wash) 585T/d
concentrator(
evaporator)
Filter cake Spent lees

Conc. Vinasse
Process cond. 2.4 tons/day
350 tons/day to 236t/d Power Alcohol
bio compost
40 tons/day

Fig 2.1 Total material balance of Metahara Sugar Factory distillery Plant

Table 2.1 Type and total quantity of effluent generated

Types of Concentrated Spent Lees Process Total
effluent Spent wash condensate
Quantity
(tons/day) 350 87.4 236 655.4
Source : KBK Chemical Eng’g Operation Manual

2.5. Distillery Effluent Treatment

Distilleries are producing higher quantities of effluents which are produced from Primary
Column, Rectifier and Recovery Column, Spent wash concentrating falling film evaporators as
well as washing of process equipments and leakages of pumps, lines and different equipments.
Leftover which is removed from the bottom of primary column is called Spent wash. The
proportion of spent wash is nearly 14– 15 times the total alcohol production. The disposal of
large quantities of distillery effluent poses environmental problems, as it contains a

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 17
 Optimization of Molasses Fermentation to Reduce Effluent

considerable amount of organic material and has high biological oxygen demand (BOD) and
chemical oxygen demand (COD) (Manonmani et al., 1990, Raghukumar et al., 2004).
Distillery spent wash is the unwanted liquid waste produced during the production of alcohol
and it is one of the most important environmental issues. The distillery spent wash with its
characteristic unpleasant odor and dark brownish colour poses a serious threat to the water
quality around the world (Joshi etal,1994). The ever increasing amount of distillery spent wash
and its disposal has stimulated the need for developing new technologies to process this
effluent efficiently and economically including growth and yield of different crops in
agriculture as supplemented with irrigation water (Sarayu etal,2009), Bio methanation to
produce methane gas for electric generation, to use in fermentation of molasses in place of
dilution water, mixing with press mud to produce biocompost etc. Presently, some distilleries
are started to reduce the BOD level of distilleries effluents by using special types of fish as
well as earth worms.
General characteristics of distillery Effluent is given in the following table 2.2 it contains
high BOD, COD, Potassium content and some salts but the effluent does not contain any toxic
heavy metals as it is a waste from plant materials. Hence, it is possible to utilize this waste
effluent for the purpose fermentation of molasses directly without any treatment.

Table 2.2- General characteristics of distillery effluent

S/N Characteristics Continuous Batch Process

Process(Biostil)
6- 9 liters per liter 15 liters per liter
1 Quantity of spent wash production of Alcohol production of Alcohol
2 Temperature of spent wash 60 -100 60- 100
3 Odor Jiggery Jiggery
4 Color Dark Brown Dark Brown
5 PH 4.3-4.6 4.3 – 4.6
6 BOD. Mg/lit. 90000 – 95000 30000 – 40000
7 COD. Mg/lit. 190000-200000 85000-95000
8 Total solids, Mg/lit. 270000-280000 80000-90000
9 Potassium (K), mg/lit. 18000-20000 8000-10000

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 18
 Optimization of Molasses Fermentation to Reduce Effluent

Table 2.2 continued

10 Chlorides. Mg/lit. 13000-15000 5000-6000
11 Sulphates, Mg/lit. 15000-18000 2000-5000
12 Phosphates as PO 4 , Mg/lit. 1000-1500 800-1200
13 Calcium (Ca), Mg/lit. 2600 – 2700 500 – 600
14 Nitrogen (KTN), Mg/lit. 2000 – 2500 1000 – 1200
15 Sodium (Na), Mg/lit. 300 – 500 200 – 300
(Source : N.K. Saha etal Elsevier, 2005)
2.5.1. Recirculation of Distillery Effluents to Fermentors
One of the method to avoid pollution of river and water streams is to use the distillery effluent
in the place of dilution water in the fermentation section. According to the suggestion made by
Case Tech Consultancy Private limited Company it is possible to use all effluent generated
from Rectifier Column which is called spent lees and process condensate from vinasse
concentrating evaporator without any further treatment except lowering the temperature of
effluents to normal operating condition of fermentation.
According to KBK chemical engineering PLC and the study made by Praj it is possible to
recirculate about 50% of spent wash generated in to fermentation plant. This addition of spent
wash to fermenter is commonly called Back Sloping.
In addition to reduction of effluent generation from distilleries back sloping will be done in the
fermentation station with the view of improving
• pH to control bacteria growth.
• Nutrients that are needed by the yeast for rapid growth.
• Buffering action to fermented wash.
• Utilization of process water and steam (Peter Rein Cane Sugar Engineering 2007, J.
God Bole 2002,).
2.5.2. Effects of Re-circulating Effluent on the Food Quality and Safety of Spirit
Drinking liquors must fulfill certain characteristics to be consumed by people. Among this
Easter as ethyl acetate, miscibility with water, ethyl alcohol concentration, total solid content,
total ash, volatile acidity, copper content and aldehyde content are the predominant one in
judging the quality of sprit (SAO#257,1976 to SAO#358,1978).

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 19
 Optimization of Molasses Fermentation to Reduce Effluent

In order to use rectified sprit produced on re circulating effluent to fermenters for the purpose
of drinking liquors the parameters which are mentioned in Table 2.3 must be fulfilled.

Table 2.3. Standard Administrative Orders (SOS) for different types of distilled Sprit

Type of drink
S/N Chemical Brandy Rum Vodka Whiskey
Composition SAO#358s.1978 SAO#257s.1976 SAO#258s.1976 SAO#259s.1976
1 Ethyl 32.5%, min(free 25˚ proof-42.85 32.5%, minimum
Alcohol from added 30˚ proof-40.01 (free from added
colouring 35˚ proof- 37.15 colouring except
Except carmel carmel prepared
prepared from from sugar)
sugar
2 Total 1%, maximum 0.005%, 0.2% maximum
solids% maximum

3 Total ash 0.02% Free from 0.02% maximum

maximum suspended
matter
4 Volatile acid 50 grams per Heavy 25-60 2 gram per mg 40 grams, maximum
as acetic 100L of Medium 11-24 100 ml of 100 100L absolute
acid Absolute Light 5-10 mg proof, Alcohol.
Alcohol, per 100 ml maximum
maximum
5 Easters as 20 grams per Heavy: 56-565 10 mg per 100 8 grams per 100ml
Ethyl 100L of Medium: 13-55 ml of 100 proof of absolute alcohol,
Acetate absolute alcohol Light: 0.6-12 maximum minimum.
mg per 100mL
of 100 proof

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 20
 Optimization of Molasses Fermentation to Reduce Effluent

Table 2.3 continued

6 Higher 30-350 grams Heavy: 114-238 0.03 (V/V), 30 – 350 grams,
Alcohol as per 100L of Medium:36-136 Maximum per 100Lt of
Amyl absolute alcohol Light: 0-35 Absolute Alcohol
Alcohol Per 100 L Of
Absolute
alcohol
7 Furfural 5 grams per heavy: 1-5.4 5 grams per 100 l
(per 100L medium: 0.6 – of absolute alcohol,
100L of of absolute 0.9 maximum
absolute alcohol, light: 0-0.5
alcohol) maximum per 100 l of
absolute alcohol
8 aldehydes as nil heavy: 17-19 2 grams as 60 grams per 100 l
acetaldehyde medium: 8-16 acetaldehyde, of absolute alcohol,
(per 100l of light: 0-7 mg maximum
absolute mg per 100 mg per 100 ml of
alcohol) proof 100
proof,
maximum
9 copper as cu heavy: 11-14 10 ppm
medium: 7-10
light: 3-6
ppm tannins
(Source : Annex J FAO -2013)

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 21
 Optimization of Molasses Fermentation to Reduce Effluent

3. Materials and Methods

In this thesis mainly two studies were conducted. First, optimization of nutrients such as DAP
and Urea input to fermentors was carried out to know the optimum requirements of nutrients.
After, identifying the optimum nutrient demand the second study was carried out on varying
the proportion of spent wash, spent lees, process condensate and water for the purpose of
making fermented wash.
3.1. Characterization of Input Materials
Compositional analysis of molasses as well as process water used for fermentation were
conducted for fermentable sugar content, brix, Sulphated ash, conductivity ash and volatile
acid content of molasses. Besides, the quality of input water for residual chlorine content, pH
and hardness is also analyzed.
3.1.1. Materials
A. Samples
The sample used to characterize input materials are the molasses collected from Metahara
Sugar Factory ethanol plant and process water from treated water tank has been taken for
analysis
B. Reagents
Concentrated Sulfuric acid is used for determining ash percent of molasses and volatile acidity.
6.34M HCL solution, phenolphthalein solution, 2M NaoH solution, Fehling solution, dry yeast
are used for determination of fermentable sugar in molasses. Lead acetate, ammonia liquid and
Ferro cyanide to determine calcium content. 0.01 Molar EDTA solution, Ammonium buffer
solution and sodium thiosulphate were used for determination of process water hardness.
Residual chlorine content of the process water was done by using 10% potassium iodide
solution, 1% starch solution and distilled water.
C. Apparatus
Equipments used for analysis of input materials and effluents were beaker 100ml, glass rod,
200ml and 250 ml volumetric flasks, thermostatically controlled water bath, heating appliance,
refractometer, filter paper, vacuum pump, crucible, spectro photometer and pH meter.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 22
 Optimization of Molasses Fermentation to Reduce Effluent

3.1.2. Methods
3.1.2.1. Compositional Analysis of Molasses
I.Fermentable Sugar content of Molasses ( ICUMSA, GS 4/3-7,1995)
Fermentable sugar were determined by deducting un fermentable sugar from total reducing
sugar (TRS) of molasses. Total reducing sugar of molasses can be determined by using Lane
and Eynon Methods.
Procedure
i. Determination of Total Reducing Sugar
Total reducing sugar of molasses can be determined by using Lane and Eynon Methods.
ii. Un fermentable Sugar Determination
Take 50 gram molasses add 20 gram dry yeast then make up volume to 500ml with water and
keep for 24 hours at 32 degree centigrade, then after fermentation is complete centrifuge and
titrate the solution with 10ml Fehling solution (5ml Fehling A+5ml Fehling B) and 15ml water
using methylene blue (0.5%in water) as indicator. Add solution through burette while Fehling
solution in boiling. A brick red color appears at the end point.
Unfermentable sugar = Fehling factor x dilution (10) x 100 (3.1)
Titration value in ml
iii. Fermentable Sugar determination
Fermentable Sugar = Total Reducing Sugar – un fermentable Sugar (3.2)
II. Brix Analysis
Analysis of brix are done by using ICUMSA GS4-13 (1994).
III. Determination of Volatile Acid in Molasses
Take 50 gram molasses sample, then dilute to 500 ml with distilled water, then take 100ml of
this solution and add 5ml conc. H 2 SO 4 and 100ml water then collect 100ml distillate and titrate
with 0.1N NaOH solution.
Volatile acids in PPM = 60000 x 100 x 0.1N x 10 (dilution x volume of 0.1N NaOH used
70x100 (ML of sample taken for distillation (3.3)
IV. Analysis of sulfated ash in molasses
Sulfated ash % of molasses is determined by ICUMSA method GS1/3/4/7/8-11,1996.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 23
 Optimization of Molasses Fermentation to Reduce Effluent

V. Determination of final molasses carmel content

Analysis of carmel content will be done by using ICUMSA,GS2/3-9, 1996.
VI. Calcium content of molasses( ICUMSA,1995 GS8/2/3/4-9)
Determination of calcium oxide content will be done by using ICUMSA,1995 GS8/2/3/4-9.
VII. Phosphate content of molasses as P 2 O 5
Determination of phosphate in molasses are done by using ICUMSA,1996 GS7-15.
3.1.2.2. Physico chemical analysis of process water
A. Analysis of hardness as CaCO 3
The method implemented to test the hardness of water is WHO 26R.1,1999.
B. Determination of residual chlorine in process water(AOAC- Method)
AOAC method has been used for testing of chlorine content in the process water.
3.2. Optimization of Nutrients
3.2.1. Materials
A. Samples
The samples used for analysis of nutrient optimization was Molasses taken from bulk
molasses storage tank, process water, Dap, Urea and concentrated sulfuric acid.
B. Apparatus
Apparatus used for this analysis were volumetric flasks having 1 liter capacity, pH meter, 15
liter capacity tank, refracto meter, Microscope with heamo cytometer and Ebuillo meter.
C. Reagents
Fehling solution, methylene blue and 10% H 2 SO 4 for determination of residual sugar and
yeast count.
3.2.2. Design for Optimization of Nutrients
Response surface methodology (RSM) is implemented as it is a collection of mathematical and
statistical techniques useful for the modeling and analysis of problems in which a response of
interest is influenced by several variables and the objective is to optimize the response. The
main advantage of RSM is the reduced number of experimental runs needed to provide
Sufficient information for statistically acceptable results (Montgomery 2001). In this portion of
study DAP and Urea input as nutrient were taken as a factor to be studied at three levels and
alcohol% wash, yeast cell biomass and residual sugar% of fermented wash were taken as a

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 24
 Optimization of Molasses Fermentation to Reduce Effluent

response variable. The factor levels are decided based on practical knowledge and the
concentration of nutrient to be examined were as follows:
• Urea: - 20.84, 41.67 and 62.51 gram/m3 of Fermented wash (Beer).
• DAP:- 10.42, 20.84 and 31.26 gram/m3 of wash(Beer).
Table 3-1 Central Composite design
Factors Response Variable
Std Run Urea DAP Alcohol% Cell mass Residual
(mg/liter) (mg/liter) fermented count of Sugar % of
wash fermented fermented
wash wash
13 1 41.67 20.84 * * *
12 2 41.67 20.84 * * *
10 3 41.67 20.84 * * *
4 4 62.51 31.26 * * *
11 5 41.67 20.84 * * *
2 6 62.51 10.42 * * *
1 7 20.84 10.42 * * *
3 8 20.84 31.26 * * *
9 9 41.67 20.84 * * *
8 10 41.67 35.58 * * *
7 11 41.67 6.1 * * *
6 12 71.14 20.84 * * *
5 13 12.21 20.84 * * *

3.2.3. Methods
A. Determining the effect of nutrients on fermentation
The method used for determining the effect of nutrients application is annexed in appendix-A1.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 25
 Optimization of Molasses Fermentation to Reduce Effluent

3.3. Optimization of effluent re-circulation to fermenters

3.3.1. Materials
A. Samples
The samples used for analyzing the effect of re-circulating effluents to fermenters were
Molasses taken from bulk molasses storage tank, process water, DAP, Urea and concentrated
sulfuric acid. In addition effluents such as spent wash (Vinasse), spent lees and process
condensate.
B. Apparatus
Apparatus used for this analysis were lab scale fermentation vessels having 10 liter capacity,
pH meter, refracto meter, microscope with heamocytometer and Ebuillo meter.
C. Reagents
Fehling solution, methylene blue and 10% H 2 SO 4 for determination of residual sugar and yeast
count. 0.01 Molar EDTA solution, ammonium buffer solution and sodium thiosulphate will be
used for determination of effluents hardness.
3.3.2. Design for optimizing recirculation of effluent to fermentors
Response surface methodology (RSM) is implemented to identify the optimum recirculation
flow rate of distillery effluents such as spent wash, spent lees and process condensate to
fermentors. Three factors at three levels are employed, the factors are volume of spent wash,
spent lees and process condensate. The levels were determined based on the study conducted
by Praj and Case Tech consultancy private limited company. Praj recommends to re-circulate
about 50% of spent wash generated to fermentors and Case Tech also recommends to use both
spent lees and process condensate generated from rectifier column and spent wash
concentrating evaporators to fermentor without any treatment to dilute molasses for the
purpose of fermentation. For this specific thesis the factor levels are selected as 0,25 and 50%
for spent wash, 0,50 and 100% recirculation for both process condensate and spent lees. The
trial will be conducted on optimum sugar concentration level of 15 brix (current operating
condition of MSF distillery) and at different fermentable sugar concentration such as 15,20 and
25 brix for the purpose of identifying future operating condition. Residual sugar concentration,
yeast cell mass and alcohol% fermented wash are considered to be response variable.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 26
 Optimization of Molasses Fermentation to Reduce Effluent

Table 3.2 the Box Behnken experimental design for the three variables with three levels
employed for optimization of distillery effluent recirculation at 15 brix FW
Std Run Factors Response Variable
% Spent % Spent % Process Alcohol% Yeast count Residual
wash lees condensate fermented sugar %
wash of wash
8 1 50 50 100 * * *
17 2 25 50 50 * * *
10 3 25 100 0 * * *
11 4 25 0 100 * * *
15 5 25 50 50 * * *
4 6 50 100 50 * * *
5 7 0 50 0 * * *
12 8 25 100 100 * * *
2 9 50 0 50 * * *
16 10 25 50 50 * * *
14 11 25 50 50 * * *
1 12 0 0 50 * * *
6 13 50 50 0 * * *
13 14 25 50 50 * * *
7 15 0 50 100 * * *
3 16 0 100 50 * * *
9 17 25 0 0 * * *

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 27
 Optimization of Molasses Fermentation to Reduce Effluent

Table 3.3: the Box Behnken experimental design employed for optimization of distillery
effluent re-circulation at different sugar concentration
Std Run Factors Response Variable
%Concent % % % Process Alcohol% Cell mass Residual
ration of Spent Spent condensate fermented count of sugar %
sugar wash lees wash fermented fermented
wash wash
16 1 11.5 50 100 50 * * *
13 2 11.5 0 0 50 * * *
15 3 11.5 0 100 50 * * *
28 4 11.5 25 50 50 * * *
19 5 8.6 25 100 50 * * *
11 6 8.6 25 50 100 * * *
27 7 11.5 25 50 50 * * *
4 8 14.3 50 50 50 * * *
1 9 8.6 0 50 50 * * *
8 10 11.5 25 100 100 * * *
20 11 14.3 25 100 50 * * *
25 12 11.5 25 50 50 * * *
18 13 14.3 25 0 50 * * *
24 14 11.5 50 50 100 * * *
14 15 11.5 50 0 50 * * *
5 16 11.5 25 0 0 * * *
2 17 14.3 0 50 50 * * *
29 18 11.5 25 50 50 * * *
26 19 11.5 25 50 50 * * *
10 20 14.3 25 50 0 * * *
3 21 8.6 50 50 50 * * *
9 22 8.6 25 50 0 * * *
6 23 11.5 25 100 0 * * *

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 28
 Optimization of Molasses Fermentation to Reduce Effluent

Table 3.3 continued

23 24 11.5 0 50 100 * * *

21 25 11.5 0 50 0 * * *

12 26 14.3 25 50 100 * * *
22 27 11.5 50 50 0 * * *
7 28 11.5 25 0 100 * * *
17 29 8.6 25 0 50 * * *

3.3.3. Methods
I. Characterization of distillery Effluent
A. Determination of pH of Effluents
pH of effluent will be determined on using ICUMSA 1994 GS1/2/3/4/7/8-23 method.
B. Determination of effluent Brix (ICUMSA 1994 GS4-13)
Analysis of effluent brix is done by ICUMSA 1994 GS4-13.
II. Determining the Effect of recirculation of Distillery Effluent to Fermentors
The method used for determining the effect of re-circulating effluent to fermenter is annexed
in appendix-A2.
3.4. Effect of Yeast Recycling on Bulk Fermentation Process
3.4.1. Materials
A. Samples
Samples of fermented wash from wash holding tank and fermenter number -4 and yeast cream
from yeast separator outlet line were taken and distilled water was used to dilute wash and
cream.
B. Apparatus
Microscope with hemocyto meter to count yeast cell, 100ml measuring cylinder and 200 ml
flask with stopper.
C. Reagents
Methylene blue indicator is used to identify dead and live cell.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 29
 Optimization of Molasses Fermentation to Reduce Effluent

3.4.2. Methods
a. Take samples of fermented wash from fermentor – 4 outlet line, de yeasted wash from clear
wash holding tank and yeast cream from cream outlet line of yeast separator.
b.Take 1 ml wash and cream liquor separately in 100 ml measuring cylinder and add 1 ml
methyline blue solution. Then makeup volume to 100 ml by distilled water mix well.
c. Take hemocyto meter slide put a drop of diluted yeast culture solution on the
Hemocyto meter and cover with slip.
d.The viable yeast cells remains transparent and dead cells becomes dark blue while seen
under microscope.
3.5. Determining the effects of recirculating distillery effluent to food safety
and quality of rectified spirit alcohol
3.5.1. Materials
A. Samples
Samples used for this experiment were fermented wash prepared at laboratory by using
optimum recirculation of effluents and by using only process water.
B. Apparatus
Apparatus used are 250, 500 ml conical flasks, burette, water bath, 25 ml colourless glass
cylinder, silica crucible.
C. Reagents
Reagents used are 0.05N aqueous sodium bisulfate and aqueous iodine solution for
determination of Aldehyde content in sprit. 0.5N KOH solution, H 2 SO 4 and phenolphthalein
indicator to determine Ester. 0.8% aqueous salicylic aldehyde and concentrated H 2 SO 4 for
Fusel oil in sprit. Furfural in sprit was determined by using aniline and glacial acetic acid.
Potassium permanganate, oxalic acid, phosphoric acid and sulfuric acid were used to determine
Methyl alcohol. Copper in rectified sprit was analyzed on using sulfuric acid, aqua-ragia
solution, HCL, Potassium Ferro cyanide, NH 4 Cl, NH 4 OH and acetic acid.
3.5.2. Methods
The methods followed to test the alcohol produced are annexed in appendix –E.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 30
 Optimization of Molasses Fermentation to Reduce Effluent

4. Results and Discussion

4.1. Characterization of Input Materials
Input materials such as molasses, spent wash(vinasse), spent lees, process condensate and
process water samples are analyzed for different parameters and the results were correlated
with optimization of effluent generation.
4.1.1.Characterization of Molasses Input
Analysis of molasses for fermentible sugar, soluble solid content(brix), phosphorous, nitrogen,
volatile acid and other parameters were also done for 45 days by taking from bulk tanks daily.
Besides,composit samples were taken from molasses bulk tank which has a capacity to hold
molasses for three months. Analysis were conducted in South Africa at SMRI for calcium,
phosphorous, potassium, sodium, total nitrogen and at Novozymes laboratories in Denmark for
starch, NIR ash, Brix , NIR dry solids, NIR fructose, NIR glucose, Pol, NIR sucrose and
protein content. The results are presented in table 4.1.
The results indicate that almost all analysis results of molasses are in the required ranges
except calcium salts and sulphated ash which are beyound the required for fermentation and
distillation operation as ash content beyuond 10% in molasses reduces the activity of yeasts
and increases sludge formation in fermentor there by reducing the effective volume of
fermentors. Besides, the sludge from the wash will adhere on the nozzles of yeast cream
centrifugal separator machines and creates a huge problem on the performance of fermentors
alcohol production. Moreover, as it precipitate on heating surfaces of distillation trays as well
as falling film evaporators it will increases effluent generation by reducing heat transfer
coffecients of these units. Volatile acid content was slightly higher even if it is below the
standard different causes must be assessed on molasses handling areas. volatile acid if
exceeds 5000ppm will reduce rate of fermentation by 30-40% and if it exceeds 7000 ppm will
kill the yeast cells on reducing viablity of the cells by 40 – 50%( Jyant Godbole, Praj. 2002) .

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 31
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.1. Molasses analysis results

S/N Type of analysis Unit Results Standard
1 Total reducing sugar % 50.79 44-55
2 Un fermentible sugar % 3.13 <5
3 Fermentible sugar % 47.81 40-55
4 Calcium as (CaO) ppm 22171 < 16500
5 Phosphate as P 2 O 5 ppm 208 NA
6 Brix % 81.98 83-91
7 Volatile acidity ppm 4750 < 5000
8 Carmel at 375nm Absorbance 0.215 < 0.3
9 Calcium mg/kg 2600 1800

10 Phosphorous “ ” 140 <20000

11 Potassium “ “ 12000 <50000
12 Sodium “ “ 230 <4000
13 Total nitrogen % 0.12 0.9
14 Starch mg/kg 3060 NA
15 NIR ash % 17.73 < 15
16 NIR brix Deg-brix 85.27 83-91
17 NIR dry solids % 82.33 80-90
18 NIR fructose % 8.94 9.4
19 NIR glucose % 6.5 6.8
20 NIR pol % 29.4 28
21 NIR sucrose % 32.88 30

22 Protein% m/m % 2.59 2-6

4.1.2.Characterization of Process water

Samples of process condensate used for fermentation were analyzed for 30 days and results
are presented in table 4.2 below and the characteristics of water used at Metahara is found
suitable for fermentation of molasses as well as distillation process as its hardness is below
200ppm.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 32
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.2. analysis results of process water

S/N Types of analysis Unit Results
1 Hardness ppm 102
2 Ph pH 6.98
3 Free chlorine before activated carbon ppm 0.5-1.0
4 Free chlorine after activated carbon ppm Nil

4.2. Optimization of Nutrients

Fermented wash with brix of 15, 20 and 25 % was prepared and poured into 2 litres capacity
bottles (flasks) and to each flask nutrient such as Urea and DAP are provided as per the setted
recipe. Then fermentation was conducted for 48 hours at room temprature and the following
data were generated.
4.2.1. Optimization of nutrient at 15% brix
Table 4.3. Optimization results at 15% brix( 8.67% fermentable sugar) fermented wash
Factors Response Variable
Std Run Urea(gm/liter) DAP(gm/liter) Alcohol% Residual Cell mass
Sugar% (number of
cell/ml)
13 1 41.67 20.84 5.25 0.83 225,000,000
12 2 41.67 20.84 5.59 0.8 225,000,000
10 3 41.67 20.84 5.16 0.8 200,000,000
4 4 62.51 31.26 4.75 0.8 275,000,000
11 5 41.67 20.84 4.53 1.1 240,000,000
2 6 62.51 10.42 4.00 1.4 250,000,000
1 7 20.84 10.42 4.20 1.51 175,000,000
3 8 20.84 31.26 4.63 1.29 200,000,000
9 9 41.67 20.84 4.82 0.91 215,000,000
8 10 41.67 35.58 4.93 1.0 295,000,000
7 11 41.67 6.1 4.06 1.43 275,000,000
6 12 71.14 20.84 4.02 1.32 300,000,000
5 13 12.21 20.84 4.13 1.48 200,000,000

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 33
 Optimization of Molasses Fermentation to Reduce Effluent

4.2.1.1. Data analysis for optimization of nutrient at 8.67 sugar

concentration

Data were modeled by multiple regression analysis and the statistical significance of the terms
was examined by analysis of variance for each response. The statistical analysis of the data and
three dimensional plotting were performed using design expert software (6.0.8, 2002). The
adequacy of regression model was checked by R2, Adj R2, Pred R2, Adeq precision and F-test
(montgomery 2001). The significance of F value was judged At 95% confidence level. The
regression coefficients were then used to make statistical calculation to generate three-
dimensional plots from the regression model. The degree of relationship between the variables
was also checked by using correlation matrix of the Factors and response variables.
I. Data analysis for response-1 alcohol% fermented wash
Quadratic model is suggested by the design program for this response to test for its adequacy
and to describe its variation with independent variables. From ANOVA test in Table 4.6, the
model F-value of 15363.10 implies the model is significant. There is only a 0.01% chance that
a "model F-value" this large could occur due to noise.
Table 4.4 ANOVA test for fermented wash alcohol
Source Sum of DF Mean F-Value Prob >F
Squares Square
Model 2.55 5 0.51 4.98 0.029 Significant
A 0.51 1 6.930E-003 0.068 0.8021
B 0.062 1 0.51 5.03 0.0598
A2 0.82 1 1.48 14.46 0.0067

B2 0.27 1 0.35 3.42 0.1070

AB 8.100E-003 1 0.026 0.25 0.6322
Residual 1.45 7 0.1
Lack of Fit 0.33 4 0.032 0.17 0.9428 Not
Significant
Pure Error 0.59 3 0.20
Cor Total 3.27 12

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 34
 Optimization of Molasses Fermentation to Reduce Effluent

Values of "prob > F" less than 0.0500 indicate model terms are significant. In this case A2 are
significant model terms. The "lack of fit f-value" of 0.17 implies the lack of fit is not
significant relative to the pure error. There is a 94.28% chance that a "lack of fit f-value" this
large could occur due to noise. Non-significant lack of fit is good -- we want the model to fit.
Table 4.5 Post ANOVA statistics for fermented wash alcohol
Std. dev. 0.45 R-Squared 0.78
Mean 4.62 AdjR-Squared 0.62
C.V. 9.84 PredR-Squared 0.58
PRESS 4.09 Adeq.Precision 5.06

The "Pred R-squared" of 0.5764 is in reasonable agreement with the "Adj R-squared" of
0.6241. "Adeq precision" measures the signal to noise ratio. A ratio greater than 4 is desirable.
The ratio of 5.06 indicates an adequate signal. This model can be used to navigate the design
space (Montgomery 2001).

II. Data analysis for response-2 cell mass count of fermented wash
Linear model is suggested by the design program for this response. All statistical analysis
including ANOVA test, Post ANOVA statistics, lack of fit test, normal plot of residuals etc are
done for the cell mass data. All the tests indicated that the model is statistically acceptable.
From ANOVA test, the model F-value was 6.82 which implies that the model is significant.
Values of "Prob > F" was less than 0.0500 indicating that model terms A and B are significant.
The "lack of fit F-value" of 0.94 implies the lack of fit is not significant relative to the pure
error. Non-significant lack of fit is good -- we want the model to fit.
Table 4.6 Post ANOVA statistics for fermented wash cell mass
Std. dev. 3.887E+007 R-Squared 0.58
Mean 2.365E+008 AdjR-Squared 0.49
C.V. 16.43 PredR-Squared 0.47
PRESS 2.595E+016 Adeq.Precision 7.54

The "Pred R-squared" of 0.47 is in resonable agreement with the "Adj R-squared" of 0.49 as
one might normally expect."adeq precision" measures the signal to noise ratio. A ratio greater
than 4 is desirable. The ratio of 7.54 indicates an adequate signal. This model can be used to
navigate the design space.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 35
 Optimization of Molasses Fermentation to Reduce Effluent

III. Data analysis for response-3 residual sugar of fermented wash

Quadratic model is suggested by the design program for this response. All statistical analysis
including ANOVA test, Post ANOVA statistics, lack of fit test, normal plot of residuals etc are
done for the Residual sugar analysis data. All the tests indicated that the model is statistically
acceptable. From ANOVA test, the model F-value was 6.82 which implies that the model is
significant.
Values of "Prob > F" was less than 0.0500 indicating that model terms B, A2, B2 are
significant. The "lack of fit F-value" of 0.53 implies the lack of fit is not significant relative to
the pure error. Non-significant lack of fit is good -- we want the model to fit.
Table 4.7 Post ANOVA Statistics for Fermented wash Residual Sugar
Std. dev. 0.28 R-Squared 0.89
Mean 1.13 AdjR-Squared 0.81
C.V. 25.19 PredR-Squared 0.87
PRESS 1.44 Adeq.Precision 7.61

The "Pred R-squared" of 0.87 is in resonable agreement with the "Adj R-squared" of 0.81 as
one might normally expect."adeq precision" measures the signal to noise ratio. A ratio greater
than 4 is desirable. The ratio of 7.61 indicates an adequate signal. This model can be used to
navigate the design space.
4.2.1.2. Correlation matrics for nutrient optimization at 8.67 sugar
concentration
The correlation matrix is one of the most common and most useful statistics. The
correlationcoefficient measures the strength of a linear relationship between the variables. The
correlation coefficient is always between -1 and +1. The closer the correlation is to +/-1, the
closer to a perfect linear relationship. The correlation matrix for nutrient optimization at 8.67%
fermentable sugar concentration is described in the table 4.8 and as it is seen clearly almost all
the factor- to-factor correlation is zero for the factors- to - factors correlation, off-diagonal
values close to zero are better.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 36
 Optimization of Molasses Fermentation to Reduce Effluent

Table-4.8: Correlation matrix of factors [Pearson's R]

A B A2 B2 AB

A 1.000

B 0.000 1.000

A2 0.000 0.036 1.000

B2 -0.000 -0.266 -0.051 1.000

AB -0.000 -0.000 -0.000 0.000 1.000

4.2.1.3. Equation for nutrient optimization at 8.67 sugar concentration

Model equations are given in terms of coded factors and actual factors. Coded factors indicate
when the minimum and maximum values of the factors are represented by -1 and +1
respectively instead of their actual values.
I. Alcohol % fermented wash
Equation interms of coded factors:
Alcohol Content = +4.53 +0.25 * A - 0.088 * B +0.34 * A2
-0.20 * B2 +0.045 * A * B (4.1)
Equation in terms of Actual factors:
Alcohol Content = +4.61 - 0.05 * Urea +0.076* DAP +791873E-004 * Urea2 –
1.81E-003 * DAP2 – 2.07E-004 * Urea * DAP (4.2)
II. Cell Mass Count in Fermented Wash
Final equation in terms of coded factors:
Yeast Cell Count = +2.36E+008 +2.13E+ 007 * A -6.94E+006* B (4.3)
Final equation in terms of actual factors:
Yeast cell count = +2.08E+008 +1.02E+006* Urea–6.66E+005*DAP (4.4)
III. Residual Sugar Percent of Fermented Wash
Final equation in terms of coded factors:
Residual Sugar = +1.13-0.14*A + 0.051*B (4.5)

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 37
 Optimization of Molasses Fermentation to Reduce Effluent

Final equation in terms of actual factors:

Residual Sugar = +1.30 - 6.48E-003* Urea – 4.88E-003 * DAP (4.6)
Where A and B are coded value for urea and dap concetration.
Diagnostic test for the responses
All diagnostic plots are also tested for all responses for adequacy of the models (normal plot of
residuals, residuals vs predicted value, residuals vs factor, box cox plot, studentized residuals,
leverage, etc.). For example fig 4.1 shows how precisely the alcohol% is modeled, because all
the points line up nicely and the deviation of points for alcohol% fermented wash from
normality is insignificant. Similar results were observed for the remaining responses (Cell mass
and Residual sugar content).

DESIGN-EXPERT Plot Normal Plot of Residuals

Alcohol Content

99

95
90
80
70
50
30
20
10
5

-1.83 -0.91 0.01 0.94 1.86

X: Studentized Residuals
Y: Normal % Probability

Fig 4.1 Normal plot of residual for alcohol % fermented wash

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 38
 Optimization of Molasses Fermentation to Reduce Effluent

4.2.1.4. Response surface plot analysis for nutrient optimization at 8.67 %

sugar concentration
The effect of different treatment parameters on alcohol% fermented wash is given in Table 4.5.
To aid visualization of variation in responses with respect to processing variables, series of
three dimensional Response surfaces were drawn for each response using design expert
software 6.0-8.0.

I. Response-1 alcohol% fermented wash

The effect of DAP and Urea on alcohol% fermented wash

5.149
4.915
Alcohol 4.681
Content 4.447
4.214

31.26
62.51
26.05 52.09
15.63 41.67
B: DAP
10.42 31.26 A: Urea

Fig 4.2 Response surface plot for fermented wash alcohol content

From the response surface graph it is clearly seen that alcohol content of fermented wash
increases as the DAP and urea application up to some level and after the optimum application
of DAP production of alcohol goes down due to additional increment in calcium content of the
wash as DAP contains calcium in it. It is well known that calcium salt will affect the activity of
yeast (Heggart et al,1999).

II. Response -2 Yeast cell mass count

From the response surface graph(Fig 4.3) it is clearly seen that Yeast cell count of fermented
wash increases with the urea application but as the concentration of Dap increases from 10.42
mg/l to 31.25 mg/l the cell mass growth shows a declining trend.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 39
 Optimization of Molasses Fermentation to Reduce Effluent

The effect of DAP and Urea concentration on Yeast cell biomass count

DESIGN-EXPERT Plot
X = A: Urea
Yeast Cell Count
Y = B: DAP 2.65E+008
2.51E+008
2.37E+008
Yeast Cell 2.22E+008
2.08E+008
Count

31.26
26.05 62.51
52.09
20.84 41.67
B: DAP 15.63 31.26 A: Urea
10.42 20.84

Fig 4.3- Response surface plot for yeast cell biomass in fermented wash
III. Response-3 Residual Sugar
The effect of DAP and Urea concentration on the residual sugar content of fermented wash

DESIGN-EXPERT Plot

Y = B: DAP
1.31
X = A: Urea
Residual Sugar 1.22
1.13
Residual Sugar
1.04
0.94

31.26
26.05 62.51
20.84 52.09
15.63 41.67
B: DAP 31.26
10.42 20.84
A: Urea

Fig 4.4 Response Surface plot for residual sugar content of fermented wash
Response surface graphs on Fig 4.3 and 4.4 shows that reduction of residual sugar and increase
in cell mass content of fermented wash has been observed as the Urea application increases
from 20.84-62.51mg/l but the reverse condition will be prevail while increasing DAP
concentration increases from 20.84 to 31.26mg/l. This is mainly because of the molasses

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 40
 Optimization of Molasses Fermentation to Reduce Effluent

quality of MSF which contains high calcium content which has the tendency to form
precipitate on reacting with the phosphate of DAP. The precipitate will trap the yeast cell mass
and causes settling with sludge without contributing much to alcohol production.
4.2.1.5. Optimization solution for nutrients at 15ºBrix fermented wash

The optimum parameter values for fermented wash; alcohol%, yeast cell mass count and
residual sugar were predicted by partially differentiating equations (4.2), (4.4) and (4.6) with
respect to each parameter and equating to zero (montgomery 2001). The optimization was to
get maximum alcohol % fermented wash, optimum cell mass and minimum residual sugar as
much as possible. The first response (Alcohol% fermented wash) was considered as an
important response and more weight is given to it since it is the major determining factor of
this optimization. The predicted values were Urea 62.51 mg/l, DAP 14.51 mg/l, yeast cell mass
count 2.62045E+008, Alcohol% fermented wash 5.136 and residual sugar content of 0.9626%.
In order to verify this prediction experiment was conducted at 15°Bx fermented wash by
applying 62.51 Urea and 14.51 DAP mg/l and 5.1% alcohol , 2.6804E+008 yeast cell mass
count and 0.958% residual sugar content in fermented wash had been obtained which are in
good agreement with the predicted values. Besides, equations 4.2 and 4.6 were differentiated
twice inorder to know whether the predicted values of Urea and DAP will give minimum
residual sugar and maximum alcohol and the value at predicted points are – 2.6 and 0 for
equation (4.2) and (4.6) respectively. Since -2.6 is below zero, the predicted alcohol %
fermented wash is maximum and residual sugar is minimum as it is greater or equal to zero.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 41
 Optimization of Molasses Fermentation to Reduce Effluent

4.2.2. Optimization of nutrient at 20 % brix (fermentable sugar

concentration of 11.5%)
Table 4.9. Optimization Results at 20 % brix fermented wash
Factors Response Variable
Std Run Urea(gm/liter) Dap(gm/liter) Alcohol% Residual Cell mass
Sugar% (cfu)

13 1 41.67 20.84 5.08 0.753 150,000,000

12 2 41.67 20.84 6.01 0.63 150,000,000
10 3 41.67 20.84 5.95 0.52 300,000,000
4 4 62.51 31.26 5.9 0.62 325,000,000
11 5 41.67 20.84 5.66 0.6 350,000,000
2 6 62.51 10.42 6.18 0.43 375,000,000
1 7 20.84 10.42 5.86 0.55 250,000,000
3 8 20.84 31.26 6.28 0.53 350,000,000
9 9 41.67 20.84 5.86 0.62 375,000,000
8 10 41.67 35.58 6.21 0.693 400,000,000
7 11 41.67 6.1 5.47 0.63 325,000,000
6 12 71.14 20.84 5.95 0.61 295,000,000
5 13 12.21 20.84 6 0.58 325,000,000

4.2.2.1. Data analysis for optimization of nutrient at 11.5% sugar

concentration
I. Data analysis for response-1 alcohol% fermented wash

2Fi model is suggested by the design program for this response. All statistical analysis
including ANOVA test, Post ANOVA statistics, lack of fit test, normal plot of residuals etc are
done for the Alcohol% fermented wash data. All the tests indicated that the model is
statistically acceptable. Values of "Prob > F" was less than 0.0500 indicating that model terms
are significant but in this case there are no significant model terms. The "lack of fit F-value" of
0.98 implies the lack of fit is not significant relative to the pure error. Non-significant lack of
fit is good -- we want the model to fit.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 42
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.10 Post ANOVA statistics for fermented wash alcohol

Std. dev. 0.29 R-Squared 0.40

Mean 5.88 AdjR-Squared 0.38
C.V. 4.94 PredR-Squared 0.17
PRESS 1.86 Adeq.Precision 4.83

A deq precision measures the signal to noise ratio. A ratio greater than 4 is desirable. Ratio of
4.830 indicates an adequate signal. This model can be used to navigate the design space.

II. Data analysis for response-2 cell mass count of fermented wash
2FI- model is suggested by the design program for this response. All statistical analysis
including ANOVA test, Post ANOVA statistics, lack of fit test, normal plot of residuals etc are
done for the cell mass data. All the tests indicated that the model is statistically acceptable.
The "lack of fit F-value" of 0.45 implies the lack of fit is not significant relative to the pure
error. Non-significant lack of fit is good -- we want the model to fit.
Table 4.11 Post ANOVA statistics for fermented wash cell mass
Std. dev. 5.178E+007 R-Squared 0.68
Mean 3.054E+008 AdjR-Squared 0.57
C.V. 16.93 PredR-Squared 0.55
PRESS 6.845E+016 Adeq.Precision 7.88
A ‘’Pred R-Squared’’ is as close to adj R-square is a better predictor the model.

III. Data analysis for response-3 residual sugar content of fermented wash
Linear model is suggested by the design program for this response. All statistical analysis
including ANOVA test, Post ANOVA statistics, lack of fit test, normal plot of residuals etc are
done for the residual sugar analysis data. All the tests indicated that the model is statistically
acceptable. Values of "Prob > F" was less than 0.0500 indicating that model terms are
significant but in this case there are no significant terms. The "lack of fit F-value" of 2.43
implies the lack of fit is not significant relative to the pure error. Non-significant lack of fit is
good -- we want the model to fit.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 43
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.12 Post ANOVA Statistics for Fermented wash residual sugar
Std. dev. 0.057 R-Squared 0.59
Mean 0.60 AdjR-Squared 0.50
C.V. 9.49 PredR-Squared 049
PRESS 0.064 Adeq.Precision 7.77

A pred R-Squared is as close to AdjR-Square the model is good. Besides, adeq. precision
measures signal to noise ratio since it is 7.77 which is greater than 4. Hence, the model is
desirable to navigate the design space.
4.2.2.2. Correlation matrics for nutrient optimization at 11.5% sugar
concentration
The correlation matrix is one of the most common and most useful statistics. The correlation
coefficient measures the strength of a linear relationship between the variables. The correlation
coefficient is always between -1 and +1. The closer the correlation is to +/-1, the closer to a
perfect linear relationship.
Table-4.13: Correlation Matrix of Factors [Pearson's R]
A B A2 B2 AB

A 1.000

B -0.000 1.000

A2 0.000 0.000 1.000

B2 -0.000 -0.000 -0.131 1.000

AB 0.000 -0.000 -0.000 -0.000 1.000

Almost all the factor- to-factor correlation is zero for the factors- to - factors correlation, off-
diagonal values close to zero are better.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 44
 Optimization of Molasses Fermentation to Reduce Effluent

4.2.2.3. Regression equation for nutrient optimization at 11.5% sugar

concentration
Model equations are given in terms of coded factors and actual factors. Coded factors indicate
when the minimum and maximum values of the factors are represented by -1 and +1
respectively instead of their actual values.
I. Alcohol % Fermented Wash
For this case both final equation interms of coded factors :
Alcohol % fermented wash = + 5.88 - 0.11*A - 0.11* B- 0.28 * A * B (4.7)
Final equation interms of actual factors :
Alcohol % Fermented wash = + 5.172 + 0.022*Urea+ 0.044 *DAP
-1.301E-003 * Urea * DAP (4.8)

II. Cell Mass Count in Fermented Wash

Since the model suggested by the experimental method for this specific design was mean
modal, the final Equation In Terms of Coded Factors :
Yeast Cell Count = +3.05E+008 - 4.54E + 007*A - 4.96E+007* B- 6.13E+007 *A*B (4.9)
Yeast Cell Count = + 2.50E+008 + 3.70E+006 * Urea + 7.0E+006 * DAP
-2.82E+005 * Urea * DAP (4.10)
III. Residual Sugar Percent of Fermented Wash
Final Equation In Terms of Coded Factors:
Residual Sugar = + 0.60 + 0.075 *A + 9.91E-003 * B (4.11)
Final Equation In Terms of Actual Factors:
Residual Sugar = + 0.428 + 3.593E-003 * Urea + 9.512E-004 * DAP (4.12)
Where A is coded factor for Urea and B is for DAP
Diagnostic Test for the Responses
All diagnostic plots are also tested for all responses for adequacy of the models (normal plot of
residuals, residuals vs. predicted value, residuals vs. factor, box cox plot, studentized residuals,
leverage, etc).

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 45
 Optimization of Molasses Fermentation to Reduce Effluent

4.2.2.4. Response surface plot analysis for nutrient optimization at 11.5 %

sugar concentration
I. Response-1 Alcohol% fermented Wash

DESIGN-EXPERT Plot
X = A: Urea
Alcohol Content 6.16
Y = B: Dap 5.97
5.77
5.58
Alcohol Content 5.38

31.26
26.05 62.51
52.09
20.84 41.67
B: Dap 15.63 31.26 A: Urea
10.42 20.84

Fig 4.5 Response Surface plot for alcohol content of fermented wash

From the Response surface graph of fermented wash it clearly shows that increasing Urea or
DAP has an increasing effect on alcohol% for fermentable sugar concentration of 11.5% and
maximum alcohol production occurred either at31.26 mg/l DAP or 62.51mg/l Urea application
rate. This is mainly due to the increase in the yeast cell mass of wash due to better availability
of nitrogen and phosphorous. But, simultaneously increasing both nutrients above the optimum
points as well as at lower application rate of these nutrients will reduce alcohol production.
Since at higher application rate availability of free nitrogen and phosphorous will increases in
the wash and creates conducive environment for bacterial proliferation which causes stuck
fermentation. On the other hand at lower nutrient supply multiplication of yeast inside
fermenter will be reduced and resulted in lower production of alcohol.

II. Response-2 Yeast cell mass count of fermented Wash

Response surface graph of fermented wash plotted on Fig 4.6 clearly shows that increasing
simultaneously Urea and Dap will reduce multiplication of yeast cell for fermentable sugar
concentration of 11.5% with the same reasons for response-1 above.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 46
 Optimization of Molasses Fermentation to Reduce Effluent

3.71E+00
3.15E+00
Yeast Cell 2.60E+00
2.04E+00
1.49E+00

31. 62.51
26.
20. 41.67
B: Dap 15.
A: Urea
10. 20.84

Fig 4.6 Response Surface plot for yeast cell mass content of fermented wash

III. Response-3 Residual Sugar % of Fermented Wash

DESIGN-EXPERT Plot
Residual Sugar
X = A: Urea
Y = B: Dap

0.78
0.67
Residual Sugar 0.56
0.45
0.34

31.26
62.51
26.05
52.09
20.84 41.67
B: Dap 15.63 31.26 A: Urea
6.10 20.84
Fig 4.7 Response Surface plot for Residual sugar content of fermented wash
At 11.5%(20degree brix) fermented wash concentration the concentration of residual sugar at
application rate of 6.10 mg/lt of DAP and 62.51mg/lt of Urea was to the lower side. But, it is

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 47
 Optimization of Molasses Fermentation to Reduce Effluent

higher when both Urea and DAP has been applied simultaneously with the same reasoning as
for the above two response variables.
4.2.2.5. Optimization solution for nutrient application for sugar
concentration of 11.5 % fermented wash
About 10 solutions were suggested by design expert 6.0-8.0 for this typical case and from
practical observation on fermentation process at Metahara distillery higher consumption of
DAP will cause higher incrustation on distillation column and heat exchanging equipments.
Hence out of this 10 combination of urea and DAP, the one with lower DAP amount and urea
concentration was preferred. As a result, the predicted result became Urea 20.84mg/lt, DAP
31.26mg/lt, Alcohol %fermented wash 6.16,Yeast cell mass count 3.63E+008, residual
sugar0.53, desirability 0.78. To verify the results, experiment was conducted at sugar
concentration of 15° brix fermented wash, Urea and DAP concentration rate of 20.84 and 31.26
mg/l respectively and the result became 6.18 alcohol% wash, 3.632E+008 yeast count and 0.52
residual sugar% had been obtained which is identical with predicted value. Besides, equations
4.8 and 4.12 were differentiated twice in order to know whether the predicted values of Urea
and DAP will give minimum residual sugar and maximum alcohol and the value at predicted
points are – 2.07 and 0 for equation (4.8) and (4.12) respectively. Since -2.07 is below zero,
alcohol % fermented wash predicted obtained at predicted value is maximum and residual
sugar is minimum as it is greater or equal to zero.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 48
 Optimization of Molasses Fermentation to Reduce Effluent

4.2.3. Optimization of Nutrient at 25 % Brix fermented wash

Table 4.14. Optimization Results at 25 % brix(fermentable sugar conc. of 14.3%

fermented wash
Factors Response Variable
Std Run Urea(gm/liter) Dap(gm/liter) Alcohol% Residual Cell mass
Sugar% (cfu)

13 1 41.67 20.84 7.08 1.6 240,000,000

12 2 41.67 20.84 6.99 1.6 240,000,000
10 3 41.67 20.84 7.31 1.5 240,000,000
4 4 62.51 31.26 7.08 1.28 295,000,000
11 5 41.67 20.84 6.88 1.6 250,000,000
2 6 62.51 10.42 7.34 1.7 240,000,000
1 7 20.84 10.42 7.53 1.2 240,000,000
3 8 20.84 31.26 7.51 1.2 240,000,000
9 9 41.67 20.84 7.43 1.2 255,000,000
8 10 41.67 35.58 7.43 0.83 300,000,000
7 11 41.67 6.1 7.43 1.1 240,000,000
6 12 71.14 20.84 7.56 0.9 270,000,000
5 13 12.21 20.84 7.51 1.1 220,000,000

4.2.3.1. Data analysis for optimization of nutrient at 14.3% sugar

concentration
I. Data analysis for response-1 alcohol% fermented wash
Linear model is suggested by the design program for this response. All statistical analysis
including ANOVA test, Post ANOVA statistics, lack of fit test, normal plot of residuals etc are
done for the Alcohol% fermented wash data. All the tests indicated that the model is
statistically acceptable. Values of "Prob > F" was less than 0.0500 indicating that model terms
are significant but in this case there are no significant model terms. The "lack of fit F-value" of

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 49
 Optimization of Molasses Fermentation to Reduce Effluent

0.33 implies the lack of fit is not significant relative to the pure error. Non-significant lack of
fit is good -- we want the model to fit.
Table 4.15. Post ANOVA Statistics for Fermented wash Alcohol%
Std. dev. 0.22 R-Squared 0.19
Mean 7.33 AdjR-Squared 0.038
C.V. 3.07 PredR-Squared 0.044
PRESS 0.76 Adeq.Precision 4.56
A pred R-Squared is as close to AdjR-Square the model is good. Moreover, adeq. precision
measures signal to noise ratio since it is 4.56 which is greater than 4. Hence, the model is
desirable to navigate the design space.

II. Data Analysis for Response-2 Yeast cell mass count of fermented Wash
2Fi- model is suggested by the design program for this response. All statistical analysis
including ANOVA test, Post ANOVA statistics, lack of fit test, normal plot of residuals etc are
done for the Residual sugar analysis data. All the tests indicated that the model is statistically
acceptable. From ANOVA test, the model F-value was 23.24 which implies that the model is
significant.
Values of "Prob > F" was less than 0.0500 indicating that model terms A,B, B2 , and AB are
significant. The "lack of fit F-value" of 2.57 implies the lack of fit is not significant relative to
the pure error. Non-significant lack of fit is good -- we want the model to fit.
Table 4.16 Post ANOVA statistics for fermented wash cell mass
Std. dev. 1.836E+007 R-Squared 0.53
Mean 2.515E+008 AdjR-Squared 0.38
C.V. 7.30 PredR-Squared 0.51
PRESS 9.865E+015 Adeq.Precision 6.56

The "Pred R-squared" of 0.53 is in resonable agreement with the "Adj R-squared" of 0.38 as
one might normally expect."adeq precision" measures the signal to noise ratio. A ratio greater
than 4 is desirable. The ratio of 6.56 indicates an adequate signal. This model can be used to
navigate the design space.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 50
 Optimization of Molasses Fermentation to Reduce Effluent

III. Response -3 residual sugar % of fermented wash

2Fi- model is suggested by the design program for this response. All statistical analysis
including ANOVA test, Post ANOVA statistics, lack of fit test, normal plot of residuals etc are
done for the Residual sugar analysis data. All the tests indicated that the model is statistically
acceptable. Values of "Prob > F" was less than 0.0500 indicating that model terms are
significant but in this case no significant terms are present. The "lack of fit F-value" of 1.00
implies the lack of fit is not significant relative to the pure error. Non-significant lack of fit is
good -- we want the model to fit.
Table 4.17 Post ANOVA Statistics for Fermented wash Residual Sugar
Std. dev. 0.24 R-Squared 0.46
Mean 1.29 AdjR-Squared 0.27
C.V. 18.64 PredR-Squared 0.23
PRESS 1.10 Adeq.Precision 5.9

Since adeq, precision of the model is 5.9 which is greater than 4. Hence it is adequate to
navigate the design space.

4.2.3.2. Correlation matrics for nutrient optimization at 14.3%(25 Brix)

sugar concentration
From correlation matrix of factors (Pearson’s R), it was clearly observed that almost all factor
to factor correlations are zero and it is good.

Table-4.18: Correlation matrix of factors [Pearson's R]

A B A2 B2 AB

A 1.000

B 0.000 1.000

A2 0.000 0.000 1.000

B2 0.000 0.000 -0.131 1.000

AB 0.000 0.000 0.000 0.000 1.000

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 51
 Optimization of Molasses Fermentation to Reduce Effluent

4.2.3.3. Equation for nutrient optimization at 14.3% sugar concentration

I. Response-1 Alcohol % Fermented Wash
Equation in terms of coded factors are given by
Alcohol% of fermented wash = +7.31+0.12*A-0.023*B (4.13)
Equation in terms of Actual Factors is given by:
Alcohol % fermented wash = +7.11+ 5.90E-003*Urea – 2.21E-003*Dap (4.14)
II. Response-2 Yeast Cell mass Count
Equations in terms of coded and Actual factors for cell mass count were:
For Coded Factors:
Yeast Cell Count = +2.52E+008 – 6.25E+005 *A +4.68E+006 * B
- 2.88E+007 *A*B (4.15)
Based on Actual Factors:
Yeast Cell Count = +3.74659E+008 -2.72978E+006 * Urea -5.96787E+006 * DAP
+ 1.32427E+005*Urea*Dap (4.16)
III. Response-3 Residual Sugar % fermented wash
Equation in terms of coded factors for residual sugar % of fermented wash is
Residual Sugar = +1.29 – 0.097*A- 0.033*B-0.30*A*B (4.17)
Equation in terms of Actual Factors is:
Residual Sugar% fermented wash = + 0.36+ 0.024*Urea + 0.054*Dap –
1.37E-003*Urea*Dap (4.18)

Diagnostic Test for the Responses

All diagnostic plots are also tested for all responses for adequacy of the models (normal plot of
residuals, residuals vs predicted value, residuals vs factor, box cox plot, studentized residuals,
leverage, etc.) and found in good agreement.

4.2.3.4. Response surface plot analysis for nutrient optimization at 14.3 %

(25 Brix) sugar concentration
I. Response-1 alcohol% fermented wash

The plot indicates that the effect of varying DAP or Urea alone in the ranges of 10.42- 31.26
for DAP and 20.84- 62.51mg/lt for urea for sugar concentration of 14.3% or 25 degree brix
fermented wash has negligible impact on alcohol% fermented wash. But, simultaneously
AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 52
 Optimization of Molasses Fermentation to Reduce Effluent

increasing both nutrients was conducive for productivity of alcohol inside fermenter. This is
mainly due to the increase in availability of nitrogen and phosphorous proportionally with high
sugar concentration and un favorable condition created by high sugar concentration for bacteria
proliferation.

DESIGN-EXPERT Plot
Alcohol Content
Y = B: Dap
X = A: Urea 7.46
7.39
Alcohol 7.31
7.24
Content 7.17

31.26
26.05 62.51
52.09
20.84 41.67
B: Dap 15.63 31.26 A: Urea
10.42 20.84

Fig 4.8 Response Surface plot for alcohol% fermented wash at 14.3% fermentable sugar

II. Response-2 yeast cell mass % fermented wash

DESIGN-EXPERT Plot
Yeast Cell Count
Y = B: Dap
X = A: Urea 2.84E+008
2.67E+008
2.51E+008
Yeast Cell 2.34E+008
Count 2.17E+008

31.26
26.05 62.51
52.09
20.84 41.67
B: Dap 15.63 31.26
10.42 20.84 A: Urea

Fig 4.9. Response surface plot for Yeast cell mass at concentration level of 14.3% (25
Brix) sugar

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 53
 Optimization of Molasses Fermentation to Reduce Effluent

As it was clearly seen on response surface plot the cell mass count shows a reducing trend
whenever separately increase the amount of DAP and Urea the level of 31.26 and 62.51mg/l .
Increasing DAP concentration alone at this sugar concentration will not favor either cell
production or alcohol production because of insufficient nutrient availability. On the other
hand increasing application of Urea alone improves availability of nitrogen alone which can
only encourages alcohol production rather than cell mass production as the quantity of
phosphorous is used for ATP production and multiplication of cell is to the lower sides. But
simultaneous increase of both DAP and urea concentration to 31.26 and 62.51 mg/l will have
huge impact on improvements of cell mass and alcohol production due to availability of good
nutrients.

III. Response-3 residual sugar % of fermented wash

DESIGN-EXPERT Plot
Residual Sugar
Y = B: Dap
X = A: Urea 1.65
1.46
Residual 1.26
1.06
Sugar 0.87

31.26
26.05 62.51
52.09
20.84 41.67
B: Dap 15.63 31.26 A: Urea
10.42 20.84

Fig 4.10 Response surface plot of residual sugar% fermented wash at 14.3% sugar

The plot indicates that varying Dap and Urea concentration in the ranges of 10.42- 31.26 for
Dap and 20.84- 62.51mg/lt for urea for sugar concentration of 14.3% or 25 degree brix
fermented wash has an increasing effect on Residual Sugar % fermented wash. Minimum
values of residual sugar is obtained as it was seen from the plot that at higher urea and DAP
application.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 54
 Optimization of Molasses Fermentation to Reduce Effluent

4.2.3.5. Optimization solution for nutrient application for sugar

concentration of 14.3 % fermented wash

About 3 solutions were suggested by design expert 6.0-8.0 for this typical case and the one
with optimum application of DAP and Urea were selected and the result became Urea 62.51
mg/lt, DAP 31.26 mg/lt, Alcohol %fermented wash 7.41,Yeast cell mass count 2.76E+008 and
Residual sugar content of 0.87. In order to verify the results experiment was conducted on
preparing 25°brix fermented wash and applying the predicted quantity of nutrient such as Urea
62.51mg/l and DAP 31.26 mg/l and the result became 7.38 alcohol% wash, yeast cell count of
2.76E+008 and residual sugar concentration of 0.88. The actual results are nearly the same
with the predicted values. In addition equations 4.14 and 4.18 are differentiated twice in order
to know whether the predicted values of Urea and DAP will give minimum residual sugar and
maximum alcohol and the value at predicted points are 0 and 2.74 for equation (4.14) and
(4.18) respectively. Since the double differentiated value is zero, alcohol % fermented wash
predicted is maximum and residual sugar is minimum as it is greater or equal to zero.

4.3. Optimization of effluent recirculation to fermentors

Optimization of effluent recirculation is conducted at three level of sugar concentration in the

fermentors such as at 15 degree brix ( fermentable sugar concentration of 8.66%) this sugar
concentration is taken from current working condition of Metahara distillery , 20 degree
brix(fermentable sugar concentration of 11.5%) and 25 degree brix (14.3% sugar
concentration) these sugar concentration is selected as the future plan of MSF to generate more
alcohol on comparatively controlling effluent generation.
4.3.1. Characterization of effluent
4.3.1.1. Compositional analysis of spent wash
Samples of spent wash was analyzed daily for 30 days to determine alcohol%, brix, calcium
oxide, phosphate contents at Metahara Sugar factory quality control laboratory. Composite
samples of vinasse is sent to South Africa SMRI laboratory for analysis of inorganic minerals
and samples of spent wash also analyzed for Calcium, Nitrogen, Potassium, Sodium and
Phosphorous and the results were described in table 4.19 below. Except the calcium and
phosphorous content of spent wash all values are in the required ranges. The calcium contents

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 55
 Optimization of Molasses Fermentation to Reduce Effluent

of vinasse is to the higher side and the phosphorous content is to the lower side. This is mainly
due to the nature of molasses which inturn caused by the type of cane crushed in Methara
which is harvested from alkaline soil. As described above returning vinasse to fermenter with
respect to calcium content might creat difficulties on fermentation station.
Table 4.19. spent wash analysis results
S/N Types of analysis Unit Results Standard
1 Calcium mg/kg 3500 <2700
2 Phosphorous mg/kg 170 325-490
3 Potassium mg/kg 15000 18000-20000
4 Sodium mg/kg 290 300-500
5 Total nitrogen % 0.12 2000-2500
6 Brix % 13.6 <22
7 pH pH 4.1 4.3-4.6
8 Alcohol % 0.11 Nil
9 Calcium as Cao Ppm 4043
10 Phosphate as P 2 O 5 Ppm 254 485-700

4.3.1.2. Compositional analysis of process condensate and spent lees

Samples of process condensate and spent lees were analyzed for brix, hardness in the form of
calcium, pH and alcohol%. The results were written in table 4.3 and it indicates that if
successfully recirculate these effluent streams to fermenters, inaddition to safeguarding the
environment from pollution it will avoide loss of alcohol and reduce acid consumption as the
pH of effluent is below 4.5.
Table 4.20. process condensate and spent lees analysis results
S/N Types of Unit Spent lees Process
analysis condensate
1 Hardness Ppm Nil Nil
2 pH pH 3.2 3.69
3 Brix % Nil Nil
4 Alcohol % 0.01 0.33

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 56
 Optimization of Molasses Fermentation to Reduce Effluent

4.3.2.Optimization at 15˚Brix sugar concentration

Table 4.21 Optimization results at 15˚brix

Std Run Factors Response Variable
% Spent % Spent % Process Alcohol% Yeast count Residual
wash lees Condensate fermented sugar %
wash of wash
8 1 50 50 100 3.86 350,000,000 3.6
17 2 25 50 50 5.16 300,000,000 1.35
10 3 25 100 0 4.71 200,000,000 2.5
11 4 25 0 100 4.52 225,000,000 2.65
15 5 25 50 50 5.23 295,000,000 1.5
4 6 50 100 50 2.93 275,000,000 2.2
5 7 0 50 0 4.74 325,000,000 1.8
12 8 25 100 100 4.74 235,000,000 1.3
2 9 50 0 50 5.26 290,000,000 1.1
16 10 25 50 50 5.28 235,000,000 1.32
14 11 25 50 50 5.3 230,000,000 1.2
1 12 0 0 50 5.23 240,000,000 1.1
6 13 50 50 0 5.82 275,000,000 1.16
13 14 25 50 50 5.69 245,000,000 1.3
7 15 0 50 100 6.37 225,000,000 0.92
3 16 0 100 50 4.92 235,000,000 1.65
9 17 25 0 0 5.3 240,000,000 1.34

4.3.2.1. Data analysis for optimization of effluent at 8.67 sugar

concentration
Data were modeled by multiple regression analysis and the statistical significance of the terms
was examined by analysis of variance for each response. The statistical analysis of the data and
three dimensional plotting were performed using design expert software (6.0.8, 2002). The
adequacy of regression model was checked by R2, Adj R2, Pred R2, Adeq precision and F-test
AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 57
 Optimization of Molasses Fermentation to Reduce Effluent

(montgomery 2001). The significance of F value was judged At 95% confidence level. The
regression coefficients were then used to make statistical calculation to generate three-
dimensional plots from the regression model. The degree of relationship between the variables
was also checked by using correlation matrix of the Factors and response variables.
I. Data analysis for response-1 alcohol% fermented wash
2FI model is suggested by the design program for this response to test for its adequacy and to
describe its variation with independent variables. From ANOVA test in table 4.17, the model
F-value of 4.89 implies the model is significant. There is only a 0.41% chance that a "model F-
value" this large could occur due to noise.
Table 4.22 ANOVA test for fermented wash alcohol
Source Sum of DF Mean F-Value Prob >F
Squares Square
Model 2.55 5 0.51 4.98 0.029 Significant
A 6.930E-003 1 6.930E-003 0.068 0.8021
B 0.51 1 0.51 5.03 0.0598
2
A 1.48 1 1.48 14.46 0.0067

B2 0.35 1 0.35 3.42 0.1070

AB 0.026 1 0.026 0.25 0.6322
Residual 0.72 7 0.1
Lack of Fit 0.13 4 0.032 0.17 0.9428 Not
Significant
Pure Error 0.59 3 0.20
Cor Total 3.27 12

Values of "prob > F" less than 0.0500 indicate model terms are significant. In this case A2 are
significant model terms. The "lack of fit f-value" of 3.75 implies the lack of fit is not
significant relative to the pure error. There is a 11.09% chance that a "lack of fit F-value" this
large could occur due to noise. Non-significant lack of fit is good -- we want the model to fit.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 58
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.23 Post ANOVA statistics for fermented wash alcohol

Std. dev. 0.34 R-Squared 0.81

Mean 5.12 AdjR-Squared 0.69
C.V. 6.55 PredR-Squared 0.28
PRESS 4.18 Adeq.Precision 10.80

"Adeq precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. The
ratio of 10.80 indicates an adequate signal. This model can be used to navigate the design
space.

II. Data analysis for response-2 cell mass count of fermented wash

Quadratic model is suggested for analysis of cell count data all statistical data analysis are
within acceptable agreement. The model F-value of 3.44 implies there is a 5.89% chance that a
“model F-Value” this large could occur due to noise. The lack of fit F-Value of 0.09 implies
the lack of fit is not significant relative to the pure error. There is a 96.42% chance that a lack
of fit F-value.this large could occure due to noise and non significant lack of fit is good.
Table 4.24 Post ANOVA statistics for fermented wash yeast cell count
Std. dev. 2.64E+007 R-Squared 0.18
Mean 2.60E+008 AdjR-Squared 0.58
C.V. 10.14 PredR-Squared 0.55
PRESS 1.18E+016 Adeq.Precision 7.45

The “Pred R-Squared” of 0.55 is in reasonable agreement with the “ Adj R-Squared” of 0.58.
“Adeq. Precision” measures the signal to noise ratio. A ratio greater than 4 is desirable. The
ratio of 7.45 indicates an adequate signal. This model can be used to navigate the design space.
III. Data analysis for response-3 residual sugar of fermented wash

2FI model is suggested for analysis of cell count data all statistical data analysis are within
acceptable agreement. The model F-value of 4.58 implies there is a 1.72% chance that a
“model F-Value” this large could occur due to noise. The lack of fit F-Value of 29.51 implies
the lack of fit is not significant relative to the pure error. There is a 0.28% chance that a lack
of fit F-value.this large could occure due to noise and non significant lack of fit is good.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 59
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.25 Post ANOVA statistics for FW residual sugar concentration

Std. dev. 0.46 R-Squared 0.73
Mean 1.65 AdjR-Squared 0.57
C.V. 28.05 PredR-Squared 0.50
PRESS 8.81 Adeq.Precision 8.04

Adeq. Precision measures the signal to noise ratio. A ratio greater than 4 is desirable. The ratio
of 8.04 indicates an adequate signal. This model can be used to navigate the design space.

4.3.2.2. Correlation matrics for effluent optimization at 8.67%(15 Brix)

sugar concentration
From correlation matrix of factors (Pearson’s R), it was clearly observed that almost all factor
to factor correlations are zero and it is good.
Table-4.26: Correlation Matrix of Factors [Pearson's R]
A B C A2 B2 C2 AB AC BC

A 1.000

B 0.000 1.000

C 0.000 0.000 1.000

A2 0.000 0.000 0.000 1.000

B2 0.000 0.000 0.000 0.056 1.000

C2 0.000 0.000 0.000 0.056 0.056 1.000

AB 0.000 0.000 0.000 0.000 0.000 0.000 1.000
AC 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000
BC 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000

4.3.2.3. Equation for effluent optimization at 8.6 sugar concentration

Equations are given in terms of coded factors and actual factors. Coded factors indicate when
the minimum and maximum values of the factors are represented by -1 and +1 respectively
instead of their actual values.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 60
 Optimization of Molasses Fermentation to Reduce Effluent

I. Alcohol % fermented wash

Equation in terms of coded factors:
Alcohol%fermented wash = +5.12-0.26*A-0.13*B-0.045*C-0.19*A*B-0.90*A*C
+0.38*B*C (4.19)
Equation in terms of Actual factors:
Alcohol% Content =+4.86+3.28*SW-0.65*SL+0.94*PC-1.5*SW*SL
-7.18*SW*PC+1.53*SL*PC (4.20)

II. Cell mass count fermented wash

Final Equation In Terms of Coded Factors:
Yeast Cell Count = +2.61E+008 +2.06E+ 007 * A -6.25E+006 * B-6.25E+005*C
+3.39E+007A2 -3.49E+007B2-1.13E+006C2 -2.5E+006*A*B
+4.375E+007*A*C+1.250E+007*B*C (4.21)
Final equation interms of Actual Factors:
Yeast cell mass count = +2.99E+008 – 3.54E+008*S +1.07E+008*SL -
1.09E+008* PC+5.42E+008 * SW2 -1.40E+008* S L2 -
4.50E +006*PC2 – 2.0E+007*SW*SL+ 3.50E+008*SW* PC
+5.0E+007 * SL * PC (4.22)
III. Residual sugar percent of fermented wash
Final Equation In Terms of Coded Factors:
Residual Sugar = +1.65+0.32 * A+0.18 * B+0.21 * C+0.14 * A * B
+0.83 * A * C -0.63 * B * C (4.23)
Final Equation in terms of Actual Factors:
Residual Sugar = +1.27147 -2.58*SW+1.35*SL +0.013*PC +1.1*SW*SL
+6.64*SW*PC -2.51*SL*PC (4.24)
Where A,B andC are coded value for Spent wash, Spent lees and Process condensate.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 61
 Optimization of Molasses Fermentation to Reduce Effluent

Diagnostic test for the Responses

All diagnostic plots are also tested for all responses for adequacy of the models (normal plot of
residuals, residuals vs predicted value, residuals vs factor, box cox plot, studentized residuals,
leverage, etc.) and all found in good condition.
4.3.2.4. Response surface plot analysis for effluent optimization at 8.67 %
(15 Brix) sugar concentration
I. Response-1 alcohol% fermented wash
DESIGN-EXPERT Plot
Alcohol% fermented Wash
Y = B: Spent Lees
X = A: Spent wash
5.44676
Actual Factor 5.22051
4.99426
C: PC = 50% 4.76801
Alcohol% 4.54176
fermented Wash

1.00
0.50
0.75
0.38
0.50 0.25
B: Spent Lees 0.25 0.13 A: Spent wash
0.00 0.00

Fig 4.11 Response surface plot of alcohol% fermented wash at constant

process condensate volume of 50% recirculation rate.

At constant process condensate application or returning 50% of the generated process

condensate to fermentors, increasing spent wash and spent lees recirculation rate to fermentors
simultaneously will reduce alcohol% fermented wash.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 62
 Optimization of Molasses Fermentation to Reduce Effluent

DESIGN-EXPERT Plot
Alcohol% fermented Wash
Y = C: Process Cond
X = A: Spent wash 6.68
Actual Factor 6.01
5.33
4.66
B:SL= 1.00 3.98

Alcohol% FW 1.00
0.75 0.50
0.38
0.50 0.25
C: Process Cond 0.25 0.13 A: Spent wash
0.00 0.00

Fig 4.12. Response surface plot of alcohol% fermented wash at constant spent
lees volume of 1.00.
It has been clearly seen from the graph that simultaneously increasing process condensate and
spent wash has side effect on fermented wash alcohol production.

DESIGN-EXPERT Plot Cube Graph

Alcohol% fermented Wash Alcohol% fermented Wash
X = A: Spent wash 6.68 3.98
Y = B: Spent Lees
Z = C: Process Cond
B+ 4.21 5.10
B: Spent Lees

5.80 3.85 C+
C: Process Cond
B- 6.50 A+ C-
A- 4.86
A: Spent wash

Fig 4.13 cubic graph of alcohol% fermented wash at 15 brix

From cubic plot for alcohol% fermented wash it is clearly seen that increasing recirculation of
spent wash alone will have a positive effect on alcohol production. Similarly, increasing
recirculation of process condensate alone to fermenter has also the same effect with spent

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 63
 Optimization of Molasses Fermentation to Reduce Effluent

wash. On contrary to this, applying recirculation of spent lees alone to fermenters has harsh
effect on alcohol production. But applying spent lees either with spent wash or Process
condensate will have good effect on alcohol production. In the other hand applying spent wash
together with process condensate had a negative effect on alcohol production.
II. Response-2 Cell mass count

DESIGN-EXPERT Plot

X= A:SW
Y=B:SL 3.16E+008
2.91E+008
Actual Factor 2.67E+008
2.42E+008
C: PC= 50% 2.17E+008
Yeast mass

1.00
0.75 0.50
0.38
0.50 0.25
B: Spent Lees 0.25 0.13 A: Spent wash
0.00 0.00

Fig 4.14. Response surface plot for cell mass at constant process condensate flow.

The cell mass increases with spent wash application rate but it reaches maximum at 50% spent
wash and nearly around 50% spent lees recirculation rate to fermentors. Increasing spent lees
recirculation above 50% and reducing below 50% shows a reduction in cell mass
concentration.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 64
 Optimization of Molasses Fermentation to Reduce Effluent

DESIGN-EXPERT Plot
X = A: SW
Y= C: PC 3.58E+008
Actual Factor 3.25E+008
2.93E+008
B: SL=50% 2.61E+008
2.29E+008
Yeast mass

1.00
0.75 0.50
0.38
0.50 0.25
C: PC 0.25 0.13 A: SW
0.00 0.00

Fig 4.15. Response surface plot at constant recirculation rate of spent lees

DESIGN-EXPERT Plot Cube Graph

Yeast mass Yeast mass
X = A: Spent wash 202625000.00
Y = B: Spent Lees 326375000.00
Z = C: Process Cond

266375000.00 215125000.00
B+

B: Spent Lees

318875000.00
185125000.00
C+

C: Process Cond
298875000.00 257625000.00
B- C-
A- A+
A: Spent wash

Fig 4.16 cubic graph of yeast cell mass count which shows interaction of factors.

The cubic graph indicates that applying or re-circulating one effluent stream only has drastic
effect on the cell mass growth but recirculating all effluents simultaneously to fermentors have
very good effect on the cell mass growth.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 65
 Optimization of Molasses Fermentation to Reduce Effluent

III. Residual Sugar % Fermented Wash

DESIGN-EXPERT Plot
Residual Sugar
X=A:SW
2.29
Y=B:SL 2.04
1.78
Actual Factor 1.53
C: PC=50% 1.28
Residual Sugar

1.00
0.75 0.50
0.38
0.50 0.25
B: Spent Lees 0.25 0.13 A: Spent wash
0.00 0.00

Fig 4.17 Response surface plot for residual sugar concentration

DESIGN-EXPERT Plot Cube Graph

Residual Sugar Residual Sugar
X = A: Spent wash 0.12 2.70
Y = B: Spent Lees
Z = C: PC =50%
B+ 2.62 1.88

B:Spent Lees 3.32

1.28 C+
C: Process Cond
B- 1.27 0.00 C-
A- A+
A: Spent wash

Fig 4.18 Cubic graph for residual sugar concentration of fermented wash at 15 brix

The cubic plot of residual sugar indicates that increasing the amount of recirculation of spent
wash to fermenters reduces the residual sugar concentration on converting most of the sugar to
alcohol. Whereas applying spent lees or process condensate alone was increases residual sugar
in fermented wash.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 66
 Optimization of Molasses Fermentation to Reduce Effluent

4.3.2.5. Optimization solution for effluent recirculation for sugar

concentration of 8.67% fermented wash

The optimum parameter values for fermented wash; alcohol%, yeast cell mass count and
residual sugar were calculated by partially differentiating equations (4.20), (4.22) and (4.24)
with respect to each parameter and equating to zero (montgomery 2001). The optimization was
to get maximum alcohol % fermented wash, optimum cell mass and minimum residual sugar
as much as possible. The first response (alcohol% fermented wash) was considered as an
important response and more weight is given to it since it is the major determining factor of
this optimization. Predicted value for this optimization was:
Spent wash 1%, Spent lees 92%, process condensate 99%, yeast cell mass count
2.15488E+008, Alcohol% fermented wash 6.57, residual sugar content of 0.274 and
desirability of 1.000. To verify the predicted value an experiment was conducted on using the
predicted effluent re-circulation rate and preparing fermented wash at 15°brix sugar
concentration. Then pH of fermented wash obtained was 5.5, alcohol% fermented wash of 6.5,
residual sugar content of 0.31 and yeast cell mass count of 2.21E+008. The predicted and
experimental analysis result was in good agreement. In addition , equation (4.20) and (4.24)
were differentiated twice and it was confirmed that the alcohol% obtained was maximum with
minimum residual sugar.

4.3.3.Optimization of effluent at different fermentable sugar concentration

Fermented wash with brix of 15(8.67% fermentible sugar), 20(11.5% FS) and 25 %(14.3%FS)
was prepared and poured on 4 litres capacity laboratory scale fermentor. To each fermentor
nutrient such as Urea and DAP are provided as per the optimization results for each level of
fermentable sugar such as at rate of 45.16 mg/lt urea and 27.05 mg/lt DAP for fermentable
sugar concentration of 8.6%, 30.46 and 11.48 mg/Lt urea and DAP respectively for 11.5%
fermentible sugar wash and 28.76 and 26.98 mg/lt Urea and DAP for fermentible sugar of
14.3% wash. Then fermentation was conducted for 48 hours at room temprature and the
following data were generated.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 67
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.27 Optimized result of effluent at different fermentabile sugar

Std Run Factors Response Variables
%Concen % % % Process Alcohol% Residual Yeast Cell
tration of Spent Spent Condensate wash sugar% mass
Sugar Wash lees
16 1 11.5 50 100 50 7.8 1.92 230,000,000
13 2 11.5 0 0 50 6.47 1.61 250,000,000
15 3 11.5 0 100 50 4.19 3.9 175,000,000
28 4 11.5 25 50 50 7.09 1.2 300,000,000
19 5 8.6 25 100 50 5.25 1.51 180,000,000
11 6 8.6 25 50 100 5 1.3 250,000,000
27 7 11.5 25 50 50 6.65 1.6 375,000,000
4 8 14.3 50 50 50 6 4.57 250,000,000
1 9 8.6 0 50 50 4.65 1.98 150,000,000
8 10 11.5 25 100 100 6.9 1.25 250,000,000
20 11 14.3 25 100 50 7.2 3.3 275,000,000
25 12 11.5 25 50 50 7.09 1.1 315,000,000
18 13 14.3 25 0 50 5.2 7.2 150,000,000
24 14 11.5 50 50 100 7.06 1.3 275,000,000
14 15 11.5 50 0 50 6.83 1.29 295,000,000
5 16 11.5 25 0 0 6.8 1.28 285,000,000
2 17 14.3 0 50 50 8 1.07 300,000,000
29 18 11.5 25 50 50 7.13 1.13 300,000,000
26 19 11.5 25 50 50 7.2 1.1 305,000,000
10 20 14.3 25 50 0 6.7 4.4 250,000,000
3 21 8.6 50 50 50 5.86 0.91 215,000,000
9 22 8.6 25 50 0 5.9 0.86 300,000,000
6 23 11.5 25 100 0 6.9 1.4 295,000,000
23 24 11.5 0 50 100 7.19 1.15 275,000,000
21 25 11.5 0 50 0 6.8 1.28 265,000,000

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 68
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.27 continued

12 26 14.3 25 50 100 8.4 1.3 275,000,000
22 27 11.5 50 50 0 6.7 2 250,000,000
7 28 11.5 25 0 100 6.5 2.4 215,000,000
17 29 8.6 25 0 50 5.87 0.89 220,000,000

4.3.3.1. Data analysis for optimization of effluent at different sugar

concentration
I. Data analysis for response-1 alcohol% fermented wash
Quadratic model is suggested by the design program for this response to test for its adequacy
and to describe its variation with independent variables. From ANOVA test in table 4.23, the
model F-value of 3.56 implies the model is significant. There is only a 1.19 % chance that a
"model F-value" this large could occur due to noise.
Table 4.28 ANOVA test for fermented wash alcohol
Source Sum of DF Mean F-Value Prob >F
Squares Square
Model 20.90 14 1.49 3.56 0.0119 Significant
A 6.71 1 6.71 15.97 0.0013
B 0.73 1 0.73 1.73 0.2098
C 0.027 1 0.027 0.064 0.8032
D 0.13 1 0.13 0.31 0.5864
A2 3.05 1 3.05 7.27 0.0174

B2 0.39 1 0.39 0.94 0.3499

C2 1.32 1 1.32 3.14 0.0980
D2 0.18 1 0.18 0.43 0.5234
AB 2.58 1 2.58 6.14 0.0266
AC 1.72 1 1.72 4.09 0.0627

AD 1.69 1 1.69 4.03 0.0645

BC 2.64 1 2.64 6.29 0.0251

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 69
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.28 continued

BD 2.250E-004 1 2.250E-004 5.360E-004 0.9819

CD 0.023 1 0.023 0.054 0.8203

Residual 58.8 14 4.2
Lack of Fit 56.9 10 5.7 Not
Significant
Pure Error 1.9 4 0.048
Cor Total 26.78 28

The "lack of fit f-value" of 5.7 implies the lack of fit is not significant relative to the pure error.
There is a 11.09% chance that a "lack of fit f-value" this large could occur due to noise. Non-
significant lack of fit is good -- we want the model to fit.
Table 4.29 Post ANOVA statistics for fermented wash alcohol
Std. dev. 0.65 R-Squared 0.78
Mean 6.53 AdjR-Squared 0.56
C.V. 9.92 PredR-Squared 0.23
PRESS 33.05 Adeq.Precision 7.96

"Adeq precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. The
ratio of 7.96 indicates an adequate signal. This model can be used to navigate the design
space.
II. Data analysis for response-2 yeast cell mass count
Quadratic model is suggested by design expert and all parameters are found in acceptabile
ranges. The lack of fit F-value of 1.87 implies the lack of fit is not significant relative to pure
error. There is a 28.71% chance that a lack of fit F-value this large could occure due to noise.
Non significant lack of fit is good.
As it is described in table 4.30 Adeq precision measures the signal to noise ratio. A ratio
greater than 4 is desirable.The ratio of 5.540 indicates an adequate signal. this model can be
used to navigate the design space.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 70
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.30 Post ANOVA statistics for fermented wash yeast cell count
Std. dev. 4.06E+007 R-Squared 0.69
Mean 2.58E+008 AdjR-Squared 0.37
C.V. 15.76 PredR-Squared 0.58
PRESS 1.16E+017 Adeq.Precision 5.54

III. Data analysis for response-3 residual sugar % fermented wash

2FI model is suggested for inverse transformed response of residual sugar % fermented wash.
ANOVA and Post Anova test as well as the diagnostic test for this model are found within the
norm. the model F-value of 3.89 implies the model is significant. There is only a 0.60% chance
that a model F-Value this large could occre due to noise. The Lack of fit F-value of 3.08
implies the lack of fit is not significant relative to pure error. There is a 14.32% chance that a
lack of fit f-value this large could occure due to noise. Non significant lack of fit is good.
Table 4.31 Post ANOVA statistics for fermented wash residual sugar
Std. dev. 0.19 R-Squared 0.68
Mean 0.69 AdjR-Squared 0.50
C.V. 28.31 PredR-Squared 0.12
PRESS 1.88 Adeq.Precision 9.29

Adeq precision measures the signal to noise ratio. A ratio greater than 4 is desirable. Thus a
ratio of 9.285 indicates an Adequate signal. This model can be used to navigate the design
space.

4.3.3.2. Correlation matrics for effluent optimization at different sugar

concentration
From correlation matrix of factors (Pearson’s R), it was clearly observed that almost all factor
to factor correlations are zero and it is good.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 71
 Optimization of Molasses Fermentation to Reduce Effluent

Table-4.32: Correlation matrix of factors [Pearson’s R]

A B C D A2 B2 C2 D2 AB AC AD BC BD CD
A 1.000

B 0.000 1.000

C 0.000 0.000 1.000

D 0.000 0.000 0.00 1.000

A2 0.000 0.000 0.00 0.000 1.000

B2 0.000 0.000 0.00 0.000 -0.137 1.000

C2 0.000 0.000 0.000 0.000 -0.137 -0.137 1.000

D2 0.000 0.000 0.000 0.000 -0.137 -0.137 -0.137 1.000

AB 0.00 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000

AC 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000

AD 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000

BC 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000

BD 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000

CD 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000

4.3.3.3. Equation for effluent optimization at different sugar concentration

I. Response -1 alcohol% fermented wash

Final equation in terms of coded and actual factors are :
In terms of Coded factors:
Alcohol% FW = +7.03 + 0.75*A + 0.25*B + 0.048*C + 0.1*D- 0.69A2-0.25* B2 –
0.45*C2 +0.17*D2 -0.8*A*B +0.66*A*C+0.65*A*D+ 0.81*B*C –
7.500E-003*B*D +0.075*C*D (4.25)
In terms of Actual Factors :
Alcohol% FW = -4.97+1.15*Sugar concentration+ 0.13* SW%- 0.05*S L -0.06* P c
– 0.027* sugar concentration2 – 3.94E-004 SW2 – 1.80E-004*SL2 +
6.66E-005*Pc2- 6.42E-003* Sugar * SW + 2.62E-003*Sugar conc*SL
+2.60E-003* sugarConc*Pc + 6.50E-004*SW*SL – 6.0E-006*Sw*Pc +
3.0E-005*SL*Pc (4.26)

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 72
 Optimization of Molasses Fermentation to Reduce Effluent

II. Response-2 yeast cell mass count

Final equation interms of coded factors
Cell mass count= +3.19E+008 +1.54E+007 * A +8.33E+006 * B -8.33E+005*C
-8.75E+006*D -5.24E+007*A2 - 3.80E+007*B2 - 5.18E+007 * C2 –
6.17E+006 * D2 - 2.89E+007 * A * B+ 4.13E+007 * A * C
+1.88E+007*A*D + 2.5E+006 * B + 3.75E+006*B*D
+ 6.250E+006*C*D (4.27)
Final equation interms of actual factors
Cell mass count = -5.39E+008+8.07E+007*Sugar concentration+7.73e+006 * SW %-
1.42E+006*S L % -1.63E+006 * Pc% - 2.10E+006 * sugar concentration2
– 60866.67* SW%2 -20716.67 * SL %2 -2466.66667* Pc%2 –
2.30E+005* Sugar Concentration * SW%+ 1.65E+005 * SugarConc*SL%
+75000* Sugar conc* PC% +2000 * SW%* SL % +3000* SW% * Pc%
+2500* SL% * Pc% (4.28)
III. Response -3 residual sugar concentration
Final equation interms of coded factors
1.0/(Residual Sugar) = +0.69 - 0.23 * A - 7.08E-003 * B- 0.05 * C + 0.019 * D
- 0.33*A*B+0.16*A*C+0.23*A*D+0.028*B*C
+ 0.045 * B * D + 0.11 * C * D (4.29)
Final equation in terms of actual factors
1.0/(Residual Sugar) =+2.07 - 0.06* Sugar Concentration + 0.05* SW %
- 0.016* SL%-0.02*PC%-2.62E-003*Sugar Concn* SW% +
6.26E-004* Sugar Conc * SL% + 9.36E-004 * Sugar Conc*Pc%
+2.21E-005*SW%*SL%+ 3.62E-005 * SW% *PC% +
4.50E-005* SL % * Pc% (4.30)

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 73
 Optimization of Molasses Fermentation to Reduce Effluent

4.3.3.4. Response surface plot analysis for effluent optimization at different

sugar concentration
I. Response-1 alcohol% fermented wash

DESIGN-EXPERT Plot
Alcohol%FW
X=A: Sugar Conc. 7.40
Y=B:SW% Alc% FW 6.63
5.85
Actual Factors 5.08
C: SL = 50% 4.30

D: PC=50%
50.00
37.50 25.00
22.50
25.00 20.00
B: SW % 12.50 17.50
0.00 15.00 A: Sugar Concn

Fig 4.19 Response surface plot for alcohol% fermented wash at different sugar
concentration
Response surface plot showed that increasing sugar concentration as well as spent wash re-
circulation increases alcohol production. It is due to the increase in the nutrient supply and
creation of unfavorable condition for bacterial growth(J.God Bole,2002).
DESIGN-EXPERT Plot Cube Graph
Alcohol %FW Alcohol % FW
6.15 7.35
X=A:Sugar Concn
Y=B: SW%
Z=C:SL% B+ 5.74 4.32

Actual Factor B:SW %

D: PC=50% 2.43 6.84 C+
C: SL%
B- 7.06 A+ C-
A- 5.27
A: Sugar Concn

Fig 4.20 cubic graph for alcohol% fermented wash

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 74
 Optimization of Molasses Fermentation to Reduce Effluent

Cubic plot showed as the interaction effects of different factors such as sugar concentration,
spent wash and spent lees on alcohol production. As the concentration of sugar in fermenter
increases the production of alcohol will also increases. Similarly increasing spent wash
recirculation rate to fermenter at constant sugar concentration has an increasing effect to
alcohol production. But increasing spent wash and sugar concentration with in the ranges of
study has a negative effect on alcohol production as the yeast cell suppresed with high soluble
solid concentration. From the plot it is clearly seen that increasing spent lees recirculation
alone at any sugar concentration reduces alcohol production but if it is recirculated together
with spent wash and process condensate it increases production of alcohol.

DESIGN-EXPERT Plot Cube Graph

Alcohol FW% Alcohol
5.84 8.34
X=A: Sugar Concn.
Y= B:SW%
Z=D:PC% B+ 6.80 6.70

Actual Factor B:SW%

C: SL% = 100 2.14 7.85 D+

D: PC%
B- D-
A- 3.06 6.17 A+
A: Sugar concn.

Fig 4.21 cubic graph of alcohol% fermented wash at 100% spent lees
recirculation rate

The cubic graph in Fig 4.21 revealed that increasing sugar concentration has positive effect on
alcohol production and the cummulative effect of adding spent wash, process condensate and
sugar concentration at 100% recirculation of spent lees favours alcohol production.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 75
 Optimization of Molasses Fermentation to Reduce Effluent

II. Response-2 yeast cell mass count

DESIGN-EXPERT Plot
Cell mass%FW
X=A: Sugar Conc.
Y=B:SW% 2.85E+008
2.26E+008
Actual Factors 1.67E+008
C:SL=100% Cell 1.08E+008
4.85E+007
D:PC=100% Mass
% FW

50.00
25.00
37.50
22.50
25.00 20.00
B: SW % 12.50 17.50
0.00 15.00 A: Sugar Concn.

Fig 4.22 Response surface plot for cell mass count at different sugar
concentration and effluent recirculation rate.
Cube Graph
DESIGN-EXPERT Plot
135,166,666.67 Cell mass
Cell mass% FW 228,500,000.00
X=A:Sugar Concn
Y=B:SW% 201,833,333. 130,166,666.67
Z=C:SL% B+

Actual Factor B: SW % 256,833,333.33

48,500,000.00
D: PC=50% C+
C: Spent Lees %
125,166,666.67 168,500,000.00
B- C-
A- A+
A: Sugar concentration

Fig 4.23 Cubic graph for cell mass count at different sugar
concentration and effluent recirculation rate

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 76
 Optimization of Molasses Fermentation to Reduce Effluent

The Response surface plot indicates at 100% application of process condensate and spent lees
the cell mass reaches its maximum at 14.3% fermentable sugar(25˚brix) and at 25% spent wash
recirculation rate. The cube graph indicates that as the sugar concentration increases together
with spent lees the cell mass shows an increasing trend. Where as increasing spent wash with
sugar concentration suppress the cell mass growth due to presence of higher solid. But at lower
sugar concentration (8.6% FS) increasing spent wash recirculation increases the cell mass. On
the other hand increasing spent lees application at lower sugar concentration will suppres cell
mass growth. It might be due to presence of more acidic volatile compounds.

III. Response-3 Residual sugar concentration

DESIGN-EXPERT Plot
1.0/(Residual Sugar)
X = A: Sugar concentration
Y = B: SW %
Actual Factors
C: SL% = 100.00 4.73557
D: Pc = 100.00 3.7615
2.78743
Residual Sugar
1.81336
0.839296

50.00
25.00
37.50
22.50
25.00 20.00
B: Spent wash % 12.50 17.50
0.00 15.00 A: Sugar concentration
Fig 4.24 response surface plot for residual sugar of fermented wash

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 77
 Optimization of Molasses Fermentation to Reduce Effluent

DESIGN-EXPERT Plot
Cube Graph
1.0/(RS) Residual Sugar
X=A:Sugar Conc. 1.00 1.50

Y= B:SW%
Z=C:SL% B+ 0.89 5.71
Actual Factor
D:PC% = 100 B: SW %
4.74 0.84 C+
C: SL %
B- 1.24 A+ C-
A- 2.20
A: Sugar Concn.

Fig 4.25 cubic graph for residual sugar concentration of fermented wash
Response surface plot of residual sugar at 100% spent lees and process condensate re-
circulation rate shows a reducing trend as the spent wash re-circulation rate increases due to
improvement in availablity of nutrients which enhances cell mass growth. The cubic graph at
100% process condensate re-circulation rate shows increasing spent lees application to
fermenters will reduce consumption of sugar due to low growth rate of cell mass.
4.3.3.5. Optimization solution for effluent recirculation for different sugar
concentration of fermented wash

The optimum parameter values for fermented wash; alcohol%, yeast cell mass count and
residual sugar were calculated by partially differentiating equations (4.28), (4.30) and (4.32)
with respect to each parameter and equating to zero (montgomery 2001). The optimization was
to get maximum Alcohol % fermented wash, optimum cell mass and minimum residual sugar
as much as possible. The first response (alcohol% fermented wash) was considered as an
important response and more weight is given to it since it is the major determining factor of
this optimization so that to get higher alcohol and lower effluent generation. Solution for this
optimization was sugar concentration of 25%, spent wash 50%, spent lees 68.5%, process
condensate 85.83%, yeast cell mass count 2.40674E+008, alcohol% fermented wash 7.43 and
residual sugar content of 0.417. Experiment was conducted based on the predicted effluent

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 78
 Optimization of Molasses Fermentation to Reduce Effluent

recirculation rate and fermented wash sugar concentration. The analysis results were 7.5 %
alcohol content, 2.38E+008 yeast cell mass count with residual sugar concentration of 0.42 and
all results were almost identical with the predicted one. Besides, equations 4.28 and 4.32 were
differentiated twice in order to know whether the predicted values of effluent recirculation will
give minimum residual sugar and maximum alcohol and the value at predicted points are –
0.052 and 0.06 for equation (4.28) and (4.32) respectively. Since -0.052 is below zero, alcohol
% fermented wash predicted obtained at predicted value is maximum and residual sugar is
minimum as it is greater or equal to zero.

4.4. Effects of yeast cream separator on bulk production process

Assessment was done on bulk production process by selecting unit operations found in
fermentation section which having higher potential of reducing effluent generation on
maximizing alcohol production. Samples of fermented wash from fermenter-4, de yeasted
wash and yeast cream separated from yeast separator machine were taken three times and
counting of live and dead cell were done by Heamocyto meter. The separation efficiency was
found to be in the range of 41 to 71% with viability of 90-95%. But, from the specification data
sheet of the machine the efficiency of separation was 97% which is comparatively higher than
the actual values. Hence it is necessary to see different parameters like nozzle diameter of the
centrifuge and sediment concentration of fermented wash to improve separation efficiency so
as to get recirculation of more matured cell to fermenters with the view of increasing alcohol
production and reduction of effluent generation from distillery.
4.5. Effect of re-circulating effluent to fermenter on food safety and quality
of spirit
In order to study the effect of re-circulating spent wash, process condensate and spent lees to
fermentor comparative analysis was carried out between the distilled spirit obtained from
fermentation of molasses by using pure process water only and by using optimum recirculation
of process condensate, spent wash and spent lees. The results obtained were mentioned in
Table 4.33 below.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 79
 Optimization of Molasses Fermentation to Reduce Effluent

Table 4.33 Comparative analysis test results of rectified spirit produced with or
without recirculation of effluent

S/N Types of test Results for Results for

Standard
Unit sample A sample B
1 Aldehyde mg/l 0.707 0.712 < 40
content
2 Methyl alcohol Nil Nil NA
3 Fusel oil Ppm 50 50 < 450

4 Furfural Ppm Nil Nil < 64

5 Ester as ethyl gm/100ml 0.0088 0.0176 < 0.02
acetate
6 Total solids % 0.004 0.0038 Max 0.005
7 Copper gm/100ml 0.000356 0.000356 Max 0.0004
8 Acidity as gm/100ml 0.1893 0.2019 NA
acetic acid

The qualities of distilled sprit from fermented wash obtained by using optimum effluent
recirculation rate and process water were identical and no appreciable difference is found.
Moreover, all analysis results on quality parameters are maintained within the standard.

4.6. Material and Energy Balance

Material and energy balance was done for effluent recirculation system in order to suggest
proper equipment size, such as pumps, cooler, piping and cooling water required in order to
use vinasse, spent lees and process condensate. The balance was done twice based on the
current operating brix of fermenters (15˚Bx) and the near future operating brix after installing
molasses clarification plant( as per this research 25˚Bx).

4.6.1. Material and Energy Balance for 15˚Bx Fermenter Brix

Currently the distillery plant of Metahara Sugar Factory generates about 568 tons of Spent
wash from Primary Column, 87 tons Spent lees from Rectifier Column and 238 tons of process
condensate from Spent wash Evaporators respectively per day. Out of this, 1% of spent wash

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 80
 Optimization of Molasses Fermentation to Reduce Effluent

about 5.85 tons/day, 92% of Spent lees about 80 tons/day and 99% process condensate about
236 tons/day will be expected to recirculate according to optimization of Effluent recirculation.

4.6.1.1. Solid Balance of the Effluent

From the analysis results of spent wash, spent lees and process condensate mentioned in
Table4.2 and 4.3 above spent wash contains 13.6% soluble solids and the other two effluent
streams are with negligible solid content. Hence, the solid balance
Spent wash(5.85 t/d, brix = 13.6 %)

Spent lees(80.4 t/d) Buffer Tanks Recirculated

Process cond(236 t/d) Effluent (322.25t/d)
Fig 4.26 Solid Balance around Buffer tank
Mass of inflow stream to buffer tank = Mass of effluent recirculated to fermenters
Spent wash+ Spent lees+ Process Condensate = Effluent Recirculated
5.85 t/d + 80.4 t/d+ 236 t/d = 322.25 t/d
Effluent mass flow Rate re circulated to fermenter = 322.25 tons per day
Solid in each effluent stream = solid in recirculated effluent
Solid in SW + Solid in SL+ Solid in PC = Solid in effluent recirculated
5.85*0.136+ 80.4*0+ 236* 0 = 322.25* BXe
Effluent soluble solid content ( BXe) = 0.24 %
4.6.1.2. Energy Balance
Energy balance was done on Buffer tank and on re circulated effluent cooler to determine final
temperature of the re circulated effluent before cooler and after cooler. Besides, physical
characteristics of the effluent were determined.

Energy Balance on Buffer Tank :

Spent Wash(5.85 t/d, brix=13.6 %, temp= 85 ˚C)

Spent lees(80.4 t/d, 120˚C) Buffer Tanks Re circulated

Process Cond(236 t/d, 55˚C ) Effluent (322.25t/d)
Fig 4.27 Energy Balance on Buffer Tank
Energy in put to Buffer Tank with Effluent Stream = Energy out with Re circulated Stream
AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 81
 Optimization of Molasses Fermentation to Reduce Effluent

Energy in : Mass of SW* Cp 1 *T 1 + Mass of SL*Cp 2 *T 2 + Mass of Pc* Cp 3 * T 3

Energy in = 5.85 * Cp 1 *85˚C + 80.4t/d* Cp 2 * T 2 + 236 * Cp 3 * 55˚C
Where Cp1 is the Specific heat capacity of vinasse and it is found more accurately by using the
correlation derived by Gucker and Ayres(E.Hugot cane Sugar Hand Book 3rd edition page449).
Cp 1 = [1- (0.6- 0.0018*T + 0.0008(100-P)*(B/100)]*4.182 KJ/kg/˚C
T = temperature of spent wash in ˚C = 85˚C
B = brix of spent wash in % = 13.6%
P = Purity of Spent wash which is zero for this case.
Cp 1 = 0.928*4.182= 3.88KJ/kg ˚C
Cp 2 andCp 3 are the specific heat of Spent lees and Process condensate; and estimated by the
following equations (Perry Chemical Engineers hand Book page 2-174, Table 2-196).
Cp = C 1 + C 2 * T 2 +C 3 *(T 2 )2 + C 4 *(T 2 )3 + C 5 *(T 2 )4
Where the value of C 1 is 15359, C 2 = -116.12, C 3 = 0.4514, C 4 = -0.00078422 and 5.206 E-
007 forC 5 . Taking the temperature of spent lees to be 120˚C (393˚K) and 55˚C (328˚K) for
process condensate. The value of Cp 2 and Cp 3 became 4253.24 J/kg.˚K and 4181.7 J/kg.˚K.
Substituting the above values in energy equation yields
Energy input = 97,188,568KJ
Energy out : Mass of Re circulated Stream* Cp 4 * T 4
Energy out = 322.25 t/d* Cp 4 * T 4
Where: Cp 4 = [1- [0.6- 0.0018*T + 0.0008(100-P)]*(B/100)]*4.182 KJ/kg/ºC
= 4.176 kj/kg/ºC
Equating the equation of Energy in put = Energy out put
The final temperature of effluent in buffer tank became 72.4 ˚C

Energy Balance on Cooler

Cooling water out

Cooling water in
Fig 4.28 Energy balance on cooler

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 82
 Optimization of Molasses Fermentation to Reduce Effluent

Thermal energy entered to the cooler = Thermal energy leaving the cooler
Thermal Energy input = Energy with Effluent in + Energy in with cooling water
Thermal Energy output = Energy with cooled effluent+ Energy with warm water
Thermal Energy in = Mass of effluent* Cp 4 * T4 + mass of cooling water* Cp*T cooling water in
Thermal Energy out= Mass of effluent* Cp 4 * T5 + mass of cooling water* Cp *T cooling water out
Where T 4 = 72.4 ˚C and T 5 is the desired temperature of fermenter which is 34˚C. Cooling
water inlet and outlet temperatures will be maintained at 5˚ and 33˚C respectively.
Equating the energy balance equations and on rearranging
Mass of effluent*(72.4-34)*4.176 = Mass of cooling water* 4.182* (33-28)
322.25*(72.4-34)*4.176 = Mass of cooling water* 4.182*(33-28)
Mass of cooling water required = 441 tons/day.
4.6.2. Equipment Design
Cooler for cooling re circulated effluent has been designed. Plate type heat exchanger was
proposed in this study to satisfy the demand of current situation and future expansion of the
factory. The feature of PHEs are their compactness, higher heat transfer coefficient, cost
effectiveness and unique ability to handle fouling fluids. Moreover, its flexibility to increase or
reduce heat transfer area by adding or reducing the number of plate makes it special for
distillery like Metahara sugar factory for future expansion.

The cooler size is decided based on the heat load removed from the re circulated effluent and
it is calculated by using general heat transfer equation.
Heat transfer area = Q/ (U*LMTD).
Where Q = the thermal energy needed to be removed from the Re circulated effluent(KJ).
U= the overall heat transfer coefficient( W/m2K).
LMTD = the log mean temperature difference
Q = Mass of Re circulated effluent* Cp* Δ T
= 322,250*(72.4-34)*4.176 = 598.1 KW
24* 60*60
The overall heat transfer coefficient of plate heat exchanger for the duty of water to water
cooling and heat transfer area less than 2500m2 and number of plates between 3- 700 is lying in
the ranges of 3000- 7000 ( W/m2K) and for this specific case 3000( W/m2K) was selected
(Boem R.F. etal, 1999).
AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 83
 Optimization of Molasses Fermentation to Reduce Effluent

LMTD = ( T hot, in - T cold, out ) - ( T hot, out - T cold, in )

Ln( ( T hot, in - T cold, out ) / ( T hot,out - T cold, in ))
= ((72.4-33) - ( 34-5))/ Ln(( 72.4-33)/ ( 34 – 5)) = 33.93
Heat transfer Area required = 598.1 KW / (3( KW/m2K) * 33.93) = 5.9 m2
Taking safety factor of 5%, the heat transfer area required will be 6.2 m2
In Methara Sugar Factory distillery already installed PHE cooler having 5 m2 is available to
cool recycle vinasse and it is possible to increase flow area by adding plates to 5.9 m2.
Pipe Sizing
The pipe selected for effluent recycling system has the duty to handle 322.25 tons in a day.
Taking densities of effluent to be 1 ton per meter cube, normal recommended velocity of the
effluent to be 1.25-2.5m/second. Hence, pipe diameter was determined by using the following
equation:

D=

Where ρ is densities of an effluent and at 0.24° brix it is taken as 1005kg/m3

M - is mass flow rate of effluent to be recirculated and taken to be 321.68tons/day
U- is allowable velocities of effluent with in a pipe and taken to be 2 m/s
The diameter of pipe became 1.92 inch and taking in safety factor of 5%, then the diameter of
pipe required was 2.016 inch and from standard 2 inch pipe is selected.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 84
 Optimization of Molasses Fermentation to Reduce Effluent

5. Financial Analysis of Effluent Recirculation

Experimentally and Engineering wise it has been seen that recirculation of distillery effluent to
fermentation plant at optimum flow rate and temperature is feasible. But, its feasibility should
be checked financially. In this section, the proposed project to be analyzed financially
depending on the proposed and designed recirculation system according to Max s. Peters,
Klaus D. Timmerhaus and James C. Van Horne (Max S. et.al (1991) James et.al(2004)).
5.1. Investment Cost Estimation
In this cost some costs like land moving, land lease, building cost and other minor costs have
not been included since the plant already exists and the only work now is the expansion work.
Since 5m3 tank to serve as a buffer tank, recirculated effluent transfer pump with 15 m3
capacity and PHE with 5m2 are available in the factory. Hence there is no need for civil work.
But cost for piping, installation of three flow meters for proportionate the effluent and PHE
plates are included. As shown in Table 5.1 to implement effluent recirculation system cost to
install 3 flow meter, about 6 PHE plate and pipes with contingency are considered.
Table 5.1. Investment cost to install effluent recirculation system
S/N Description Unit Quantity Birr per unit Total cost
1 Flow meter Pcs 3 120,000 360,000
2 PHE plates and gasket Pcs 6 3,220 19,320
3 Pipes Meter 24 60 1440
4 Contingency % 20 76,152
Total 456,912

5.2. Operating Cost Estimation

This cost is the cost incurred due to day to day operation activity of the proposed effluent re
circulation system to fermenters. In this case, overhead cost does not included as there is no
additional overhead cost due to implementation of this plant. Besides, no additional operator is
required to operate the system. The only thing required to run this system is the cost of
additional chilling water to cool the effluent.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 85
 Optimization of Molasses Fermentation to Reduce Effluent

Table 5.2. Operating cost to run effluent recirculation system

S/N Description Unit Quantity Cost per unit Total cost
in Birr
1 Water M3 107,136 5.6 birr 599,964
2 Electricity KWH 145,071 0.58 84,141
3 Contingency % 15 102,616
Total 786,721

5.3. Cost to treat Effluents

Process condensate and spent lees generated from vinasse concentrating evaporators and
Rectifier column are treated in waste water treatment plant having anaerobic and aerobic
compartments combined with neutralization tank. In this treatment plant different materials
like urea, DAP, dung, caustic soda and electrical energy are required.
Table 5.3. Treatment cost of effluent

S/N Description Unit Quantity Cost per unit Total cost

in Birr in Birr
1 Caustic Soda Kg 100,800 9 907,200
2 Cow dung Qts 200 70 14,000
3 Urea kg 1460 10 14,600
4 Dap kg 1460 12 17,520
5 Electricity KWH 9363 0.58 5430
Total 958,750

Applying optimization results and on re circulating 92% spent lees and 99% process
condensate to fermenter without any waste water treatment techniques will reduce the
cost mentioned in Table 5.3 to 97.2 % and become 26,550 Birr on saving 932,200 Birr.
Besides, treated water requirement for fermentation plant also reduced by 316 m3 and
chemical as well as energy requirement to treat this amount of water, the saving
obtained will be 430,013 Birr per annum. In addition to safeguarding the environment
from pollution by applying optimization of effluent recirculation to fermenters, a Total
of 1,362,213 Birr will be saved. Moreover, daily sulfuric acid consumption also
reduced from 121 litres to 98 litres which increases the annual saving by 115,862Birr.
AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 86
 Optimization of Molasses Fermentation to Reduce Effluent

5.4. Projected Cash Flow

All cost data are summarized in the following table in order to evaluate the proposed effluent system financially.
Table 5.4 projected cash flow of the proposed system
Year 0 1 2 3 4 5 6 7 8 9 10
Capital utilization % 100 100 100 100 100 100 100 100 100 100 100
Cash in flow
In Birr’000’ 1503 1503 1503 1503 1503 1503 1503 1503 1503 1503 1503
Saving of raw water in 430 430 430 430 430 430 430 430 430 430 430
take
Saving of chemicals 1068 1068 1068 1068 1068 1068 1068 1068 1068 1068 1068
and materials
Saving of electricity 5.430 5.430 5.430 5.430 5.430 5.430 5.430 5.430 5.430 5.430 5.430
Cash out flow
In Birr ‘000’ 1243.6 887.2 887.2 887.2 887.2 887.2 887.2 887.2 887.2 887.2 887.2
Investment 456.9
Utility cost 786.7 786.7 786.7 786.7 786.7 786.7 786.7 786.7 786.7 786.7 786.7
Depreciation(10%) 45.69 45.69 45.69 45.69 45.69 45.69 45.69 45.69 45.69 45.69
Interest(12%) 54.83 54.83 54.83 54.83 54.83 54.83 54.83 54.83 54.83 54.83
Gross Profit 259.4 615.78 615.78 615.78 615.78 615.78 615.78 615.78 615.78 615.78 615.78

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 87
 Optimization of Molasses Fermentation to Reduce Effluent

5.5. Profitability
According to the income statement, the project will start generating profit in the first year of
operation. In order to check the feasibility of the project other indicators of profitability were
calculated.
Pay Back Period
The investment cost and income statement projection are used to project the pay-back period as
follows:
Table 5.5 payback period on investment for the proposed effluent recirculation system
Year 0 1 2 3
Cumulative cash flow 259.4 875.18 1490.96 2106.74
Yearly cash flow 259.4 615.78 615.78 615.78

Payback period = 0 Years + 456.9 * 12 months = 8.9 months

615.78
The payback period indicates that implementation of effluent re circulation to fermenters
returns its investment in 9 months.
Table 5.6 NPV Analysis at 12%DCF
Year Cash flow(103 Birr) DCF PV(103 Birr)
0 -456.9 1 -456.9
1 615.78 0.893 549.89
2 615.78 0.797 490.8
3 615.78 0.711 437.82
4 615.78 0.636 391.64
5 615.78 0.567 349.14
6 615.78 0.507 312.2
7 615.78 0.452 278.33
8 615.78 0.404 248.78
9 615.78 0.361 222.3
10 615.78 0.322 198.28

NPV= ∑ PV= 3,022,257 Birr, Which is positive.

All financial terms indicates the system proposed to re circulate effluent to fermentation plant
was feasible financially.
AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 88
 Optimization of Molasses Fermentation to Reduce Effluent

6. Conclusion and Recommendation

6.1. Conclusion
This study clearly indicates that there can be significant technical potential for maximum
economic use of effluent recirculation to fermenters with the view of implementing cleaner
industrial production system. The experiment was designed in series of steps aiming on
reducing effluent production in increasing alcohol productivity through optimizing nutrient
application to fermenters and increasing cell mass on efficiently using yeast recycling
machines followed by returning optimum amount of effluent to fermenters. Beside to this the
quality of input materials such as process water and molasses were also checked and found in
optimum ranges. But the calcium oxide content of molasses is found beyond the required
22,171 ppm which is above 16,500 ppm. This can cause incrustation on evaporator and
distillation column which resulted in high effluent generation. In addition to this accumulation
of sludge inside fermenters can occur as the calcium gets precipitated by the action of acid
added to maintain the pH of fermenters around 4.5 and this can reduce the effective volume of
the fermenters which has the consequence of reducing alcohol% fermented wash and
increasing effluent generation.
Response surface method was used to analyze the data generated from nutrient optimization
and effluent recirculation system. The experiments were conducted in the current operating
condition of MSF fermentation (15˚bx) of molasses and future optimum molasses feed rate.
The optimum nutrient dose for the current operating condition of fermentation was 62.51 and
14.51mg/lt Urea and DAP respectively. It was found in the experiment that the optimum macro
nutrient required slightly varies according to fermentable sugar concentration of fermented
wash. The study shows nutrient requirements of fermentation process increases for DAP with
the increase in concentration of fermentable sugar but remains the same for Urea. At 15ᵒ bx
fermented wash concentration the optimum amount of Urea and DAP application was 62.51
and 14.51 mg/l where as at 25ᵒbx fermented wash sugar concentration 62.51 and 31.26 mg/l
Urea and DAP quantity are required. Experiment were conducted on the existing fermentable
sugar concentration of fermented wash(15 brix) and at different sugar concentration (15, 20
and 25 brix) to get optimum recirculation of effluent and sugar concentration which resulted in
higher alcohol production with low effluent generation. The optimum effluent rate for the
existing sugar concentration were Spent wash 1%, Spent lees 92% and process condensate

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 89
 Optimization of Molasses Fermentation to Reduce Effluent

99%. As a result the response variable became yeast cell mass count 2.15488 E+008,
Alcohol% fermented wash 6.57 and residual sugar content of 0.274. The optimum effluent
recirculation to fermenter at different sugar concentration became sugar concentration of 25%,
spent wash 50%, spent lees 68.5% and process condensate 85.83%. Yeast cell mass count
2.40674E+008, Alcohol% fermented wash 7.43, residual sugar content of 0.417 were obtained
as response variables. After obtaining the optimum effluent recirculation rate to fermenter
comparative test has been conducted by preparing fermented wash on using optimum effluent
recirculation and by using only process water and fermenting them for 40 hours, distilling the
fermented wash by using lab scale distillation mantle to produce rectified sprit alcohol. Then
analysis of the sprit collected were done for total solid , furfural, fusel oil, aldehyde, ester and
other attributes which is related from food safety point of view and all the results were found in
the required ranges for both samples.
Material and Energy balance was conducted based on optimized effluent recirculation rate and
suitable types of cooler, pumps and piping were suggested in order to implement the findings
of the research. Besides, cost benefit analysis was done for effluent recirculation system and
all financial indicators have shown the viability of the project.
6.2. Recommendation
According to the experimental results and practical observations of Metahara Sugar Factory
distillery plant the following recommendations to be implemented to reduce effluent discharge
from the plant to environment and increase productivity:
 To utilize strictly the research results of this thesis.
 In order to reduce effluent generation per liter of alcohol from the plant the factory has to
focus on increasing alcohol % of fermented wash on increasing retention time from the current
32-40 hours to 48-60 hours by installing additional fermenter and applying more molasses to it.
Besides, to install molasses clarification system to avoid excess lime salts from molasses feed
which will avoid accumulation of sludge inside fermenters which has a possibility of reducing
reaction volume and retention time of fermented wash. Furthermore, increasing availability of
yeast separator machines through proper maintenance practice so as to increase the count of
cell mass inside fermenter on re circulating yeast cream.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 90
 Optimization of Molasses Fermentation to Reduce Effluent

6.2.1. Recommended future study

Further study is recommended on the following areas:
 Yeast strain selection which will withstand high sugar concentration so as to ferment the
molasses relatively on higher concentration of sugar on reducing the use of water for dilution
and effluent generation.
 Suitable enzyme selection and application to fermenter with the view of increasing alcohol %
fermented wash and reducing effluent generation.
 Alternative disposal methods for distillery effluents like ferti irrigation on cane plantation
fields, to install methanator and convert the high BOD effluent to lower one on simultaneously
producing electric energy and to use the vinasse produced as animal fodder on processing it
further.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 91
 Optimization of Molasses Fermentation to Reduce Effluent

References
Agrawal P K, and Kumar S (1998) Studies on Alcohol Production from Sugarcane
Juice, Sugarcane Molasses, Sugar beet Juice and Sugar beet Molasses,
Saccharomyces Cerevisiae Nsi-113. Proceedings of the 60th Annual Convention
of the Sugar Technologists Association of India, Shimla, India.
Ajay K.etal (2012) Physico-chemical characterization of distillery effluent and its dilution
Effects at different levels. Archives of applied Science Research.
Annex J, (2013) Summary of Current food Standards as of april 2013.
Bansal R and Singh Rs (2003) a Comparative Study on Ethanol Production From Molasses
Using Saccharomyces Cerevisiae and Zymomonas Mobilis. Indian JMicrobio
l43:261-64.
Bhairappa C. A novel use and Re use of molasses based distillery effluent using Ferm Aid P.
E.Hugot (1986) Hand book of cane sugar engineering, 3rd edition, Elesevier science
Publisher company INC. The Netherlands.
Fal Allen(1994) The micro brewery Laboratory manual- a practical guide to Laboratory
Techniques and Quality control procedures for small scale brewers.
GB Deshpande(2002) Overview of Continuous Alcohol fermentation and Multi pressure
Distillation technology. Praj Industrial limited, Pune.
Godbole J. 2002. Ethanol from cane molasses Hawaii workshop .PRAJ industries LTD. Pune
,India
Harrigan, W.F. Laboratory Methods in Food Microbiology 3rd Ed. Academic Press, San
Diego, CA. 1998
IAN P. Willington, Gerald G.M.(1982) Options for Handling stillage Waste from sugar based
fuel ethanol production. Elsevier scientific publishing company Amsterdam.
International Commission for Uniform Methods of Sugar Analysis (ICMSA), 1994.
ICUMSA Methods Book
Joshi, h. C., karla, n., choudhury, a. And deb, d. L. 1994.Environmental issues related with
distillery effluent Utilization in agriculture in india. Asia pacific journal of
Environment development 1:92 103.
Kadambini Gaur.(2006). Process Optimization for the Production of Ethanol Via Fermentation
Kaur U (1995) Studies on Alcoholic Fermentation at Higher Substrate Concentration. M.Sc.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 92
 Optimization of Molasses Fermentation to Reduce Effluent

Thesis, Punjab Agricultural University,Ludhiana, India.

K.A.Jacques, T.P.Lyons, DR Kellsall, (2004) Alcohol text book 4th edition A reference for fuel
and industrial Alcohol industries, published by Nottingham University Press.
KBK chemical engineering PLC operating Manual (2009), Maharashtra, India.
Madhavi, A, & Rao, A. P. Effect of Industrial Effluent on Properties of Groundwater.
Journal of Environmental Biology (2003). , 24(2), 187-192.
Mahimairaja, S., Bolan, N. S. (2004): Problems and Prospects of Agricultural Use of distillery
Spent Wash In India. In: Super Soil 2004. Proceedings of the 3rd Australian New
Zealand Soils Conf.,University Of Sydney Australia,
(Www.Regional.Org.Au/Au/Assoi/Supersoil 2004)
Manonmani, K., Chitrarasu, G., Swaminathan, K. (1990): Effect of Spent wash on the
Physicochemical Properties of Saline Calcareous Soil. J. Poll. Res., 9, 79–82.
Max S. Peters, Klaus D. Timmerhaus (1991) Plant Design and Economics for Chemical
Engineers, 4th Edition, University of Colorado, USA.
N.K. Saha etal(2004) Improving water usage for Indian distilleries published by Elsevier.
Peter Rein.(2007). Cane Sugar Engineering Verlas Dr. Albert Bartens Kg - Berlin 2007.
Rao M. J. P., 1997. Industrial Utilization of Sugar Cane and Its Co-Products. IAPCK
Publishers and distributors. New Delhi India.
Robert H. Perry (1999) Perry’s Chemical Engineers’ Handbook, 7th edition, The
McGraw-Hill Companies, Inc., USA
Sarayu Mohana, Bhavik, K, Acharya and Datta Madamwar. 2009. Distillery Spent Wash
Treatment Technologies and Potential Applications. J. Hazardous Material
163(1):12-25.
Sharma K.K. (2012). Study Report on Ethanol manufacturing Process of Metahara Sugar
Factory. Integrated CaseTech Pvt LTD. New Delhi.
Sh.S.Suresh etal(2011) Report on assessment of grain based fermentation technology, waste
Treatment options, disposal of treated effluents.
Training Manual on manufacturing of Ethyl alcohol from cane molasses by Ethiopian Sugar
Corporation Research and Training Directorate, Wonji-Shoa.
WHO (2006) World Health Organization Guidelines for drinking-water quality[electronic
resource].Volume. 1, Recommendations. – 3rd edition

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 93
 Optimization of Molasses Fermentation to Reduce Effluent

Appendices
Appendix- A: Preparation of fermented wash samples
1.Samples preparation and analysis for Nutrient Optimization
- Take samples of molasses having 85% brix 0.56, 0.76 and 0.97 kg for preparation of 15, 20
and 25 brix molasses solution.
- Dilute the molasses with process water till the volume of solution reached 13 liters and check
the brix of the solution with refractometer.
- Adjust the brix of solution to 15, 20 and 25% by adding slightly process water and molasses.
- Adjust the pH to 4.5+ 0.2 by adding H 2 SO 4 .
- Add 1.3 liter of laboratory yeast media.
- Pour the prepared molasses solution in to thirteen 1 liter volumetric flasks.
- Add Macro nutrients as per the randomized combination given in the design expert in the
range of 12.21 to 71.14 gm/ lt and 6.1 to 35.58 mg/ lt for Urea and DAP respectively.
- Let the samples to ferment in room maintained at 28-32˚C for 40 hours.
- Determine alcohol % FW with Ebullometer, count the cell mass with heamocytometer and
residual sugar by using Lane and Eynon methods.
2.Samples Preparation and Analysis for Optimizing Effluent Recirculation
- To prepare molasses solutions with 15, 20 and 25 brix , take samples 0.56, 0.76 and 0.97 kg
of molasses with 85 brix respectively and dilute them to 4 litres with a combination of spent
wash, spent lees, process condensate and process water as per the combination ratio mentioned
in the design expert.
- Adjust the pH of solution to 4.5 + 0.2 by adding H 2 SO 4 .
- Add Urea and DAP to fermented wash sample in the amount of 115 and 108 mg for 25 brix
(14.3 % sugar concentration), 122 and 46 mg for 20 brix(11.5% sugar) and 180 and 108 mg for
15 brix ( 8.6% sugar) solution respectively.
- Add Laboratory Yeast Media 400 ml to the prepared Molasses media and left the samples in
the room maintained at 28-32˚C for 40 hours.
- Determine alcohol % FW, yeast cell count and residual sugar content.
Appendix B:- Modeling Experimental Samples from Bulk Fermentation Process.
In the bulk production process of Metahara sugar factory distillery plant the quantity of input
materials used and output generated on daily bases are:

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 94
 Optimization of Molasses Fermentation to Reduce Effluent

Table 7.1 – Mass balance on each unit process

S/N Process Input Output
1 Fermentation - Process water 571 tons - Fermented wash 713 tons
- Molasses of 206 tons - Sludge 20 tons
- CO 2 44 tons
2 Distillation - Fermented wash 713 tons - Rectified sprit 42.5 tons
- Spent wash 585.5 tons
- Spent lees 85 tons
3 Dehydration - Rectified sprit 42.5 tons - Power alcohol 40 tons
- Spent lees 2.5tons

4 Spent Wash Evaporation - Spent wash 585.5 tons - Concentrated Spent wash 350
tons
- Process condensate 235.5 tons
In the lab scale it was decided to produce 4 liters of fermented wash samples for optimizing
effluent recirculation at fermented wash brix of 15(the existing Bulk production process) and
20& 25 (future proposed sugar concentration). The input material required to prepare sample is
calculated as follows:
For 15˚ brix fermented Wash
Molasses having 85˚ brix with specific gravity of 1.34 is used to prepare fermented wash with
brix of 15˚ and Specific gravity of 1.06.
- Molasses Required in kilo gram = 15%*4Lt*1.06kg/lt = 0.56 kg
85%* 1.34
- Mass of Fermented wash Prepared = 4Lt*1.06Kg/lt = 4.24 kg
Effluent generated on distilling the fermented wash was calculated based on the following
- Spent wash generated = (585/713)*4.24 kg = 3.48 kg
- Spent lees generated = (87.5/713)*4.24 kg = 0.52 kg
- Process Condensate = (235.5/ 713)*4.24 kg = 1.4 kg
- 50% Spent wash application= 0.5*3.84 = 1.74 kg
- 25% Spent wash application = 0.25*3.84= 0.87 kg
- 50% Process condensate = 0.5*1.56 = 0.70 kg

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 95
 Optimization of Molasses Fermentation to Reduce Effluent

- 100% process condensate = 1*1.56 = 1.40 kg

- 50% Spent lees recycling = 0.5*0.59 = 0.26 kg
For 20˚ brix fermented Wash
- Molasses Required in kilo gram = 20%*4Lt*1.08kg/lt = 0.76 kg
85%* 1.34
- Mass of fermented wash prepared = 4Lt*1.08kg/lt = 4.32 kg
- Spent wash generated = (585.5/713)*4.32 kg = 3.55 kg
- Spent lees generated = (87.5/713)*4.32 kg = 0.53 kg
- Process Condensate = (235.5/ 713)*4.32kg = 1.43 kg
For 25˚ brix fermented Wash
- Molasses Required in kilo gram = 20%*4Lt*1.1kg/lt = 0.97 kg
85%* 1.34
- Mass of fermented wash prepared = 4Lt* 1.1kg/Lt = 4.4 kg
- Spent wash generated = (585.5/713)*4.4 kg = 3.61 kg
- Spent lees generated = (87.5/713)*4.4 kg = 0.54 kg
- Process Condensate = (235.5/ 713)*4.4kg = 1.45 kg
Appendix C- Analysis Results and Experimental runs for Effluent Re circulation
1. For 15 Brix (8.67% fermentable Sugar Concentration) fermented wash
Table 7.2- Analysis results of Run -1
Run Factors Response Variable
% Spent % Spent % Process Process Alcohol% Yeast count Residual
Water in kg
wash or lees or Condensate fermented sugar %
kg kg or in kg wash of wash
1 1.92 0.295 1.56 0 3.87 325,000,000 3.4
2 0.96 0.295 0.78 1.645 5.16 350,000,000 1.34
3 0.96 0.59 0 2.13 4.65 175,000,000 2.7
4 0.96 0 1.56 1.16 4.55 200,000,000 2.48
5 0.96 0.295 0.78 1.645 5.25 250,000,000 1.47
6 1.92 0.59 0.78 0.39 2.92 275,000,000 2.2
7 0 0.295 0 3.385 4.74 325,000,000 1.75
8 0.96 0.59 1.56 0.57 4.76 225,000,000 1.1

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 96
 Optimization of Molasses Fermentation to Reduce Effluent

Table 7.2 continued

9 1.92 0 0.78 0.98 5.24 300,000,000 1.2
10 0.96 0.295 0.78 1.645 5.26 275,000,000 1.34
11 0.96 0.295 0.78 1.645 5.28 235,000,000 1.3
12 0 0 0.78 2.9 5.22 215,000,000 1.1
13 1.92 0.295 0 1.465 5.82 300,000,000 1.1
14 0.96 0.295 0.78 1.645 5.68 235,000,000 1.32
15 0 0.295 1.56 1.825 6.37 225,500,000 0.94
16 0 0.59 0.78 2.31 4.91 235,000,000 1.63
17 0.96 0 0 2.72 5.31 237,500,000 1.32

Table 7.3 Analysis results of Run -2

Run Factors Response Variable
% Spent % Spent % Process Process Alcohol% Yeast count Residual
Water in
wash or lees or Condensate fermented sugar %
kg
kg kg or in kg wash of wash
1 1.92 0.295 1.56 0 3.84 375,000,000 3.7
2 0.96 0.295 0.78 1.645 5.15 275,000,000 1.36
3 0.96 0.59 0 2.13 4.76 225,000,000 2.4
4 0.96 0 1.56 1.16 4.48 250,000,000 2.9
5 0.96 0.295 0.78 1.645 5.23 300,000,000 1.51
6 1.92 0.59 0.78 0.39 2.91 250,000,000 2.3
7 0 0.295 0 3.385 4.7 350,000,000 1.9
8 0.96 0.59 1.56 0.57 4.72 250,000,000 1.4
9 1.92 0 0.78 0.98 5.27 290,000,000 1.1
10 0.96 0.295 0.78 1.645 5.26 230,000,000 1.33
11 0.96 0.295 0.78 1.645 5.3 240,000,000 1.2
12 0 0 0.78 2.9 5.24 245,000,000 1.1
13 1.92 0.295 0 1.465 5.78 260,000,000 1.2
14 0.96 0.295 0.78 1.645 5.69 257,500,000 1.32

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 97
 Optimization of Molasses Fermentation to Reduce Effluent

Table 7.3 continued

15 0 0.295 1.56 1.825 6.38 245,000,000 0.92
16 0 0.59 0.78 2.31 4.9 230,000,000 1.7
17 0.96 0 0 2.72 5.28 250,000,000 1.37

Table 7.4- Analysis results of Run -3

Run Factors Response Variable
% Spent % Spent % Process Process Alcohol% Yeast count Residual
Water in
wash or lees or Condensate fermented sugar %
kg
kg kg or in kg wash of wash
1 1.92 0.295 1.56 0 3.87 350,000,000 3.7
2 0.96 0.295 0.78 1.645 5.17 275,000,000 1.35
3 0.96 0.59 0 2.13 4.72 200,000,000 2.4
4 0.96 0 1.56 1.16 4.53 225,000,000 2.57
5 0.96 0.295 0.78 1.645 5.21 335,000,000 1.52
6 1.92 0.59 0.78 0.39 2.96 300,000,000 2.1
7 0 0.295 0 3.385 4.78 300,000,000 1.76
8 0.96 0.59 1.56 0.57 4.74 230,000,000 1.4
9 1.92 0 0.78 0.98 5.27 280,000,000 1
10 0.96 0.295 0.78 1.645 5.32 200,000,000 1.29
11 0.96 0.295 0.78 1.645 5.32 215,000,000 1.1
12 0 0 0.78 2.9 5.23 260,000,000 1.1
13 1.92 0.295 0 1.465 5.86 265,000,000 1.18
14 0.96 0.295 0.78 1.645 5.7 242,500,000 1.26
15 0 0.295 1.56 1.825 6.36 207,500,000 0.9
16 0 0.59 0.78 2.31 4.95 240,000,000 1.62
17 0.96 0 0 2.72 5.31 232,500,000 1.33

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 98
 Optimization of Molasses Fermentation to Reduce Effluent

2. Effluent Recirculation At Different Sugar Concentration of Fermented Wash

Table 7.5 Analysis results of effluent recirculation at different sugar concn. Run -1
Sample of fermented wash preparation Response Variables
Run Molass Spent Spent Process Process Alcohol% Residual Yeast Cell
water in
es Wash lees Condensa wash sugar% mass
kg
used in added in kg te in kg
Kg in kg
1 0.76 1.955 0.6 0.79 0.215 7.7 1.95 227,500,000
2 0.76 0 0 0.79 2.77 6.45 1.62 247,500,000
3 0.76 0 0.6 0.79 2.17 4.17 3.95 172,500,000
4 0.76 0.978 0.3 0.79 1.492 7.05 1.3 305,000,000
5 0.56 0.96 0.59 0.78 1.31 5.23 1.51 177,500,000
6 0.56 0.96 0.295 1.56 0.825 4.9 1.32 245,000,000
7 0.76 0.978 0.3 0.79 1.492 6.68 1.55 362,500,000
8 0.97 2 0.305 0.805 0.32 6.1 4.54 275,000,000
9 0.56 0 0.295 0.78 2.565 4.65 1.98 145,000,000
10 0.76 0.978 0.6 0.79 1.192 6.8 1.29 210,000,000
11 0.97 1 0.61 0.805 1.015 7.1 3.6 250,000,000
12 0.76 0.978 0.3 0.79 1.492 7.07 1.12 325,000,000
13 0.97 1 0 0.805 1.625 5.3 7.21 145,000,000
14 0.76 1.955 0.3 1.58 0 7.07 1.4 287,500,000
15 0.76 1.955 0 0.79 0.815 6.8 1.29 297,500,000
16 0.76 0.978 0 0 2.582 6.8 1.28 282,500,000
17 0.97 0 0.305 0.805 2.32 7.75 1.08 300,000,000
18 0.76 0.978 0.3 0.79 1.492 7.1 1.13 295,000,000
19 0.76 0.978 0.3 0.79 1.492 7.18 1.2 300,000,000
20 0.97 1 0.305 0 2.125 6.67 4.5 250,000,000
21 0.56 1.92 0.295 0.78 0.645 5.9 0.86 225,000,000
22 0.56 0.96 0.295 0 2.385 5.86 0.89 287,500,000
23 0.76 0.978 0.6 0 1.982 6.89 1.4 300,000,000

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 99
 Optimization of Molasses Fermentation to Reduce Effluent

Table 7.5 continued

24 0.76 0 0.3 0.79 2.47 7.21 1.13 285,000,000
25 0.76 0 0.3 0 3.26 6.9 1.2 275,000,000

26 0.97 1 0.305 1.61 0.515 8.3 1.4 300,000,000

27 0.76 1.955 0.3 0 1.305 6.71 1.95 250,000,000
28 0.76 0.978 0 1.58 1 6.48 2.55 205,000,000
29 0.56 0.96 0 0.295 2.385 5.85 0.9 212,500,000

Table 7.6- Analysis results of Run-2

Run Sample of fermented wash preparation Response Variables

Molasses Spent Spent Process Process Alcohol Residual Yeast Cell
water in
used in Wash lees Condens % wash sugar% mass
kg
Kg added in kg ate in kg
in kg
1 0.76 1.955 0.6 0.79 0.215 7.8 1.91 228,750,000
2 0.76 0 0 0.79 2.77 6.6 1.56 252,500,000
3 0.76 0 0.6 0.79 2.17 4.19 3.91 180,500,000
4 0.76 0.978 0.3 0.79 1.492 7.11 1.2 302,500,000
5 0.56 0.96 0.59 0.78 1.31 5.28 1.51 178,750,000
6 0.56 0.96 0.295 1.56 0.825 5.1 1.29 242,500,000
7 0.76 0.978 0.3 0.79 1.492 6.6 1.63 382,500,000
8 0.97 2 0.305 0.805 0.32 6.2 4.55 262,500,000
9 0.56 0 0.295 0.78 2.565 4.65 1.97 155,000,000
10 0.76 0.978 0.6 0.79 1.192 7.1 1.2 292,500,000
11 0.97 1 0.61 0.805 1.015 7.35 3.1 287,500,000
12 0.76 0.978 0.3 0.79 1.492 7.09 1.13 312,500,000
13 0.97 1 0 0.805 1.625 5.19 7.12 162,500,000
14 0.76 1.955 0.3 1.58 0 7.05 1.3 287,500,000
15 0.76 1.955 0 0.79 0.815 6.85 1.28 292,500,000
16 0.76 0.978 0 0 2.582 6.8 1.27 285,000,000

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 100
 Optimization of Molasses Fermentation to Reduce Effluent

Table 7.6 continued

17 0.97 0 0.305 0.805 2.32 8.2 1.05 300,000,000

18 0.76 0.978 0.3 0.79 1.492 7.15 1.12 302,500,000
19 0.76 0.978 0.3 0.79 1.492 7.19 1.1 302,500,000
20 0.97 1 0.305 0 2.125 6.73 4. 260,000,000
21 0.56 1.92 0.295 0.78 0.645 5.85 0.94 200,000,000
22 0.56 0.96 0.295 0 2.385 5.93 0.84 307,500,000
23 0.76 0.978 0.6 0 1.982 6.92 1.4 297,500,000
24 0.76 0 0.3 0.79 2.47 7.2 1.12 275,000,000
25 0.76 0 0.3 0 3.26 6.75 1.34 250,000,000

26 0.97 1 0.305 1.61 0.515 8.6 1.2 262,500,000

27 0.76 1.955 0.3 0 1.305 6.72 1.9 245,000,000
28 0.76 0.978 0 1.58 1 6.55 2.3 222,500,000
29 0.56 0.96 0 0.295 2.385 5.89 0.88 225,000,000

Table 7.7- Analysis results of effluent recirculation at different FS Run-3

Run Sample of fermented wash preparation Response Variables
Molasses Spent Spent Process Process Alcohol Residual Yeast Cell
water in
used in Wash lees Condens % wash sugar% mass
kg
Kg added in kg ate in kg
in kg
1 0.76 1.955 0.6 0.79 0.215 7.9 1.9 233,750,000
2 0.76 0 0 0.79 2.77 6.36 1.65 250,000,000
3 0.76 0 0.6 0.79 2.17 4.21 3.84 172,500,000
4 0.76 0.978 0.3 0.79 1.492 7.11 1.1 292,500,000
5 0.56 0.96 0.59 0.78 1.31 5.24 1.51 183,750,000
6 0.56 0.96 0.295 1.56 0.825 5 1.29 262,500,000
7 0.76 0.978 0.3 0.79 1.492 6.67 1.62 380,000,000
8 0.97 2 0.305 0.805 0.32 5.7 4.62 212,500,000
9 0.56 0 0.295 0.78 2.565 4.65 1.98 150,000,000

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 101
 Optimization of Molasses Fermentation to Reduce Effluent

Table 7.7 continued

10 0.76 0.978 0.6 0.79 1.192 6.8 1.26 247,500,000
11 0.97 1 0.61 0.805 1.015 7.15 3.2 287,500,000
12 0.76 0.978 0.3 0.79 1.492 7.11 1.05 307,500,000
13 0.97 1 0 0.805 1.625 5.12 7.27 142,500,000
14 0.76 1.955 0.3 1.58 0 7.06 1.2 250,000,000
15 0.76 1.955 0 0.79 0.815 6.84 1.29 295,000,000
16 0.76 0.978 0 0 2.582 6.8 1.28 287,500,000
17 0.97 0 0.305 0.805 2.32 8.05 1.08 300,000,000
18 0.76 0.978 0.3 0.79 1.492 7.14 1.14 302,500,000
19 0.76 0.978 0.3 0.79 1.492 7.23 1 312,500,000
20 0.97 1 0.305 0 2.125 6.7 4.7 240,000,000
21 0.56 1.92 0.295 0.78 0.645 5.83 0.93 220,000,000
22 0.56 0.96 0.295 0 2.385 5.91 0.85 305,000,000
23 0.76 0.978 0.6 0 1.982 6.89 1.4 287,500,000
24 0.76 0 0.3 0.79 2.47 7.16 1.2 265,000,000
25 0.76 0 0.3 0 3.26 6.75 1.3 270,000,000

26 0.97 1 0.305 1.61 0.515 8.3 1.3 262,500,000

27 0.76 1.955 0.3 0 1.305 6.67 2.15 255,000,000
28 0.76 0.978 0 1.58 1 6.47 2.35 217,500,000
29 0.56 0.96 0 0.295 2.385 5.87 0.89 222,500,000

Appendix –D Yeast Counting by HeamoCytometer

Take 1ml sample in a 100ml measuring cylinder and add 1ml methylene blue solution; Then
makeup volume to 100ml by distilled water. Mix well. Take heamocytometer slide and put a
drop of diluted yeast culture solution on the heamocytometer and cover with a cover slip. The
viable yeast cells remains transparent and dead cells becomes dark blue.
- To count the cells within an area, switch over to the 10X objective lens with the 40X
objective lens, frame up the counting area of one of the 25 large squares, and take a
count. Use your hand-held counter to keep an accurate tally of the cells.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 102
 Optimization of Molasses Fermentation to Reduce Effluent

- Count the number of yeast cells found inside grid 1, 2, 3,4 and 5 and sum up.
- Total number of yeast cell is obtained considering the sum of yeast cell found in
chamber 1-5 and dilution factor as follows:
Yeast Count = total cell in grid(1-5)*5*DF*104
Grid- 1 Grid- 2
Grid- 4 Grid- 5 Grid-3

1. 2

4 3

FigA1. Cell count by Hemocytometer slide

Source: Fal Allen,1994- A practical guide to laboratory and quality control procedures for
small scale brewers.

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 103
 Optimization of Molasses Fermentation to Reduce Effluent

Appendix- E compositional analysis results of input materials and effluents

Table 7.8. compositional analysis results of molasses
Date TRS UFS FS CaO VAC P 2 O5 Brix Carmel
1 52 3.2 48.8 22100 4500 205 83 0.223
2 52.17 3.2 48.97 22150 4565 206 82.5 0.275
3 52 3.23 48.77 21008 4875 207 84 0.293
4 52.1 3.24 48.86 21150 4600 207 83 0.25
5 52.15 3.23 48.92 20150 4689 215 85 0.21
6 52.16 3.24 48.92 20115 4920 225 81 0.215
7 52.11 3.24 48.87 22150 4855 210 80.8 0.225
8 52 3 49 22086 4867 212 81.1 0.21
9 52.05 3 49.05 22250 4835 230 81 0.215
10 51.9 3.03 48.87 24650 4745 219 80 0.235
11 52 3.1 48.9 25845 4748 225 81 0.21
12 52 3.1 48.9 4645 4790 215 81.5 0.21
13 51.6 3.15 48.45 21500 4684 245 82 0.195
14 52 3.6 48.4 22815 4645 225 82 0.185
15 51.9 3.7 48.2 22140 4825 222 83 0.165
16 51.5 3.8 47.7 23679 4860 210 82.25 0.175
17 52 3.9 48.1 23546 4780 205 82 0.177
18 48 3.3 44.7 22800 4495 205 82.5 0.167
19 48.5 3.2 45.3 22100 4565 210 85 0.155
20 48.9 3.15 45.75 22200 4610 205 80.5 0.218
21 48.5 3.16 45.34 20500 4675 208 80.4 0.245
22 48 3.18 44.82 21280 4725 212 80.65 0.265
23 48.1 3.2 44.9 21950 4785 222 80.6 0.275
24 48 2.7 45.3 22500 4765 190 80.85 0.245
25 48.2 3 45.2 22100 4768 195 81.65 0.265
26 49.6 2.9 46.7 22250 4775 180 81.85 0.245
27 49.8 2.8 47 22250 4850 200 82 0.222

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 104
 Optimization of Molasses Fermentation to Reduce Effluent

Table 7.8 continued

28 49.4 2.8 46.6 22150 4885 205 83 0.213
29 49.5 2.7 46.8 22300 4800 204 83.5 0.215
30 49.5 2.7 46.8 22050 4850 206 84.1 0.205
31 50 3 47 26500 4725 205 85 0.213
32 49.9 3 46.9 24500 4754 200 83 0.165
33 51 3 48 21950 4730 202 82.6 0.178
34 51.2 3 48.2 20750 4680 204 81.9 0.153
35 51.4 3.3 48.1 20900 4735 208 81.8 0.2
36 51.6 2.7 48.9 20860 4755 200 81.6 0.206
37 51.6 3 48.6 21000 4765 204 81.7 0.218
38 51 3.3 47.7 23450 4810 200 81.6 0.225
39 51.2 3.1 48.1 21055 4825 202 81.4 0.225
40 52 3.1 48.9 20350 4875 206 80.8 0.213
41 51 3 48 20500 4850 205 80.4 0.21
42 51.3 3 48.3 22285 4825 204 81 0.167
43 51.5 3 48.5 21900 4760 200 81.2 0.215
44 51.6 3 48.6 21800 4710 202 81.3 0.208
45 51.8 3 48.8 21500 4620 200 81.9 0.200
Average 50.79 3.12 47.68 22171 4750 208 81.98 0.213

Table 7.9 compositional analysis of process water

Chlorine before Chlorine after
Date Hardness pH activated carbon activated carbon
(ppm) (ppm) (ppm)
1 100 6.8 0.5 Nil
2 110 7.1 0.6 Nil
3 98 7 0.7 Nil
4 85 7.1 0.5 Nil
5 115 7.1 0.8 Nil

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 105
 Optimization of Molasses Fermentation to Reduce Effluent

Table 7.9 continued

6 110 7 1 Nil
7 98 6.9 0.9 Nil
8 105 7 0.9 Nil
9 102 7 0.8 Nil
10 106 7.1 0.75 Nil
11 104 7 0.6 Nil
12 107 6.8 0.5 Nil
13 102 6.95 0.6 Nil
14 105 6.96 0.5 Nil
15 98 6.98 0.56 Nil
16 96 6.99 0.65 Nil
17 101 7 0.85 Nil
18 102 6.96 0.75 Nil
19 106 6.95 0.9 Nil
20 105 6.9 0.84 Nil
21 104 7 0.9 Nil
22 96 7.05 0.5 Nil
23 98 7.06 0.5 Nil
24 99 7 0.5 Nil
25 101 7 0.8 Nil
26 105 7 0.5 Nil
27 104 6.9 0.5 Nil
28 98 7 0.6 Nil
29 100 6.9 0.5 Nil
30 100 6.9 0.5 Nil
Average 102 6.98 0.67 Nil

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 106
 Optimization of Molasses Fermentation to Reduce Effluent

Table7.10 for compositional analysis of process condensate, vinasse and Spent lees
Spent wash Process condensate Spent lees
Date Brix Alcohol% CaO pH P 2 O5 Hardness pH Brix Alcohol% Hardness pH Brix Alcohol%
(ppm) (ppm) (ppm) (ppm)
1 13.4 0.116 4003 4.2 175 nil 3.7 nil 0.4 Nil 3.1 nil 0.1
2 13.2 0 4100 4.3 150 nil 3.8 nil 0 Nil 3.1 nil 0
3 13.5 0.393 3850 4.2 150 nil 3.8 nil 1 Nil 3.1 nil 0
4 13.7 0.084 3450 4 165 nil 3.6 nil 0.25 Nil 3.3 nil 0
5 13.8 0.137 4210 4 145 nil 3.6 nil 0.4 Nil 3.3 nil 0
6 13.6 0.3 4050 4.1 175 nil 3.65 nil 1.1 Nil 3.3 nil 0
7 13.7 0 4110 4.05 125 nil 3.6 nil 0 Nil 3.2 nil 0.1
8 13.5 0 4050 4.06 135 nil 3.6 nil 0 Nil 3.2 nil 0
9 13.6 0 4025 4 175 nil 3.6 nil 0 Nil 3.15 nil 0
10 13.3 0.1 4055 4 190 nil 3.55 nil 0.3 Nil 3.11 nil 0
11 12.9 0.077 4045 4 225 nil 3.6 nil 0.2 Nil 3.12 nil 0
12 13.5 0.63 3345 4.2 256 nil 3.7 nil 1.5 Nil 3.14 nil 0
13 13.6 0.43 3650 4.2 255 nil 3.8 nil 0.9 Nil 3.16 nil 0
14 13.7 0.15 3750 4.15 280 nil 3.8 nil 0.3 Nil 3.15 nil 0
15 14.3 0 4356 4 325 nil 3.7 nil 0 Nil 3.19 nil 0
16 14.2 0 4425 4 350 nil 3.7 nil 0 Nil 3.2 nil 0

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 107
 Optimization of Molasses Fermentation to Reduce Effluent

Table 7.10 continued

17 13.5 0.356 4438 4 354 nil 3.7 nil 1.3 Nil 3.15 nil 0
18 13.8 0.253 4425 4.2 350 nil 3.8 nil 1 Nil 3.17 nil 0
19 13.6 0.083 4410 4.1 358 nil 3.8 nil 0.2 Nil 3.13 nil 0
20 13.4 0 4352 4.1 360 nil 3.7 nil 0 Nil 3.3 nil 0
21 13.7 0 4354 4.2 375 nil 3.7 nil 0 Nil 3.2 nil 0
22 13.6 0 4100 4.05 390 nil 3.6 nil 0 Nil 3.2 nil 0
23 13.8 0 3954 4.05 350 nil 3.6 nil 0 Nil 3.3 nil 0
24 13.9 0 3875 4.1 305 nil 3.6 nil 0 Nil 3.25 nil 0
25 13.5 0 3800 4.08 275 nil 3.6 nil 0 Nil 3.24 nil 0
26 13.2 0 3888 4.2 280 nil 3.8 nil 0 Nil 3.22 nil 0
27 14.1 0 3925 4.08 240 nil 3.8 nil 0 Nil 3.3 nil 0
28 13.4 0 4030 4.1 250 nil 3.8 nil 0 Nil 3.1 nil 0
29 13.5 0.05 4265 4.09 235 nil 3.7 nil 0.1 Nil 3.2 nil 0
30 13.4 0.25 4000 4.08 225 nil 3.7 nil 0.9 Nil 3.3 nil 0
Average 13.6 0.11 4043 4.10 254.1 nil 3.69 nil 0.33 Nil 3.20 nil 0

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 108
 Optimization of Molasses Fermentation to Reduce Effluent

Appendix-F test procedures for alcohol quality analysis

I. Preparation of Rectified Sprit
a. Distill out the wash obtained from fermentation of molasses obtained by using the optimum
recirculation rate of spent wash, spent lees, process condensate and process water.
b.Distill out the wash obtained from fermentation of molasses by using process water alone.
c. Analyze the sprit obtained from distillation column during step a and b- for furfural, fusel oil
and aldehyde content and compare the results from food safety and quality point of view.
II. Determination of Aldehyde Content of Sprit (AOAC Method)
a. Place 100 ml of distillate in 500 ml flask ( stopper conical flask).
b. add 100 ml of water and 25 ml of freshly prepared 0.05N aqueous sodium bisulfate solution
containing 10% alcohol and allow to stand about 30 minutes on shaking occasionally.
c. Add 0.05N aqueous iodine solution with stirring until the mixture remains a definite yellow
colour, then add few drop of starch solution and titrate the excess iodine with 0.05N aqueous
sodium thiosulfate solution to colourless end points.
d. Run a blank contain 100ml of water and 25ml thiosulphate solution by adding the exact
amount of iodine solution as used for the sample.
e. Calculate as acetyl dehydrates the difference between titration in ml of sodium bisulfate
solution times 1.1 equal mg of acetaldehyde in the sample.
Aldehyde content = 0.022* difference in burette reading* Normality of Na 2 S 2 O 3
III. Determination of Ester( as CH 3 COOC 2 H 5 ) of Sprit
a. Transfer 100ml of the sample in to a 250 ml round bottom flask.
b. Add few drops pH phenolphthalein indicator and neutralize in the cold, the free acid if
present with 0.1N std KOH solution.
c. Add 2ml of std. KOH solution (0.5N).
d. Attach the flask to a reflux condenser provided with a soda- lime guards tube and reflux the
contents on water bath for at least one hour.
e. Cool the contents pour in to another flask; wash the Original flask with 100ml of freshly
distilled water, add the washing to the original liquor and then titrate with std. H 2 SO 4 adding a
few drops of phenolphthalein Indicator.
f. Carryout a blank, using 100ml of water in place of the neutralized sample.
g. Calculate the percentage of ester as (CH 3 COOC 2 H 5 ) = 0.088*(V 1 - V 2 )*N

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 109
 Optimization of Molasses Fermentation to Reduce Effluent

where, V 1 = volume, in ml of std. H 2 SO 4 required in the blank

V 2 = Volume in ml of std. H 2 SO 4 required with the sample.
N= Normality of Std. H 2 SO 4
IV. Determination of Fusel oil in Sprit
A simple fusel oil test for spirits consists of adding 0.5ml of 0.8% of aq. Solution of salicylic
aldehyde to 5ml of the sample. Then adding 10ml of conc. H 2 SO 4 at the rate of about 1ml/sec.
After 20 minutes the color of the solution is observed.
Table 3.3 Colour ranges of rectified sprit Fusel oil content in ppm
Grade Color range Fusel oil ppm
1 Lemon yellow 0 – 50ppm
2 Lemon yellow to slightly brownish 6 – 150ppm
3 Brown 350ppm
4 Red color 500ppm & above
V. Determination of Furfural in Sprit
Take 5ml of material under test in colorless glass cylinder (25ml) Dilute with 5ml of water
and after thorough mixing add 0.5ml of aniline preferably with a pipette and 1ml of glacial
acetic acid preferably with a burette. Gently, agitate the mixture, till it becomes homogenous
and then set aside for a period of 5 minutes at a temp above 15 . In case furfural is present
(more than 2 PPM) in the material a red color starts developing in course of few seconds and
reaches its maximum intensity in 5 – 10 minutes. If no red color either permanent or transient
develops during this time then the test for furfural content shall be taken to have been satisfied.
Result
In Rectified Spirit sample furfural test is positive/Negative
VI. Determination of Methyl Alcohol Content of Rectified Sprit
a. Dilute 0.5 ml of the samples of alcohol sprit with water to 5ml and add 2.0ml of
potassium permanganate solution in phosphoric acid and set aside for ten minutes.
b. add 2ml of Oxalic acid solution in H 2 SO 4
c. To the above colourless solution, of magneta and set aside at a temperature between 15
to 30˚C.
d. Examine the colour of the above sample after 30 minutes.
e. Compare the final change in colour against the blank produced by adding 5 ml of

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 110
 Optimization of Molasses Fermentation to Reduce Effluent

Magneta solution.
f. The material shall be taken as satisfying the requirements of this test if no colour change
is visible between the two solutions.
g. Results will be described in the test qualitatively on Positive or Negative.
VII. Determination of Copper in Sprit
a. Transfer 20ml of the material (Rectified Spirit) into Silica Crucible and add one ml of dil
H 2 SO 4 .
b. Heat gently in the beginning and then evaporate almost to dryness on water bath.
c. Ignite the crucible over smokeless flame to eliminate the H 2 SO 4 acid. Then cool and
dissolve the residue in 2ml of water.
d. Add four drop of aqua-ragia solution and evaporate to dryness on water bath (by taking care,
experiment can be conducted on burner flame also).
e. Dissolve the residue into 2ml of diluted HCI and warm gently till the residue is dissolved.
f. Add 0.5gms of NH 4 Cl and dilute to 15ml with distilled water. Add dil. (NH4OH) till the
solution comes alkaline.
g. Boil excess of NH 3 and filter into clean Nesceller’s tube. Cool and then render the solution
with acetic acid.
h. Then dilute to 40ml add 0.5ml of potassium Ferro cyanide solution. Stir well and make up to
50 ml volume with distilled water to make a sample tube.
i. Prepare series of control solution, each containing in 50ml of water.
• 0.5gms of NH 4 CI
• 3 – 5 drops of Acetic Acid (CH 3 COOH)
• 0.5ml of K 4 Fe (CN 6 ) together with increasing amount of standard containing solution
added in the control solution having nearby as possible; the same intensity of color as
that of the test solution.
a. Observe that the Control solution containing X ml of Std Cu solution is nearly matching with
test solution. V ml.
k. Calculate the copper content of sprit by using the following equation
Cu (Copper) gms/100ml = 0.00002845 x 12.5 x V

AAU AAit School of Chemical and Bio Engineering Food Engineering Stream Page 111