0

PREDICTION OF SUGAR YIELD FROM SUGAR CANE

USING PROCESS MODELING

BY

ISTIFANUS, AKILA BARDEY

(PG/Ph .D/08/49236)

DEPARTMENT OF AGRICULTURAL AND BIORESOURCES ENGINEERING

FACULTY OF ENGINEERING
UNIVERSITY OF NIGERIA, NSUKKA

DECEMBER, 2015
 ii

APPROVAL PAGE

PREDICTION OF SUGAR YIELD FROM SUGAR CANE

USING PROCESS MODELING
BY

ISTIFANUS, AKILA BARDEY

(PG/Ph .D/08/49236)

A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE

REQUIREMENTS FOR THE AWARD OF THE DEGREE OF
DOCTOR OF PHILOSOPHY (Ph .D) IN AGRICULTURAL AND
BIORESOURCES ENGINEERING, UNIVERSITY OF NIGERA,
NSUKKA

Engr. Istifanus, Akila Bardey Signature -------------------------- Date --------------------

(Student)

Engr. Prof. C. O. Akubuo Signature-------------------------- Date --------------------

(Supervisor)

Engr. Prof. I. N. Itodo Signature --------------------------- Date -------------------

(External Examiner)

Engr. Prof. W. I. Okonkwo Signature --------------------------- Date --------------------

(Head of Department)

Engr. Prof. E. S. Obe Signature --------------------------- Date --------------------

(Chairman, Faculty Post

Graduate Committee)
 iii

CERTIFICATION

Engr Istifanus, A. Bardey a post graduate student of the Department of

Agricultural and Bioresourcs Engineering, with the registration number

PG/Ph.D/08/49236 has satisfactorily completed the requirement for the research work for

the degree of Doctor of Philosophy of Engineering, in Agricultural processing

Engineering. The work embodied in this report is original and has not to the best of my

knowledge, been submitted in part or full for any Diploma, First or Masters degree of

this University or any other University.

……………………………………….. .....................................................

Engr. Prof. C. O. Akubuo Engr. Prof. W. I. Okonkwo

Supervisor Head of Department

………………………………………..

Engr. Prof. E. S. Obe

Chairman, Faculty Post Graduate Committee

 iv

DEDICATION

This research work is dedicated to the Most High God who has been the fountain

of my inspiration and sustenance. By His willpower I obtained the grace, peace, strength

and wisdom to research through.

 v

ACKNOWLEDGMENTS

I am indebted to God Almighty who has been my live wire and by whose

authority I exist and posses the undeserved empowerment to venture into life struggles.

My profound gratitude goes to my supervisor Prof. C. O. Akubuo for his unquantifiable

mentoring and large impartation of knowledge to me. I also sincerely appreciate the

Head of Agricultural and Bioresources Engineering, UNN Prof. Okonkwo W. I. and the

entire staff of the department for their various contributions to my life morally,

academically and otherwise.

My dearest wife Mrs. Akila Tabitha deserves very huge thanks for her prayers,

moral and mental sharing of my toils, and her resolute counsels and sacrifices all to

exchange for my ultimate success. I acknowledge my beloved children Joy, Hilary,

Benita and Hero for their sacrifices, goodwill and prayers.

I strongly acknowledge my Pastor, I. O. Ukoson and his wife, Mr/Mrs W. D.

Lamu, Evang/Mrs Edison Tata and Mr Kene (CRUDAN-Jos) my mother and my mother

in-law Mama Biham Istifanus and Mama Lydia Iliya respectively, Benson Istifanus

Bardey, Aunty Karim, Aunty F, Isma Iliya Mafindi and Nathan Iliya for their prayers,

encouragements and support during this period of my struggle.

I acknowledged the moral and mental support of Prof. J. N. Maduako, Dr Wada,

Engr /Mrs. Angoos S. A., T. N. Andeyarka, Dr Gauis T. Barnabas, Engr William Kesiki

Benidict, Dr B. S. C. Gani, Dr Bako Zinas, Barr Esther Tata, Engr M. B. Kamai, Mallam

Misa Gaji, Pastor/Mrs Ordinan Yunana Insp/Mrs James Kunku and beloved Brother

Austin Joshua. Their Various supports rendered to me in different capacities are

 vi

memorable.

I also thank the management of Bacita Sugar Company (Josepdam), Bacita,

Kwara State and National Cereals Research Institute (NACRI) Baddegi, Niger State,

Dangote Sugar Refinery, Apapa Lagos.

I indeed owe Dr Isa Akubor, James Ochua, Dr Sarki Din Sunday and Engr

Francis Obi a big thank you for their contributions in seeing that this work was a success.

This work would not have been complete without the cooperation and support of

the Management of Savannah Sugar Company, Numan especially Mr Yusuf Ayegboka

Lamidi, Sis Chinyere Emmanuel, Mr Khalid I. Bakari (Laboratory and Quality Control

Department); Engr. Sunday Olawoye, Mr Umar Gana and Mr Obidah

Yunani(Engineering Department): their patience, untiring contributions and support in

making my efforts realizable is highly appreciated.

The Provost of College of Agriculture Jalingo Dr. Elizabeth Wachap and the

management of the College sincerely acknowledged and appreciated for according me

the opportunity to embark on this programme.

My special thanks goes to the family of Pastor Felix Ilegbinijie of UTC Lagos for

their hospitality, love, kindness and great sacrifice demonstrated to me during my work

in Lagos.

My computer typesetters, the UK’s group of company Nkrumah Hall UNN have

been so wonderful.
 vii

ABSTRACT

A study was undertaken to developed a model that will be used to predict sugar and the

by-products from sugar cane. The model developed from MATLAB was used to predict

the sugar, bagasse, filter cake and molasses yield from sugar cane. The predicted values

from the model were compared to yield data obtained from the production of sugar cane

from the Savannah Sugar Company, Numan, Nigeria for 90 days. The analysis of

variance (ANOVA) at p ≤ 0.01 was used to determine if there were significant difference

in the yield predicted by the model and the measured factory yield. The F-LSD at p ≤

0.01 was used to separate the means. The model is validated where there was no

significant difference between its predicted yield and the factory-obtained yield. The

sugar cane input of 2,150.52 MT was obtained from the Savannah Sugar factory. The

corresponding imbibitions water pumped into the mixed juice was 673.12 MT. The

predicted sugar, bagasse, molasses and filter cake yield using the MATLAB model was

279.5MT (13%), 1,049.46MT(48%), 111.828MT(5.2%) and 101.1MT(4.7%)

respectively. The ANOVA showed that there was no significant difference between the

MATLAB model and the factory-based model. It is concluded that the ANOVA validated

MATLAB model for sugar yield prediction. Consequently, this model is recommended

for use in predicting sugar and by-products yields from sugar cane. Statistical analysis

using the F-LSD showed that the model satisfied such an invariant confidence in the

model, and provided support for its validity.

 viii

TABLE OF CONTENTS

Title Page i
Approval Page ii
Certification iii
Dedication iv
Acknowledgments v
Abstract vi
Table of Contents vii
List of Tables xi
List of Figures xii
List of Plates xiii

CHAPTER ONE
1.0 Introduction 1
1.1 Objectives 6
1.2 Justification 6

CHAPTER TWO
2.0 Literature Review 8
2.1 The Sugarcane Plant 8

2.2 Sugars 14
2.2.1 Classification of Sugar into Types 16
2.2.2 Reducing Sugars 19
2.2.3 Other Types of Sugar and their Sources 20
2.3 Sugar Cane Processing 22
2.3.1 Bagasse 22
2.3.2 Filter Cake 23
2.3.3 Molasses 23
2.4.0 Quality Control 24
2.5.0 Clarification of Soluble or Insoluble Impurities in Mixed Juice 25
2.5.1 Phosphate as a Clarification Agent 27
 ix

2.5.2 Flocculation 27
2.6.0 Pollution Control 28
2.7.0 Thermal Behavior of Sugar Extraction Process 29
2.8.0 The Concept of Mathematical Modeling Equations 30
2.9.0 Stages of Modeling 32
2.10.0 Classification of Models 35
2.10.1 Classification Based on time Reference 36
2.10.2 Classification Based on the Degree of Certainty 36
2.10.3 Classification Based on Structure: 37
2.10.4 Classification Based on Function or Purpose 38
2.11.0 Classification of Mathematical Models 39
2.12.0 Sugar Yields 39
2.13.0 Some Applications of Mathematical Models 40
2.14.0 Mathematical Modeling of Motion Systems 41
2.15.0 Programming for Systems Analysis 42
2.16.0 MATLABB 43
2.17.0 Application of MATLAB in Systems Modeling 44
2.18.0 Review of Existing Models Related to Sugar Manufacture 44
2.18.1 Thaval and Kent 44
2.18.2 Lauret et al 46
2.18.3 Sotudedeh-Gharebagh et tal 48
2.18.4 Saurez et tal 49
2.18.5 El-Belgiti and Vorobiev 50
2.18.6 Thaval and Kent Mass Balance Model 51
2.19.0 Mass and Energy Balance 53
3.20 Prevailing Extraction Theory 54
3.21.0 Extraction Performance Parameters 55
3.22.0 Filling Ratio 55
3.23.0 Reabsorption Factor 56
3.24.0 Imbibition Coefficient 58
 x

2.25.0.1 Status quo and Main Knowledge gap in the Study of Sugar
yield from Sugar cane 58

2.25.0.2 Main Knowledge Gap in the Study of Yields of Sugar from Sugar cane 60

CHAPTER THREE
3.0 MATERIALS AND METHODS 61
3.1 General 61

3.2 The Savannah Sugar Company, Numan 63

3.2.1 Description of Sugar Production Plant 64

3.2.2 Milling Department 64

3.2.3 Processing Department 64

3.2.4 Laboratory 64

3.3 Determination of Sugar Yield 64

3.4 Determination of Bagasse Yield 65

3.5 Determination of Filter cake Yield 66

3.6 Determination of Molasses Yield 66

3.7 Development of Models 67

3.7.1 The MATLAB Simulation Model 67

3.7.2.1 Source Code of the Model Developed for the MATLAB Simulation 68

3.7.3 Validation of the Models 69

3.8.0 Milling Process 70

3.8.1 The Milling Train 70

3.9 Analysis of Sugar Prediction Model using MATLAB Software 70

 xi

CHAPTER FOUR

4.0 RESULTS AND DISCUSSION 75

4.1 Results 75
4.2 Discussion 75
4.2.1 The Comparative Behavior of Factory Versus Predicted Sugar Results 75
4.3 Results and Discussion of Bye Products of Sugar 78
4.3.1 Bagasse 78

4.3.2 Filter cake (scum or mud) 80

4.3.3 Molasses 81
4.4 Analysis of Variance (ANOVA) Discussions 82

CHAPTER FIVE
5.0 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 84
5.1 Summary 84
5.2 Conclusion 85
5.3 Recommendations 86
5.4 Contributions to the Literature/Knowledge 86
REFERENCES 87
APPENDICES 92
APPENDIX I: Field Data of sugar Production and the by-products
obtained for 90 Days 92
APPENDIX II: Theoretical Results (data) of sugar Production and the by-
products Obtained from MATLAB Simulation for 90 Replications 95
APPENDIX III: Analysis of Variance (ANOVA) Calculations 99
APPENDIX V: World Production of Sugar Cane cont’d From table 1.2 112

APPENDIX IV: Plates of bye products taken at the savannah sugar

company, Numan 116
 xii

LIST OF TABLES

Table 1.1: World Sugar Production, Consumption, Trade and Stock 4

Table 1.2 World Sugar cane Production Quantity (tons) 5

Table 2.1: Chemical Composition of Sugar cane 9

Table 2.2: Chemical Composition of Sugarcane 11

Table 2.3 Four Other Sugars and their Sources 20

Table 2.4 List of Top Ten (10) Largest Sugar Producing Countries in the World 21

Table 2.5 Summary of Various Clarification Processes 28

Table 2.6: Yield of Brown Sugar 40

Table 4.1: Least Significant Difference Obtained from the ANOVA 83

 xiii

LIST OF FIGURES

Figure 1.0: Sugar process Flow chart 3

Figure 2.2: Modeling Stages 32

Figure 3.0: Block Diagram of Sugar Manufacture Process of Savannah

Sugar Company 63

Figure 4.1: Sugar Comparison Curves Between Field and Model

Predicted Values 76

Figure 4.2: Bagasse Comparison Curves between Field and Model Predicted
Values 79

Figure 4.3: Comparison Curves of Filter Cake Field and Model Predicted
Values 80

Figure 4.4: Comparison Curves of Molasses Field and Model Predicted

Values 81
 xiv

LIST OF PLATES

Plate 1.0: The Structure of Sugar cane Plant 2

Plate 2.1: Mature Sugarcane Plants 13

PLATE 2a : Bagasse at Savannah Sugar Company, Numan 116

Plate 2b: Bagacillo at Savannah Sugar Company, Numan 117

Plate 2c: Filter cake Savannah Sugar Company, Numan 118

Plate 2d : Sugar Produced at the bagging room of Savannah Sugar Company,

Numan 119
 1

CHAPTER ONE

1.0 INTRODUCTION

Sugar, particularly edible sugar is a global item found in the recipes and menus of

the diets consumed in almost every home. It is a major product of sugarcane processing.

Sugar cane contributes well about 100% of all the sugar manufactured in Nigeria.

However, sugar can also be manufactured in other parts of the world from other plants

such as sugar beets (Atiku, 1999).

Industrial cultivation and processing of raw and refined sugar in Nigeria is

currently being undertaken by Savannah sugar company, Numan; Bacita sugar company

(now Josepdam Sugar Company), Dangote and Bua refineries in Apapa Lagos. These

companies import raw sugar and manufacture white sugar from it to complement the

requirements demanded by the Nigeria populace.

The process of manufacturing sugar from sugarcane is a very interesting subject

given the merits of this exercise. It presents us with the advantages of realizing the

production of the primary product as well as bagasse, filter cake, molasses, and so on. Of

greater interest still is the need to have an instrument through which the sugarcane

weighed to be grinded can be used to predict the end sugar that it can yield as well as the

amount all the important bye products realizable.

Process modeling is an integral part of any process industry and is undertaken to

simulate how things are done. The process model gives a description or prediction of

what the process looks like. Developing such models requires meticulous knowledge of

the process. The sugar industry is a process industry and various models have been

developed to represent the different unit operation used in the industry. The milling

1
 2

process is primarily a unit operation used to extract juice from sugarcane. Several models

have been developed to simulate the process. This thesis deals with the development of

an advanced and comprehensive model for a raw sugar factory.

Developing leaves and internodes develop in a basipetal direction in that the leaf

blade expands from the tip to the base then the internode elongates. The node consists of

a growth ring or intercalary meristem, the root band (containing root primordia) and a

bud above the leaf scar where the leaf sheath attaches, which delimits the node from the

internode below(Bull,2000). The plate below(plate 1.0) shows the structure of sugar cane

plant.

Plate 1.0: The structure of sugar cane plant

Nigeria’s sugar refining capacity, estimated at 2.1 million tons, exceeds the

country’s current total demand of 1.45 million tons. The country’s sugar refineries

depend almost exclusively on brown sugar from Brazil at five percent duty. This situation

has assisted with promoting investment in sugar refining rather than in production so far.
 3

Figure 1.0: Sugar process flow chart

. Source: Indian Sugar Mills Association(ISMA)2010 http://www.indiansugar.com/Sugar.aspx

Nigeria is a large consumer of sugar but the country’s sugar industry still depends on raw

sugar imports. Nigeria’s overall sugar consumption in May 2011/12 was estimated at

1.32 million tons. Raw sugar imports were estimated at 1.33 million tons with ending

stocks of 75,000 tons. Only 100,000 tons of refined sugar was imported for that year.

There is also no competing High-Fructose Corn Syrup (HFCS) in the market. The
 4

forecast is based on population growth as well continued industrial demand. Sugar use in

industrial activities such as manufacturing soft drinks, pharmaceuticals, biscuits, other

beverages and confectionary products demand is steady, while demand for direct

household consumption remains firm. Soft drink production alone accounts for about half

of total industrial usage (NSDC). Table 1.1 below shows the world production and

consumption of sugar (2014-15).

Table 1.1: World Sugar Production, Consumption, Trade and Stock

Particular 2012-13 2013-14 2014-15 2015-16*

World 1774.86 1757.03 1743.08 1734.05
Production
Human 1656.64 1687.34 1706.00 1734.13
Consumption
Total Import 516.97 504.81 515.5 528.67
Total Export 544.90 559.13 541.5 558.12
Ending Stock 439.78 455.15 422.81 405.29

Source: United States Department of Agriculture, (USDA)

*Estimated

Dangote Sugar Refinery PLC is Nigeria’s sugar producer. Nigeria’s consumption of

sugar continues to rise, with consumption estimated at 1.34million tonnes (basis 2013), as

an emerging class of consumers creates a bigger market for manufacturers and sellers of

FMCG products. This makes Nigeria the second-largest consumer of Sugar in Africa,

after South Africa. However at 9.5kg, per capita sugar consumption still very low in

Nigeria, compared to South Africa’s 36kg and a global average of 32kg (Dangote Sugar

Refinery Plc, 2014). According to the National Sugar Development Council (NSDC),

Nigeria has a land potential of over 500,000 hectares of suitable cane fields that can

produce over 5 million metric tons of sugarcane that when processed, can yield about 3

million metric tons of sugar. However, the sector has been neglected and depends almost

totally on refining imported raw brown sugar from Brazil worth over $500 million. Table
 5

1.2 below shows sugar cane production and consumption in 109 counties including

Nigeria.

Table 1.2 World Sugar cane Production Quantity (tons) 2013-2015

Country Rank Value as number
World - 1,911179.775
Brazil 1 768,090.444
India 2 341,200.00
China 3 128,850.908
Thailand 4 100,096.000
Pakistan 5 63.749.900
Mexico 6 61,182.077
Colombia 7 34,876.332
Indonesia 8 33,700.000
Philippines 9 31,874.000
United States 10 27,905.943
Australia 11 27,136.082
Guatemala 12 26,334.667
Argentina 13 23,700.000
Vietnam 14 20,131.089
South Africa 15 18,000.000
Egypt 16 16,100.000
Cuba 17 16,1000.000
Peru 18 10,992.240
Burma 19 9,650.000
Bolivia 20 8,065.889
Source: Factfish Sugar cane, 2015 http://www.factfish.com/country

This table continues in appendix V.

Five models which include Thaval and Kent (2012a) juice model; Lauret et al (2000)-

Hybrid Model; Sotudedeh- Gharebagh et al (2009) Model; Saurez et al (2011) model; El-
 6

Belgiti and Vorobiev(2005) Model; Thaval and Kent (2012b) Mass Balance Model and

Thaval and Kent(2012). Modeling the flow of cane constituents in the milling process

were studied. The MATLAB model which was ultimately developed and used, had the

advantage of predicting commercial or white sugar yield from sugar cane which all these

other sugar models did not attempt to do, and could not have giving the expected results.

Details of their discussions are given in section 2.18.1 to 2.18.2.

1.1 OBJECTIVES

1. To develop and validate a model for estimating sugar, bagasse, filter cake and
molasses from sugarcane.
2. To predict sugar yield using the model.

1.2 JUSTIFICATION

Sugar is a major part of the diets of most families in Nigeria and indeed the world

at large. This therefore, underscores the need for sugarcane which is the region’s major

source of table sugar to be processed into sugar. Due to increases in the standard of

living, there is a corresponding increase in the demand for sugar. Atiku (1999) revealed

that Nigeria imported 43,000 tons of sugar as at 1955 as the largest consumer in West

Africa. This prompted the Federal Government to commission some sugarcane experts in

1956 to survey the country and identify suitable areas for commercial sugarcane

cultivation and processing into production of sugar. .

The ever increasing demand for sugar necessitates improved cultivation systems,

genetic modification, and increasing land areas for cultivation in order to keep pace with

increasing demands. This effort needs be matched with a corresponding technology for

determining the yield of sugar from cane processing.

 7

With the several decades and long history of sugar production in the Nigerian

industries and refineries, there are no information in the literature on improvement on the

mode of production of sugar. These industries still rely on empirical methods of weighing

the sugar and the by-products in order to know the amount of sugar produced: hence the

need for this model for predicting yields from the production of sugar in engineering

management tool for assessing the process, which will indicate the low performing units

requiring improvements without committing materials into the production.

 8

CHAPTER TWO

2.0 LITERATURE REVIEW

2.1 The Sugarcane Plant

Sugarcane is the world's largest crop. It was cultivated on about 23.8 million

hectares, in more than 90 countries, with a worldwide harvest of 1.69 billion tons. The

world demand for sugar is the primary driver of sugarcane agriculture. Cane accounts for

80 per cent of sugar produced; most of the rest is made from sugar beets. Sugarcane

predominantly grows in the tropical and subtropical regions, and sugar beet

predominantly grows in colder temperate regions of the world (Rediff, 2015).

The sugarcane is a thick, tall, perennial plant that flourishes in tropical or sub-

tropical regions (Vermeulen, 2006). Busari and Nayagari (1999) described sugarcane as

a raw material for sugar production and as a versatile crop that can do quite well in

virtually all agro ecologies of Nigeria. The sugar cane plant can be divided into different

parts as well as different chemical components of each part. The cane plant is

distinctively made up of three parts, namely the roots, stalk and leaves. The chemical

components of these parts are shown in Table 2.1 below. It is important to note that the

cane that is supplied to the factories contains a large proportion of extraneous matter.

8
 9

Table 2.1: Chemical composition of sugar cane

Percent stalk percent roots percent leaves

Water 74.96 68.79 74.38
Ash 0.64 1.87 2.23

Fat and wax 0.38 0.54 0.69

Nitrogenous bodies 0.58 1.59 1.70

(Crude cellulose) 4.86 9.58 9.18

Fibre (pentasans) 3.04 7.04 5.49

Ligneous bodies 2.14 4.25 4.13

Sugars etc 13.40 6.34 2.21

Source: (Antoine, 2000)

The composition actually vary according to varied climate, soil conditions and age while

the information on Table 2.1 remains generic.

From sugar manufacturing point of view cane used for processing should be of

variety suited to the climate and soil condition in which it is grown and should be

harvested at the right age in order to have the highest possible sucrose content.

The cultivation of sugarcane for the production of sugar is believed to have

begun in the Indus valley on the Indian sub-continent (Okusanya and Ajogbasile, 1980).

The wild canes (saccharam barberi seretha group) of Northern India and Pakistan which

were known to be in cultivation arrived. Africa by 710AD and sugar refining was first

established in Egypt. Sugarcane is mainly grown in the alluvial soil of Kaduna and Kano

states although it is also grown on a small scale in other states of Nigeria.

 10

In a further generalized understanding, Branes (1974) also classified sugar cane as

having the constituents as shown in Table 2.2 below. In the process of cane breeding,

according to his findings, new varieties are produced and tested in a constant search for

improved characteristics. This process has been a major factor in improving productivity

in the sugarcane industry (Rein, 2007). The table below shows a typical chemical

composition of cane.
 11

Table 2.2: Chemical composition of sugarcane

Constituent Value (% by mass)

Water 69 – 75

Soluble solids (Brix)

Sucrose 8 – 16

Reducing sugars 0.5 – 3

Organic matter 0.5 – 1

Inorganic compounds 0.2 – 0.6

Nitrogenous bodies 0.5 – 1

Insoluble solids

Fibre (dry), lignin, cellulose 10 – 16

Dirt (soil, extraneous matter) 0.78 – 1.63

Source: (Barnes, 1974)

Composition of Juice:

Sugarcane is by definition a combination of juice and fibre. The mixture of brix

and water constitutes the juice of the sugarcane. Brix refers to the water-soluble solids in

the cane and includes the sugar. Technically, brix is the concentration of a solution of

pure sucrose in water having the same density as a sample of juice at the same

temperature (Bureau of Sugar Experiment Stations, 1984). Brix typically constitutes

about 17% of the cane. (Thaval, 2012). The density of the juice is a function of the brix

of the juice and is approximately 1080kg/m3 (Bureau of Sugar Experiment Stations,

2001). Soluble solids in juice include Sugars, Salts, Organic non-sugars which are made

up of several compounds each (Walford, 1996).

 12

Composition of fibre:

Fibre is the dry, water-insoluble matter in the cane (Bureau of Sugar Experiment

Stations, 2001). It typically contributes about 14% to 19% of the cane and includes any

dirt, soil and other insoluble extraneous matter as well (Thaval, 2012). The density of

fibre is approximately 1530 kg/m3 (Pidduck, 1955). The percentage of sugar in the cane

varies from 8 to 16% and depends to a large extent on the variety of the cane, its

maturity, soil condition, climate and agricultural practices. According to Foster (1956)

the weight of hygroscopic water (brix free cane water) is typically 25% of the weight of

fibre. (Thaval, 2012).

 13

Plate 2.1: Mature Sugarcane plants

The composition of the cane stalk is of major importance to the sugar processing

industry. The cane stalk can be classified into three namely, the hard outer rind, which

consists of a strong outer cuticle often covered with a layer of wax prevents evaporation

of water from the inner cells as well as giving the stalk protection against mechanical

injury and attack by micro-organisms (Antoine, 2000); Next is the soft tissue made up of

thin walled cells or parenchyma in which the sweet juice of the cane is stored. The fibres

are known as fibro-vascular bundles. They consist of the wood vessels, sieve tubes and
 14

companion cells by thick walled fibres. The stem of sugarcane is the major sink for

photosynthate (sucrose) within the sugarcane plant, rather than fruit or seed structures.

Transverse cross section through an internode reveals vascular bundles surrounded by

parenchyma cells with a thick outer epidermis covered in an external layer of wax.

2.2 Sugars

In chemistry the term ‘sugars’ denotes organic components having similar

composition and functions. They are carbohydrates of the general composition

Cn(H20)n. Commercial sugar in its raw form contains about 98.5% Sucrose(C12H22O11),

whereas refined sugar contains about 99.5%. These are extracted by processes of

clarification, evaporation, crystallization and centrifuging. Hence, the molecular weight

of sucrose is 342,296; it is very soluble in water and the solubility of sucrose, which is

directly influenced by sucrose losses in final molasses (Antoine, 2000).

According to Vermeulen (2006), sugar is a broad term applied to a large number

of carbohydrates present in many plants and characterized by more or less the taste. The

primary sugar, glucose, is a product of photosynthesis and occurs in all green plants. In

most plants, the sugars occur as a mixture that cannot be readily separated into the

components. In some plants, sugars are condensed into syrup. Juices of sugarcane

(Saccharum officinarum) and sugar sweet (Beta Vulgasis) are rich in pure sucrose,

although, sweet sugar is generally much less sweet than cane sugar.

Atiku (1999) asserted that of the many types of sugar, sucrose also called

saccharose or cane sugar is the most important of them all. Sucrose is a disaccharide

consisting of two monosaccharide components: D-glucose and D-fructose. The exact

 15

chemical name of sucrose is L-D-gluco-pyranaxyl –B-D-fructofuranaxide and the

chemical structure is as follows:

H- C
CH2 OH
O
H - C - OH
C

H - C - H OH –C - H
O

H - C - OH
H- C - OH
O

H - C
H - C

CH2

CH2 OH

[C6 H11 05] - O - [C6 H11 05] = C12 H22 05 (Sucrose) sucrose differs from

the other sugars by its ability to crystallize easily. The sucrose is produced by a process

of nature in the plant. During its early life, the sugar cane plant only contains non-

crystallisable sugars; these are sugars which cannot be formed into sugar crystals under

normal factory conditions. As the plant matures, these non-crystalilisable sugars, water

and the green matter in plants, called chlorophyll react with each other in the presence of

sunlight to form the crystallisable sugars which the factory can extract from the plant. It

is the job of the sugar factory to separate the sucrose from the many other materials,

which also form part of the plant. Basically sugar is treated as the carbohydrate especially
 16

found in fruits and vegetables. The process in which the plants transform the sun energy

into food and in the process sugar plays a great role. In the commercial level production,

sugar-beets and sugarcane are the major sources used as the raw material in production

process (Arrascaete and Friedman, 1987).

2.2.1 Classification of Sugar into Types

“Practical Action Technical Brief” (undated) identifies processed sugar as emerging

in two forms: non-crystalline and crystalline of which there are two basic types;

centrifuged and non-centrifuged. The different forms of sugar are produced in many

different countries and often have different names, therefore for clarity the different types

are described as follows:

a) Non-crystalline sugars

Syrups A non-crystalline liquid of high viscosity (thickness) concentrated from whole

cane juice. It can vary from golden brown to dark brown and contains; up to 50%

sucrose, high levels (up to 20%) of invert sugars, up to 20% moisture and the remainder

is made up of other insoluble matter (ash, proteins, bagasse etc).

b) Crystalline sugars

These can be divided into two types: non-centrifugal sugars and centrifugal

sugars. Non- centrifugal sugars are basic lump sugars where the molasses and crystals

have not been separated. Centrifugal sugars are free flowing granular sugars where the

molasses and crystals have been separated to some degree.

c) Non-centrifugal sugars

Lump sugars These sugars are a concentrated product of the cane juice and are produced

in many countries for direct consumption. They vary from yellowish brown to dark
 17

brown (almost black sometimes) in colour and contain up to 80% sucrose with the

remainder made up from moisture, invert sugars and other insoluble matter such as ash,

proteins and bagasse fines in varying proportions.

Lump sugars are produced in many countries around the world and are known by a range

of names: jaggery in Africa, gur in India and Bangladesh, desi in Pakistan, chancaca in

Peru; other names include panela, piloncillo, and rapedura.

d) Centrifugal sugars

Khandsari a basic raw granular sugar, developed in India that has been separated from

most of the molasses. Khandsari varies in colour from golden yellow to brown and

contains between 94 and 98% sucrose.

It is produced by the small to medium-scale sector and has a considerable market

in India. At its most basic, khandsari is manufactured using simple animal-drawn

crushers, is subjected to simple clarification, boiled to the consistency of a thick syrup,

and allowed to stand until sugar crystals are formed. The small crystals are then separated

in manually operated centrifuges and sun dried.

At the other end of the scale the production plant can use diesel or electrically

driven crushers, crystalliser to ensure uniform formation of crystals, power-driven

centrifuges, and forced-air driers to dry the product. Factories processing between one

and two hundred tonnes of cane per day are common, yielding between 6.5 and 13 tons of

sugar per day respectively (Factfish Sugar cane, 2015).

Since the late 1950s traditional khandsari production has been in decline in favour

of a modified process known as open pan sulphitation (OPS) sugar processing. The OPS

production method produce an off-white to white crystal sugar which can be of the same
 18

quality as that produced by modern large-scale VP sugar factories (Practical Action

Undated).

e) White granular sugars

Free flowing white granular sugars are often referred to as plantation white. These

sugars are traditionally produced in large-scale VP factories. It is possible for the

medium-scale sector to develop small-scale or mini VP plants for the production of good

quality white sugar.

f) Brown granular sugars

There are two categories of granular brown sugar: those produced directly from

the cane juice at the place of origin and those that are produced during the refining of raw

sugar. The first type includes demerara, muscovado and turbinado sugars. The second

type are coated brown or 'soft' sugars and manufactured demerara (Ramos, 1997).

Those produced directly from the cane juice at the place of origin can be made

using medium- scale open pan production methods. The refined brown sugars, however,

tend to be produced in modern large-scale VP sugar factories.

g) Demerara sugar

Named after the area in Guyana where it was first produced, demerara is a

centrifuged sugar prepared from the first crystallisation of cane syrup and has large

yellow crystals and a slightly sticky texture. Production of this sugar is suited to the

medium-scale sector as the juice needs to be carefully clarified to ensure purity and

crystalliser are required to ensure uniform grain size.

 19

Muscovado

Also known as Barbados sugar, muscovado is the product of the third

crystallisation. It is dark brown in colour with small grains and sticky texture. A by-

product of both the OPS and mini-VP scale of production, muscovado tends to be

produced as an alternative to white sugar if the standard is not very high.

Lump sugars can be produced using relatively basic low-cost equipment consisting of a

crusher and a furnace with boiling pan. Granular sugar production also uses crushers and

boiling equipment but requires additional equipment such as a clarification plant,

crystallisers, driers and packaging equipment.

2.2.2 Reducing sugars

This according to Chen and Chou (1993) implies chemical losses in which

sucrose becomes non sucrose. It is a state of change resulting from inversion and /or

carmelization. Inversion takes place under acidic conditions (PH below 7.0),

The most abundant non-sucrose components in cane are the monosaccharide; glucose and

fructose, also known as reducing sugars (Rein, 2007). The major concern with modeling

reducing sugars is the potential for inversion of the sucrose into reducing sugars. The

amount of inversion that occurs is not well known. Van der Pol and Alexander (1955)

reported that the reason for inversion of sucrose is the destruction of sucrose by enzymes.

These authors stated that the possible losses by the inversion are due to the combined

effect of temperature and pH, which are small under normal operating conditions and can

be neglected. Rein (2007) supported this view.

 20

2.2.3 Other Types of Sugar and their Sources

Although there are various types yet from these types four are most important and

popular such as Glucose, Lactose, Sucrose and Fructose. These types can be obtained

from different sources such as fruits, vegetables and milk etc. Glucose and fructose are

the types that are mostly found in Fruits, Honey and vegetables which are shown in the

below in table 2.3.

Table 2.3: Four other sugars and their Sources

Glucose Fructose Sucrose Sugars total

Source Content Content Content content
(g/100g) (g/100g) (g/100g) (g/100g)
Raw carrot 2.3 1.9 3.2 7.4
Apple 1.7 6.2 3.9 11.8
Banana 4.8 5.0 11.1 20.9
Cherry 5.9 5.3 0.2 11.4
Grape 7.6 7.8 0.1 15.5
Orange 2.2 2.4 3.9 8.5
Apricot(dried) 20.8 10 12.6 43.4
Honey 34.6 41.8 76.4
Source: Rein (2007)

· Sucrose is the kind what you can obtain in your sugar’s bowl.

· Lactose is another type and mostly found in milk and all milk products. All

these types occur naturally.

There is no country where sugar is not being used. Most of the nations produce

sufficient amount of sugar for their needs but still some countries export sugarcane

to meet there needs. A recent report revealed that about 120 countries are

producing it either on large-scale or only a small-scale just to meet their own

needs. Every year about 165 million tons of sugar is being produced all around the
 21

globe. More than 80 percent of world’s entire sugar is obtained from the sugar

cane grown in the tropical regions of the world. Rest of the production is obtained

from the sugar beet which accounts for 20% of world’s entire yield. Such type is

produced in the temperate regions of the world. Recently USDA prepared a list of

top ten countries that are contributing to near about 75% of world’s overall

production. From which Brazil is the largest producer producing 25% of world’s

overall sugar output.

The list of ten top most leading countries in sugar production is shown in table 2.4

below:

Table 2.4 List of Top 10 Largest Sugar Producing Countries in the World
Production
Rank Country
(1000 Tons)
1. Brazil 455,291
2. India 281,170
3. China 100,684
4. Mexico 50,597
5. Thailand 47,658
6. Pakistan 44,666
7. Colombia 39,849
8. Australia 38,169
9. Indonesia 30,150
10. USA 26,835
Source: USDA Foreign Agricultural Service (http://www.whichcountry.co/which-

country-is-the-largest-producer-of-sugar-cane-in-the-world/)
 22

2.3 Sugar Cane Processing

Antoine (2000) described sugar cane as having the most difficult photosynthetic

mechanism among all commercially grown crops in that it fixes around 2 to 3 percent of

radian solar energy in the form of green bio mass from atmospheric carbon dioxide and

water.

In every 1,000kg of cane biomass, the millable cane weight is around 500kg and

the rest consists of cane trash, tops and leaves considered as waste. The amount of waste

delivered to the mill depends on the degree and type of mechanization of harvesting

operation.

Other than sucrose, which happens to be the primary product for sugar

manufacture, more than 100 known products can be derived from cane at various stages

(Atiku, 1999). Some of these by-products include:

2.3.1 Bagasse – Is the fibrous residue of the cane stalk after crushing and extraction of

juice. In 1000kg of cane produces some 750-300kg of bagasse, which consists 49%

water, 4.8% fibre and 25% dry matter (sucrose and non sucrose). This is used to generate

steam and energy required by the factory. Surplus bagasse, due to its low density

(160kg/m3) and relative inflammability, is a cumbersome material to handle and

transport. It is generally used to produce paper board and electricity.

The importance of bagasse as a raw material and renewable source of energy

cannot be over emphasized. Drying bagasse using the heat from the chimney and

palletizing it really enhances its energy value.

Arrascaete and Friedman (1987) reported that drying bagasse using its own

combustion gasses has several very important advantages which include:

 23

§ Decrease in fuel consumption between 10 and 20% depending on the temperature

of the boiler stack gases.

§ Reduction of air pollution from values of about 10, 000mg/Nm2 of ash or less.

§ Reduction of losses in storage of moist bagasse reported to be as high 25%

§ Increase in density of pure-dried bales (3 times) and briquettes or pellets, 12-15%

moisture content, (6 times) to reduce storage and transportation costs.

2.3.2 Filter Cake

This consists of the impurities contained in the Cane juice which are precipitated

by treatment with lime and heat after removal by filtration. They form a cake known as

filter mud or scum. About 30-35kg of filter mud is produced per ton of cane. Filter mud

consist of 80% water and 0.9 to 1.5% sugars. It can be spread directly on agricultural

fields or stored in intermediary silo for later use as a soil conditioner. It has a large

biochemical oxygen demand (BOD) load (Antoine, 2000).

2.3.3 Molasses

This is the residual syrup from which non crystalline sucrose can be obtained following

evaporation, crystallization and fugalling of massecuite. Between 27kg to 40kg of

molasses are produced per ton of cane. Its average composition is 20% water, 35%

sucrose, 20% reducing sugar, 15% sulphated ash and 10% others. Molasses is mainly

used as animal feed or transformed into rum; alcohol or ethanol fermentation and

distillation (Cock and Carlos, 1995; Antoine 2000).

 24

2.4.0 Quality Control

The quality of granulated sugar has to conform to the standards established by the

National Food Processors Association and Pharmaceutical Industry. The quality control

process begins with the raw sugar cane before loading into the mill. Mill sanitation is an

important factor in quality control measures. Bacteriologists have shown that a small

amount of bagasse can infect the whole stream of warm juice flowing over it. Modern

mills have self cleaning troughs with a slope designed in such a way that bagasse does

not hold up but flows out with the juice stream (Vermeuden, 2006).

New analytical technologies are applied to improve the quality of industrial

processing, and the laboratory keeps abreast of standardization programs being developed

in some countries or all other institutions (Ramos, 1997). Ramos (1997) further specified

the features of cane laboratory analysis to include the following:

§ Analyzing primary juices for, apparent and real sucrose and reducing sugars.

§ Non sucrose contents, using moist disintegration and dry processing or hydraulic

press

§ Sucrose, starches, total polysaccharides, dextrants and colour precursors such as

amino nitrogen and polyphones.

The quality of sugarcane varies with age, increasing towards an optimum, and

then gradually declining. Once the cane is cut, deterioration may also begin almost

immediately. Deterioration may also begin before harvest in pest-ridden cane or in fields

affected by fire, freezes or wind storm (Clarke and Godshell, 1987).

 25

After cutting, sugarcane begins to lose water, giving an apparent increase in

sucrose content. At the same time, the activity of some enzymes such as invertase

becomes more evident as some sucrose is inverted to reducing sugars.

According to Yates, (1996), rainfall affects cane quality both from its long-term

effect in affecting growth rates, and in the short term through interfering with the

logistics of harvesting. Improvements of cane quality are the most important single factor

in maximizing profitability within the sugar industry.

2.5.0 Clarification of Soluble or Insoluble Impurities in Mixed juice

According to Bennett (1957), organic salts, consists of potassium sulphate (K2SO4 )

mixed with other essential plant components such as sodium, calcium, phosphate and

chloride. Iron, because of its involvement with colour forming reactions, is also

important..

Organic acids exists naturally in the cane plant. Of these, aconitic acid is the major

constituent, but succinic, glycolic, malic, citric, tartaric and others are also present in

trace quantities. Acetic and lactic acids are not natural constituents of sugar cane but are

formed by microbial infection. Many of these acids form insoluble calcium salts during

clarification and are therefore not removed by this process. Others form salts which are

later deposited as scales in the evaporators, eg in the case of aconitic and oxalic acids

which form calcium aconitate and oxalate. Amino acids and their amides form calcium

salts during clarification, which are not very insoluble, therefore, most will be transported

through the manufacturing process and will contribute to molasses formation.

Insoluble impurities, Antoine (2000), exist in the following forms:

Ø Field soil will settle in clarifiers;

 26

Ø Bagaccillo is caused to burst during flashing and to sink with the solids in the

clarifiers.

Ø Cane wax originates in nodes and in rind. It is physically removed with the muds

during clarification, the portion which escapes removal persists throughout the

whole manufacturing and refining process.

Ø Cane starch is not soluble in cold water but it becomes partially soluble in hot

water, and is washed out of sugar cane with hot imbibitions water. Some starch is

removed in factory clarification, but the remainder may be sufficient to retard

crystallization in the syrup and lower molasses exhaustion.

Defecation – Clarification by heat and lime, known as the simple defecation process is

the universally adopted method of cane juice purification in raw sugar manufacture.

There are many variants of the use of heat and lime (cold liming, intermediate liming and

hot liming), however, they all consist of all the following (Antoine, 2000):

· Heating – the juice is heated during which certain organic constituents e.g.

proteins are coagulated. Heat accelerates the reaction rate between the juice acids

and the lime which will be subsequently added. Note that heat also destroys

micro-organisms and enzymes preventing loss of sucrose by microbiological

activity.

· Liming – juice PH is changed from slightly acidic to neutral/slightly alkaline by

the use of milk of lime or saccharate. It causes coagulation of some colloids, and

forms a heavy precipitate of complex composition containing insoluble lime salts,

coagulated albumin and varying proportions of the fats, waxes and gums. The
 27

precipitate carries with it most of the finely suspended material of the juice that

has escaped mechanical screening (Chen and Chou, 1993).

· Flashing- the juice is suspended to 103 to 1050C, and then allowed to flash to the

constant boiling point at atmospheric pressure. This forces out the air present in

the juice and also causes bagacillo particles to burst and sink with the solids.

Flashing results in a uniform and constant temperature in the limed juice, thus

preventing thermal currents in the clarifiers and promotes good settling.

2.5.1 Phosphate as a clarification agent

According to Bennett (1975), the phosphate content of any juice is the most

important factor in efficient clarification. If the cane is grossly deficient in natural

phosphate or is otherwise very difficult to clarify, phosphate may have to be added before

liming. Nowadays, liming technique can overcome most cane deficiencies, and

satisfactorily clarify juice with neutral phosphate levels down to half or less of the

normally accepted requirement of 300 ppm pentaoxophosphate (P2O5).

When sugar solution containing soluble phosphate comes into contact with excess

calcium ions, an amorphous calcium phosphate is formed, which crystallizes as octa-

calcium phosphate andhydroxyapatite, the two- stage precipitation first causes small

particles to form which grows and rearrange into a very intricate floc that entraps and

adsorbs other non-sugars that are precipitated by the reaction change by the heat by the

calcium and by the increase in PH.

2.5.2 Flocculation

A number of synthetic water – soluble polymers bearing various trade names have

come into factory use. The action of these polyelectrolytes is to produce secondary
 28

flocculation. Generally, a good flocculation increases settling rate reduces mud volume,

decreases pol in cake, and most importantly increases the clarity of the clarified juice.

Factors that affect the performance of the polymers in juice clarification are: the time of

agitation of the stock solution, the solvent ionic strength, the polymer solution

concentration of calcium in clear juice (Bennett,1975). All these clarification processes

are shown as summary in table 2.5 below:

Table 2.5 Summary of Various Clarification Processes

Process Reagent used Precipitate Type of sugar

formed made
Defecation Lime+(phosphate)+flocculant Calcium Raw Sugar

phosphate

Sulphitation Lime + SO2 gas Calcium Sulphite White and

Refined sugar

Carbonation Lime +CO2 gas Calcium Refined sugar

carbonate

Phosphate- Lime +phosphatic acid Calcium Refined sugar

defecation phosphate

Source: Antoine, 2000.

2.6.0 Pollution Control

The process of extracting and purifying sugars from sugarcane often leads to

certain emissions of effluents from waste waters and combustion materials from

evaporators. In this era of increased awareness on environmental protection, and so

 29

pollution prevention and control in the factory is certain to prevent any further

vulnerability and degradation of the environment.

Generally, pollution can be categorized into two basic categories as follows:

i. Internal pollutants affecting the factory environment, and,

ii. External pollutants, affecting the neighborhood community.

Factory effluents constitute mainly of bagasse, filter cake, lime, molasses, solid

sugar, steam/vapour (heat sources), muds, noise, fumes, flyash, oil, grease, sugar (in

solution) entrained in various waste waters and toxic gases emanating from welding.

2.7.0 Thermal Behavior of sugar extraction process

Temperature variations influences the process of crystallization; “Based on

kinetic experiments (Smejkal, Bagherzadeh and Schick, 2008) an influence of reduced

temperatures on the energy management in sugar production was investigated. Lower

temperatures in evaporation unit resulted in decreased crystallisation temperature.

Therefore, the crystallisation rate is decreased.

The crystallisation rate as a function of temperature was studied using isothermal

crystallisation experiment. The results show that reduced crystallisation rate due to

decreased temperatures can be compensated by slightly increased supersaturation in

crystallisation unit. Moreover, already known linear dependence of mother liquor colour

and sugar colour in solution was confirmed.

On contrary to generally adopted meaning of enlarged steam consumption by temperature

decrease in evaporation it can be shown that by simultaneous reduction of the

temperature in evaporation and crystallisation, the steam consumption can be even slowly

reduced.”
 30

According to Hartel(2004), in solutes, the driving force increase as the difference

in concentration of solute above the solubility concentration increases. However,

crystallization rates do not necessarily increase as the super saturation driving forces

increase, especially in systems that easily form glassy states. In this case, as concentration

increases (or temperature decreases) and approaches the glassy transition region, mobility

of the molecules decreases that is the molecules no longer move (either translation or

rotation) with sufficient freedom to allow formation of crystalline lattice.

The driving force for crystallization of solvent in the same vane (Hartel, 2004) is

simply the temperature sub-cooling below the thermodynamic freezing/melting point. In

order to freeze in a sucrose water system, for example, the solution must be cooled below

the freezing point for the given solute concentration. The rate of freezing is generally

proportional to the driving force (sub-cooling), although temperatures too low also result

in decreased crystallization since deceased molecular mobility inhibits freezing.

2.8.0 The Concept of Mathematical Modeling Equations

According to Wikipedia (2012), a mathematical model is a description of a

system using mathematical concepts and language. The process of developing a

mathematical model is termed mathematical modeling.

Often, when engineers analyze a system to be controlled or optimized, they use a

mathematical model. In analysis, engineers build a descriptive model of the system as a

hypothesis of how the system could work, or try to estimate how an unforeseeable event

could affect the system. Similarly, in control of a system, engineers can try out different

control approaches in simulations.

 31

Quarteroni (2009) asserted that mathematical models are used not only in the

natural sciences (such as physics, biology, earth science, meteorology) and engineering

disciplines (e.g. computer science, artificial intelligence), but also in the social science.

Mathematical modeling aims to describe the different aspects of the real world, their

interaction, and their dynamics through mathematics. It constitutes the third pillar of

science and engineering, achieving the fulfillment of the two more traditional disciplines,

which are theoretical analysis and experimentation A mathematical model usually

describes a system by a set of variables and a set of equations that establish relationships

between the variables. Variables may be of many types; real or integer numbers, Boolean

values or strings etc. The variables represent some properties of the system, for example,

measured system outputs often in the form of signals, timing data, counters, and event

occurrence (yes/no). The actual model is the set of functions that describe the relations

between the different variables.

The mathematics of balance and conservation laws are straightforward at this

level of abstraction. Denoting the physical property being monitored as Q(t ) and the

independent variable time as t , we can write a balance law for the temporal or time rate

of change of that property within the system boundary depicted thus:

dQ(t ) = qin(t ) + g (t ) − qout (t ) − c(t ), (2.0)

where qin(t ) and qout (t ) represent the flow rates of Q(t ) into (the influx) and out of (the

efflux) the system boundary, g (t ) is the rate at which

Q is generated within the boundary, and c(t ) is the rate at which Q is consumed within

that boundary.
 32

2.9.0 Stages of Modeling

Depending on the event modeling processes can take various shapes the steps are

essentially the same: Curriculum K-12 Directorate (undated) proposed a modeling

process to posses the pattern illustrated in Figure 2.2 below. Although there are many

variations of the process, most of the modeling processes are essentially saying the same

thing. However, the various classroom approaches can be very different.

Stage 1
Real World Problem

Stage 6
Verify the Model Stage 2
Make Assumptions
Stage 7
Report, explain, predict

Stage 5 Stage 3
Interpret the Solution Formulate
Mathematical Problem
Stage 4
Solve Mathematical
problem

Figure 2.2: Modeling Stages

Source: Curriculum K-12 Directorate (undated)

The steps to modeling some system by Difference or Differential Equations according to

Alder(2001) have been highlighted to include the following:

i. First decide if you are interested in the changing state of some system in time. If not,

it may be that you don’t need ODEs or difference equations. This doesn’t mean that
 33

you don’t have the option on modeling the system, just that you may need others

sorts of mathematics.

ii. Try to picture the system clearly and work out what properties of the system you wish

to measure in order to describe it. This may involve. The simplest form of the Van

de Pol equation is:

y - x3
x = (2.1)
3+ x

y˙ = −x (2.2)

iii. Decide whether you prefer a discrete time system or a continuous time system.

Workout which are the measured quantities, the state variables which Change in time.

Write down any rules you can formulate which link the rate of change of the

variables, or maybe higher derivatives, to other derivatives or to the state variables

themselves. A qualitative description of the direction of change of variables should

come first (wolves eat sheep and produce more wolves, sheep breed but get eaten by

wolves). Then you need to write down some equations making these ideas precise:

S’(t) = αS(t) − βS(t)W(t) (2.3)

For example, this translates into algebra the statement: ‘Sheep breed but get eaten by

wolves’. Note that in this phase, you may have to make some simplifying choices. If you

feel that the simplifications go too far, make a note of the need to improve the model, but

keep it simple on your first pass. Make a list of the factors that increase the rate of change

of each variable, and those that decrease it.

v. Check to see if all you know about the system has been translated into algebra. Pay

careful attention to anything you know about the possible range of the variables,

e.g. always positive.

 34

vi. Do not bother about the problem of solving the equations until you are pretty sure

they are the right ones to solve. Go back and forward a few times to see if

everything that can be said in English is translated correctly into algebra.

vii. Make any estimates you can of plausible parameter values.

viii. For two variable systems, examine the phase plane and use a program to investigate

the time development from various starting places.

ix. When you have explored the initial value dependencies, check to see if the actual

system is going to be in a state which is sensitive to initial conditions, i.e. when

nearby points go to different places.

x. Check to see if you have sensitivity to parameters. Does wobbling the parameters

produce reasonably similar outcomes? Do this numerically using computer graphics

when possible.

xi. Check on the fixed points by algebra. Would you expect to have fixed points here?

Are the fixed points stable or unstable? Can you sketch the solution using a little

algebra, so you can check on whether the numerical program is behaving properly?

xii. Check your model against reality wherever possible. If it isn’t possible at all, you do

not need to waste time on it.

xiii. Are discrepancies in the model likely to arise from the simplifications you

introduced? Try developing the model to see If it can be made a better model by

bringing in more reality, a bit at a time.

xiv. If the equations are really simple, as when you have put in some simplifications,

look them up in a good book to see If they have a closed form solution. There are

many excellent books giving solvable equations, although a preferred alternative

 35

these days is to use a computer package such as MATLAB or Mathematica. Do not

however put your faith in a computer package (Billings, 2013).

2.10.0 Classification of models

Application of mathematics in social sciences outside of economics has been

criticized for unfounded models. Application of catastrophe theory in science has been

characterized as a floating model (Wikipedia, 2012)

Models may be classified into several classification schemes based on different criteria to

include the following:

§ Based on time reference

§ Based on degree of certainty

§ Based on structure

§ Based on function or purpose

§ Based on method of solution or quantification

These models bother on issues whose examples include:

· Many everyday activities carried out without a thought are uses of mathematical

models. A geographical map projection of a region of the earth onto a small, plane

surface is a model which can be used for many purposes such as planning travel.

· Another simple activity is predicting the position of a vehicle from its initial

position, direction and speed of travel, using the equation that distance traveled is

the product of time and speed. This is known as dead reckoning when used more

formally. Mathematical modeling in this way does not necessarily require formal

mathematics; animals have been shown to use dead reckoning.

 36

2.10.1 Classification Based on time Reference

Gupta and Hira (2008) outlined two models as follows:-

i) Static models – these represents a system at some specific time and does not

account for changes over time. An illustrative example is the inventory model

which can be developed and solved to determined an economic order quantity for

the next period assuming that the demand period will remain the same as that of

today.

ii) Ii) Dynamic Models – in dynamic models, time is considered as one of the

variables and allows the impact of changes due to change in time. Thus,

sequences of interrelated decision over a period of time are made to select the

optimal course of action to optimize the given objective. Dynamic programming

is an illustrative example of a dynamic model.

2.10.2 Classification Based on the Degree of Certainty

Two models according to (Sharma, 2008; Wikipedia 2012) have been discussed to

include:

Deterministic Models – in a situation where all the parameters, constants and functional

relationships are assumed to be known with certainty when a decision is made, the model

is said to be deterministic. In such a case the outcome associated with a particular course

of action is known. Hence for a specific set of input values, there is a uniquely

determined output which represents the solution of the model under conditions of

certainty. The results of the model assuming a single value. An illustrative example the

linear programming model.

 37

Probabilistic (stochastic) Model – Is a model in which at least one parameter or

decision variable is random, therefore an independent variable which is the function of

dependent variable(s) will also be random. This means that consequences or off due to

certain changes in the independent variable cannot be predicted with certainty. However,

it is possible to predict a pattern of values of both variables by their probability

distribution function. Insurance against risk of fire accident and sickness etc are examples

where the pattern of events is studied in the form of probability distribution.

2.10.3 Classification Based on Structure: this include three Models as follows-Iconic

or physical Models: in this case, properties of the real system are represented by the

properties themselves. They are often illustrated by a linear scale. Thus Iconic models

resemble the system they represent but differ in size. They may be images, for example,

globes are used to represent the orientation and shapes of various continents, oceans and

other geographical features of the earth. Iconic models may be two dimensional

(photographic, maps, blue prints, paintings, sketches of insects, etc) or three dimensional

(globes, automobile, airplanes, etc) (Sharma, 2008).

Analogue or schematic models: these models can be used to represent diagrammatic

situations and are more often useful than iconic models. They use one set of properties to

represent some other set of properties which the system under the study possesses. After

the model is solved the solution is interpreted in terms of the original system. For

example graphs are very simple analogies. They represent properties like force, speed,

age, time etc in terms of distance. A graph is well suited for representing qualitative

relationships between any two properties and predict how a change in one property
 38

affects the other. Other examples are organizational charts and flow process chart (Gupta

and Hira, 2008).

2.10.4 Classification Based on function or purpose

Models based on the purpose of their utility include the following:

Descriptive models: they characterize things as they are. The major use of this model is

to investigate the outcomes or consequences of various alternatives courses of action.

These models check the consequences only for given conditions rather than for all

conditions. There is no guarantee that an alternative selected with the aid of descriptive

analysis is optimal. They are also used when the objective is to define the problem or to

assess its seriousness rather than to select the best alternative. Simulation is an example

of descriptive technique for conducting experiments with the systems (Gupta and

Hira,2008).

Mathematical or symbolic models: employ a set of mathematical symbols (letters,

numbers etc) to represent decision the variables of the system under study. These

variables are related together by mathematical equations, inequations which describe the

properties of the system. A solution from the model is then obtained by applying a well

developed mathematical techniques. The relationship between velocity, acceleration and

distance is an example of mathematical model. In many research projects all the three

types of models are used in sequence. Iconic and analogue models are used as initial

approximations which are then refined into symbolic/mathematical model(Gupta and

Hira, 2008).

Predictive models: models of this type relate dependent and independent variables and

make provision for virtual experimentation with the computer. They are used to predict
 39

the outcomes due to a given set of alternatives for the problem. These models do not have

an objective function as part of the model to evaluate decision alternative.

Narrative or optimization Models: these models provide the best optimized solution to

problems subject to certain limitations on the use of resources. These models provide

recommended courses of action. These models are also called prescriptive models

because they prescribe what the decision-maker ought to do(Sharma,2008).

2.11.0 Classification of mathematical models

Wikipedia(2012) in their view asserts that many mathematical models can be classified in

the following ways: linear versus. nonlinear, deterministic versus. probabilistic

(stochastic), static versus. dynamic, discrete versus. continuous and deductive, inductive

or floating.

A deductive model is a logical structure based on a theory. There are many

examples of interdependence in the medicines, economic structures, and arises from

empirical findings and generalization from them. The floating model rests on neither

theory nor observation, but is merely the invocation of expected structure.

2.12.0 Sugar Yields

Yields (Chen and Chou, 1993) is an expression which is generally taken to mean

commercial sugar produced percent cane, without regard to the composition of the sugar.

It is the reciprocal of tons of cane per ton of commercial sugar. When yield is expressed

as ton sugar per ton cane the downward trend may indicate the degradation of cane; but to

be more rightly interpret the figure may be misleading because of the increasing trash in

cane.
 40

The yield of jaggery (brown sugar) from sugar cane depends mostly on the quality

of the cane and the efficiency of the extraction of juice. The table 2.6 below gives some

extreme values.

Table 2.6: Yield of Brown Sugar

High quality cane Poor quality cane

Juice per 100kg of 50kg 40kg
cane
% sugar in juice 22 17
jaggery per 100kg of 10kg 7kg
cane
Source:. Earle (2002)

High quality cane has a good juice content with high sugar levels (20%+). Poor quality

cane or cane that has been harvested early may have similar juice content but the sugar

levels will be reduced.

The efficiency with which juice can be extracted from the cane is limited by the

technology used. The simple three roller crushers used by most artisanal producers will

never extract more than 50kg of juice from each 100kg of cane.

Yields are also improved by careful control of the boiling process. Boiling should be

completed as rapidly as possible and the conditions kept as clean as possible.

2.13.0 Some applications of mathematical models

Since pre-historical times, simple models such as maps have been used. Often,

when engineers analyze a system to be controlled or optimized, they use a mathematical

model. In analysis, engineers can build a descriptive model of the system as a hypothesis

of how the system could work or try to estimate how an unforeseeable event could affect
 41

the system. Similarly, in control of a system, engineers can try out different control

approaches in simulations.

There are six basic groups of variable namely: decision variables, input variables,

state variables, exogenous variables, random variables, and output variables. Since there

can be many variables of each type, the variables are generally represented by vectors.

Decision variables are sometimes known as independent variables. Exogenous variables

are sometimes known as parameters or constants. The variables are not independent of

each other as the state variables are dependent on the decision, input, random, and

exogenous variables. Furthermore, the output variables are dependent on the state of the

system (represented by the state variables).

2.14.0 Mathematical modeling of motion systems

In linear motion(Billings, 2013), the distance law applied to a body with linear motion is
a momentum balance, often denoted as force balance:

d [m(t ) v (t )] = å Fi (t )
d
[I (x )] = i
(2.4)
dt ft

Where

I [Ns] = momentum (man x speed)

Fi = Force

I = Mv = Mx (2.5)

Where

M = man, kg

V = speed, M/S

X [M] = position
 42

If M is constant M can be moved outside the derivative term in (2.4)

Mv(t ) = m x (t ) = a (t ) = å Fi (t ) (2.6)

V = X = a is acceleration (Newton 2nd law)

S0 equation 2.6 can be used for modeling systems with rotational motion and can be

modeled in the same way as systems with linear motion(Billings, 2013).

The sugar boiling process must handle different feed stocks (syrup of different

purity) and must produce several type of sugars. So, in order to avoid the development of

separate black-box linear models for each operating point, which can be somewhat time

consuming, a nonlinear model, applicable to all conditions, is needed. As the main

nonlinearities are included in the sucrose growth rate, it will be the key parameter of the

proposed model. The crystallization phenomenon is complex, so the development of a

pure white (or totally first principles) nonlinear growth rate model is prohibited (Lauret et

tal 2000).

2.15.0 Programming for Systems Analysis

According to Wikipedia free encyclopedia (2012) computer programming

(shortened to programming or coding) is the process of designing, writing, testing,

debugging, and maintaining the source code of computer programs. In his work on the

review of some common project output data types, Enibe (2013) stated that several

computer software have been developed to facilitate efficient data analysis and

visualization which of course includes systems analysis as well. The range of these

packages he added include very high level structured programming languages as

FORTRAN, C or C++ as well as BASIC, MATLAB/SCILAB, FREEMATH, PYTHON,

etc. Computer programming, Liang (2004) is the craft of writing useful, maintainable,
 43

and extensible source code which can be interpreted or compiled by a computing system

to perform a meaningful task. Computer programming can be performed in one of the

numerous languages, ranging from a higher-level language to writing directly in low-

level machine code (that is, code that more directly controls the specifics of the

computer's hardware) all the way down to writing microcode (which does directly control

the electronics in the computer). Software development, design, and testing have become

very intricate with the advent of modern highly distributed systems, networks,

middleware, and interdependent applications. The demand for complex software systems

has increased more rapidly than the ability to design, implement, test, and maintain them,

and the reliability of software systems has become a major concern for our modern

society(Dalal, 2003).

Programming is not just syntax, classes, or objects. It is really problem solving. Loops,

methods, and arrays are fundamental techniques for problem solving. From fundamental

programming techniques to object oriented programming, there are many layers of

abstraction.

2.16.0 MATLABB

MATLAB as defined by Ale (2011) is a computer algebra system created in 1964

by Carl Engelmann at MITRE. It is a machine aid for mechanical symbolic process

encountered in analysis. It is capable of performing automatic and symbolic computation

in Mathematics, Science and Engineering. It integrates mathematical capability,

visualization, and a powerful language to provide a flexible environment for technical

computing; Patrick and Thomas (2003) said about MATLAB “MATLAB is not just a

computation and plotting package.

 44

It is a versatile and flexible tool which allows users with even the most

elementary programming capabilities to produce sophisticated graphic and graphic user

interface (GUI). The level of sophistication is only limited by ones needs, curiosity and

imaginations” One of the many things that is interesting about MATLAB (and that

distinguishes it from many other computer programming systems, such as C++ and Java)

is that you can use it interactively. This means you type some commands at the special

MATLAB prompt and get results immediately (Hahn and Valentine,2010).

2.17.0 Application of MATLAB in Systems Modeling

According to Enibe (2013) the goal of MATLAB/SCILAB is to enable the users

solve a wide spectrum of analytical and numerical problems using matrix- based

methods, attain excellent interfacing and interacting capabilities, compile with high-level

programming languages, ensure robustness in data-intensive analysis and heterogeneous

simulations, provide easy access to and straightforward implementation of state-of-the-art

numerical algorithm, guarantee powerful graphical features, etc. MATLAB has

interacting computing environment for providing both basic and sophisticated functions.

The inbuilt functions can be used to solve possibly complex but standard problems, or

one can develop one’s program by writting them as M-file, that is, written in a matrix

oriented language.

2.18.0 Review of Existing Models Related to Sugar Manufacture

2.18.1 Thaval and Kent (2012a) worked on the “modeling the flow of juice through a

mill” they came up with a juice flow model developed to estimate the juice expression

at the four nips of a six roller mill. An extended volumetric theory was applied to
 45

determine the juice expressed at each nip. The model was applied to a first and final mill,

using typical mill settings and an empirical equation to estimate reabsorption.

Results of using the model for typical heavy-duty pressure feeder settings showed

that most of the juice is expressed at the pressure feeder nip. Since the pressure feeders

are remote from the mill, a significant portion of the juice is expressed before the bagasse

enters the mill.

Juice Flow Model

In order to calculate the amount of juice expressed at a nip, the volumetric theory

of extraction was used (Russell, 1968). The theory assumes that fibre rate is constant

through the milling train and no fibre is expressed with juice.

The volume rate of juice entering the mill first needs to be determined. The mass fibre

rate through the milling train can be calculated from:

ṀcF = Ṁc x PcF (2.7)

100
Where

Ṁcf is rate of fibre (kg/s)

Ṁc is crushing rate (kg/s)

The mass fibre rate can be converted by

Vcf = Ṁcf (2.8)

df

where

Vcf is volume of fibre (m3/s)

df is density of fibre (kg/m3)

the mass rate of juice in the feed entering the nth mill is determined from
 46

ṀFnv = ṀB(n-1) + Ṁinv (2.9)

Where

ṀFnv is mass rate of feed entering the nth mill after imbibitions (kg/s)

Ṁinv is mass rate of juice in imbibitions to the nth mill (kg/s)

ṀB(n-1) is mass rate of juice in bagasse from the (n-1)th mill or in cane for the first mill

(kg/s)

In conclusion, we find out that the model was focused on the estimation of the

juice expression at each nip of a six-roller mill. Results of using the model for typical

heavy-duty pressure feeder settings show that most of the juice is expressed in the

pressure feeder.

2.18.2 Lauret et tal(2000)- Hybrid Model

The mathematical model of the sugar boiling process highlights the following state

vector:

X=[W,S, I,C,T,µ1,µ2 ] (2.10)

W is the mass of water in the pan

S is the mass of sucrose in solution

I is the mass of impurities in solution

C is the mass of sucrose in crystal form

T is the temperature of massecuite

µ1 is first moment of the distribution function

µ2 is second moment of the distribution function

The following assumptions guides the design of the model:

 47

· The pan is well-mixed thanks to a good circulation. Consequently, it is

homogeneous in temperature and super saturation

· The totality of the massecuite is in ebullition. The hydrostatic pressure effect

is ignored. The temperature of massecuite is the ebullition temperature of

water, derived from the vacuum in the pan, plus the boiling point elevation.

· Seeding takes place and the number of crystals is assumed to be constant all

along the strike. This assumption leads to the following ones

· No primary nucleation, agglomeration nor breakage.

· The Mac Cabe’s law is verified e.g. the growth rate is independent of crystal

size.

· Growth rate dispersion does not occur

The work on “Hybrid Modeling of Sugar Boiling Process (Lauret et al, 2000) addresses

the development of crystal growth rate model according to two approaches. The first is

classical and consists of the parameters o the empirical expressions of the growth rate

through the use of nonlinear programming optimization technique. The second is a novel

model strategy that combines an artificial neural network (ANN) as an approximator of

the growth rate with prior knowledge presented by the mass balance of sucrose crystals.

The first results show that the first type of model performs local fittings while the second

offers greater flexibility. The two models were developed with industrial data collected

from 60m3 batch evaporative crystallizer.

Conclusively, this Hybrid ANN model could only but infer the weight of crystals

from the level of brix of massecuite measurements and crystal content.

 48

2.18.3 Sotudedeh- Gharebagh et tal (2009) Model

Sotudedeh-Gharebagh et al (2009) carried out the “modeling and optimization of

the sugar extraction process” through the following model;

V(C1,i/z-G,i/z+dz) + Kca(Ci – Cic,i)Adzϵ - K,C1,iAdzϵ=0 (2.11)

Bulk flow interphase mass transfer invert sugar

i= Ns, S and W

where;

Ns and S are non-sugar and sugar compounds respectively in the raw juice as well as

solid phase.

Dividing both equations by Adzϵ and simplifying yields

a dC1,I - β(Ci – C1,i) + K1C1,I = 0 (2.12)

dz
where;

a=V (2.13)
Aϵ

b = K, a

The initial and boundary conditions for equations (2.13) are

atz = C1,i (2.14)

The first term in the equation (2.12) is the mass transfer of the components due to

the bulk motion of the juice flow.

The second term of this equation represents the inter-phase transport of components from

solid to liquid phase.

This work is basically done on sugar beet. The process took place in a vertical tower in

which hot water and beet cossettes were entered counter currently and the juice enriched
 49

with sugar. The amount of sugar extracted and also the amount of sugar loss at the exit

was increased by an increased in tower length.

The study therefore determined the optimal values in a way to minimize the total

operating cost based on the mathematical model using a genetic algorithm method. The

results of the study can be used as a framework to optimize the performance of industrial

extractors.

2.18.4 Saurez et tal (2011) model

Suarez et al (2011) model conducted a study on model control strategies for batch

sugar crystallization process using the following model:

M = f1[M(t), F(t), S1(t), t0≤t ≤t,f M(0) = M0 (2.15)

Where

M(t)ϵRq and F(t) ϵRm

Are the mass and flow rate vectors with q and m dimensions respectively and tf is the

final batch time.

His view was that sugar crystallization occurs through the mechanism of nucleation,

growth and agglomeration. Hence, the general phenomelogical model of the fed batch

crystallization consists of mass, energy and population balances, including relevant

kinetic rates for nucleation, linear growth and agglomeration. He asserted that the mass

and energy balances are common expressions in many chemical process models, but the

population balance related to crystallization is still an open modeling problem. So the

work was centered on achieving nucleation, growth and agglomeration.

 50

2.18.5 El-Belgiti and Vorobiev(2005) Model

El-Balgiti, Rabhi and Vorobiev (2005) had a study on “Kinetic model of sugar

diffusion” from sugar beet tissue treated by pulse electric field (PEF)

The kinetics stated thus:

Yt - Y¥ ¥
y=
Y0 - Y¥
= å Cndl
n =1
nt (2.16)

Where;

ƛn= Dq2 /l2 2l is the solid thickness (2.17)

y = the concentration ratio

0, t, and ∞ are subscripts representing the relevant concentration at t=0, t and ∞

(equilibrium) respectively.

D = diffusion coefficient

2a (1+ a )
Cn =
(
1+ a + 2 qn2 ) (2.18)

qn are non-zero positive roots of the equation

tan (qn) = aqn and a = n æç y ö÷a (2.19)

è xø

where;

Y and X are solute concentrations in the extracts and solid pressed out juice concentration

at the end of extraction respectively.

The liquid/solid ratio is

n = LS (2.20)
 51

where;

L is the extract volume and

S is the solid volume.

The Kinetic Diffusion Model studied the application of a pulsed electric field

(PEF) for pressure or solvent extraction off cellular juice from sliced sugar beets. When

exposed to an electric field of sufficiently high intensity, the cell membrane was found to

be temporarily destabilized or irreversibly ruptured (a process of electropolation). This

method was found to be promising because the PEF ensures non-thermal

permeabilisation of cellular membranes and prevents cell walls from undergoing thermal

alteration. The extraction followed various pulsed electric fields treatments (from 300 to

800v cm-1) and number of pulses varying between 50 and 1000). The slices treated were

immersed in water at ambient temperature of a liquid/solid ratio of 3. A significant

increase in extraction yield was observed. This enhancement was due to

permeassibilisation of the cellular membrane and to the addition quantity of juice

appearing on the surface of slices after PEF treatment being extracted rapidly by

convection. The optimal conditions of PEF treatments were an intensity of 670v cm-1 and

250 pulses.

2.18.6 Thaval and Kent (2012b) Mass Balance Model

Thaval and Kent (2012) conducted a study on the “Modeling the flow of cane

constituents in the milling process”. The model tracks the soluble and insoluble solid

cane components using modeling theory and experimental data, assisting in further

understanding the flow of constituents into mixed juice and final bagasse. According to

them, the work provided an opportunity to understand the factors which affects the
 52

distribution of cane constituents in juice and bagasse. Application of the model should be

improvement in the overall performance in the milling train.

The focus of this model therefore was to break down the soluble solids (brix) into

pol and soluble impurities and insoluble solids (fibre) into true fibre and mud solids

impurities and determine the flow of these components through the milling process.

Ultimately, this model is intended for use in a “whole of factory” model and will provide

the correct constituents in mixed juice and final bagasse as inputs to the downstream

station model.

The model consisted of the following milling train mass flows;

ṁc = ṁJ1 +ṁB1 (2.21)

ṁB1 +ṁJs +ṁJB = ṁB2 + ṁJ2 (2.22)

ṁB2 + ṁJ4 = ṁB3 + ṁJ3 (2.23)

ṁB2 + ṁJ5 = ṁB4 +ṁJ4 (2.24)

ṁB4 + ṁI = ṁB5 + ṁJ5 (2.25)

ṁc + ṁI = ṁJm + ṁB5 (2.26)

Where;

C = cane

I = Imbibition

B = Bagasse

J = Expressed juice stream

Js = Return stream from juice screen

Jm = Mixed juice
 53

2.19.0 Mass and Energy Balance

The law of conservation of mass leads to what is called a mass or a material balance.

Mass In = Mass Out + Mass Stored (2.27)

Raw Materials = Products + Wastes + Stored Materials. (2.28)

Σm = Σm + Σ m + Σm (2.29)
R P W S

(where Σ (sigma) denotes the sum of all terms).

Σm = Σm + Σ m + Σm = Total Raw Materials (2.30)

R R1 R2 R3

Σm = Σm + Σ m + Σm = Total Products. (2.31)

P P1 P2 P3

Σm = Σm +Σm + Σm = Total Waste Products (2.32)

W W1 W2 W3

Σm = Σm + Σ m + Σm = Total Stored Products. (2.33)

S S1 S2 S3

If there are no chemical changes occurring in the plant, the law of conservation of mass

will apply also to each component, so that for component A:

m in entering materials = m in the exit materials + m stored in plant.

A A A

For example, in a plant that is producing sugar, if the total quantity of sugar going into

the plant is not equaled by the total of the purified sugar and the sugar in the waste

liquors, then there is something wrong. Sugar is either being burned (chemically

changed) or accumulating in the plant or else it is going unnoticed down the drain

somewhere. In this case:

= (m + m +m )
A AP AW AU

where m is the unknown loss and needs to be identified( Wiley and Turner,2006).
AU

Now:

Raw Materials = Products + Waste Products + Stored Products + Losses

 54

where Losses are the unidentified materials.

Just as mass is conserved, so is energy conserved in food-processing operations.

The energy coming into a unit operation can be balanced with the energy coming out and

the energy stored.

Energy In = Energy Out + Energy Stored

ΣE = ΣE + ΣE + ΣE + ΣE
R P W L S

where

ΣE = E + E + E + ……. = Total Energy Entering

R R1 R2 R3

ΣE = E + E + E + ……. = Total Energy Leaving with Products

p P1 P2 P3

ΣE = E +E +E + … = Total Energy Leaving with Waste Materials

W W1 W2 W3

ΣE = E + E + E + ……. = Total Energy Lost to Surroundings

L L1 L2 L3

ΣE = E + E + E + ……. = Total Energy Stored (Earle,2002)

S S1 S2 S3

2.20 Prevailing Extraction Theory

According to Thaval (2012b), the fundamental equation for the mass balance of the entire

milling train is given by equation 2.34.

(2.34)

m& c is mass rate of cane (kg/s),

m& i is mass rate of added water (kg/s),

m& JM is mass rate of mixed juice (kg/s),

m& B is mass rate of final bagasse (kg/s).

This equation assumes that the final mill constitutes the cumulative output of the milling

train, number of mills in the milling train notwithstanding.

 55

Extraction is calculated by the percentage of sucrose extracted from cane. The percentage

of the sucrose in the original cane removed with the mixed juice is termed “Sucrose

extraction”, Ep (Rein, 2007).

(2.35)

Where: m& JMP is mass flow of sucrose in mixed juice (kg/s),

m& CP is mass flow of sucrose in cane (kg/s)

The mill engineer according to Thaval (2012), does not have control over the sucrose but

has control over the brix and hence it would be reasonable to report brix extraction.

(2.36)

Where: EB is brix extraction,

MCB is mass flow of brix in cane (kg/s),

MJMB is mass flow of brix in mixed juice (kg/s)

2.21.0 Extraction Performance Parameters

Extraction is accepted to be a volumetric process and extraction performance of a

single milling may be calculated from the filling ratio, reabsorption factor and imbibition

coefficient (Russel, 1968). If these parameters are known, the output quantities from the

milling unit can be calculated from the input bagasse analysis.

2.22 .0 Filling Ratio

Filling ratio is the non-dimensional representation of the delivery setting of the

mill. It is defined as the no void volume (volume without air entrapped in it) of fibre per

unit escribed volume. (Murray, 1959).

 56

( 2.3.7)

Where: is volume rate of cane fibre (m3/s),

is the volume rate escribed by the top and delivery rollers of the mill (m3/s)

The escribed volume is defined as the product of the work opening (W), the top roller

surface speed (S) and the length of roller (L).

(2.38)

Where: L, W and S are length of roller (m), work opening (m) and top roller surface

speed (m/s) respectively.

In practice, filling ratio is a function of mechanical loading and control parameters of the

milling unit.

2.23.0 Reabsorption Factor

Reabsorption factor represents the volumetric juice extraction performance of the mill

(Tharval, 2012a).

(2.39)

Where: is reabsorption factor of the mill,

is no void volume rate of bagasse of the mill (m3/s).

i) based on the assumption of perfect working condition of the mill, the efficiency of

the machine is assumed to be at least 85%, the said efficiency

comprehensively representing the reabsorption factor/filling ratio .

 57

2.24.0 Imbibition Coefficient

Imbibition coefficient is defined as the ratio of the actual brix extraction to the

theoretical brix extraction of the mill (assuming perfect mixing of the imbibition liquid

and residual juice in bagasse. It is the measure of the performance of the mill in

extracting brix.

(2.40)

Where: is imbibition coefficient of the mill,

is brix extraction of the mill,

is theoretical brix extraction of the mill.

From the above definition, it would be expected that the imbibition coefficient is a strong

function of the imbibition level and the brix levels of both the imbibition juice and the

bagasse to which it is applied.

Using the three mill parameters viz. filling ratio, reabsorption factor and

imbibition coefficient defined above, the composition of delivery bagasse and expressed

juice may be calculated from the given feed quantities. It is assumed that no fibre passes

into the expressed juice, so that the weights of fibre in the feed and delivery bagasse are

identical (Thaval, 2012b). Thus,

(2.41)

Where: ṁB(n-1)F is the mass rate of fibre in delivery bagasse of the preceding mill (kg/s).

ṁBnF is the mass rate of fibre in delivery bagasse of the current mill (kg/s).

From equation 2.37 and 2.39, the ratio of reabsorption factor to the filling ratio is given

by,
 58

(2.42)

Where Kn is the reabsorption factor of the mill,

Cn is the filling ratio of the mill,

⩒Bn is the volume of delivery bagasse of the mill,

⩒BnF is the volume of fibre in delivery bagasse of the mill.

Rearranging equation 2.42, we get,

(2.43)

The volume of juice in bagasse is given by,

(2.44)

Where: is the volume of juice in delivery bagasse of the mill.

Substituting equation 2.43 into 2.44 and rearranging the terms we get,

(2.45)

It is on the basis of the above existing models and equations that further analysis

is hereby projected to predict sugar yield from sugar cane, as treated in the next section.

The models above prove to be useful in estimating the output quantities of the milling

train when less information is known. Its application is restricted as the constant ratio

assumptions are not entirely correct and cannot be calibrated to take into account known

results.

2.25.0 Status quo and Main Knowledge gap in the Study of Sugar yield from Sugar cane

The Nigerian sugar industry is by no means young, having been first established

in the 1960s. It can however be regarded to still be in its infancy given the fact that today,
 59

it only supplies about 2% of the nation's requirement, in spite of our comparative and

competitive advantages for sugar production. The performance has deprived the country

of all the benefits derivable from a vibrant sugar sector. Chief among these are the annual

drain on the nation's foreign exchange earnings put at N101.9 billion in 2011, the loss of

hundreds of thousands of employment opportunities for skilled and semi-skilled labour

and food insecurity arising from sugar import dependence. But all these are about to

change now.

In 2008, the Federal Government of Nigeria directed the National Sugar

Development Council (NSDC), to develop a road map for the attainment of self-

sufficiency in sugar within the shortest time possible. In compliance, the Council came

up with the Nigerian Sugar Master Plan an abridged version of which is the main subject

of this slim volume.

The Plan has estimated that our demand for sugar would breach the 1.7 million

metric tonnes (MMT) mark by 2020. To be able to satisfy this from domestic production,

so the government intends to establish some 28 sugar factories of varying capacities and

bring about 250,000 hectares of land into sugarcane cultivation, over the next 10 years.

The bulk of the investment capital will come from private investors (NSDC, 2015).

According to GAIN report (2013) Nigeria’s domestic sugar production in

May2012/2013 remains unchanged at 65,000 tons (raw value), the same figure as May

2011/2012. Dangote-owned Savannah Sugar has completed the first phase of its

rehabilitation program, with about 6,700 hectares of newly planted sugar cane fields.

Dangote Sugar Company is also partnering with Nigeria’s Taraba State government to

establish a sugar processing company in Lau Local Government Area of that state. Josepdam

Sugar Company has embarked on an aggressive nursery establishment to produce enough

 60

sugar cane seeds for field expansion. Currently it has 1,250 hectares of seed cane but the

available raw material is not adequate to start operations. Other sugar estates are in varying

stages of rehabilitation. For example, Golden Sugar Company (Sunti), owned by Flour Mills

of Nigeria (FMN), has developed 2,000 hectares of land and planted over 1,000 hectares of

sugarcane (using a center-pivot irrigation system) but has not reached a minimal level

necessary for successful milling. Savannah Sugar reported a decrease in sugar cane yield

from 66 tons per hectare in 2010/2011 to 60 tons in 2011/2012. The average yield of refined

sugar from a ton of cane is estimated at approximately 0.961 or 9 percent (GAIN report

(2013).

2.25.0 Main Knowledge Gap in the Study of Yields of Sugar from Sugar cane

The main knowledge gap in the study of yields of sugar from sugar cane is that the

production process has not advanced to the level of using higher techniques such as the use of

special models for the prediction of sugar yields and its various by-products. This lasp is

occurring in spite of the fact that so much advancements have been employed in the

production of sugar or the purpose of increasing sugar output to meet the increasing demand

of the product.
 61

CHAPTER THREE

3.0 MATERIALS AND METHODS

3.1 General

This research aimed at the prediction of Sugar Yield from Sugar Cane using

process modeling. Sugar value is often not known or estimated until production is

completed in the factory at every given occasion. This method lacks the potential to

quantify the yield of sugar from sugarcane. sugar and major it’s by-products including

bagasse, molasses, and filter cake were determined in the research “Prediction of Sugar

Yield from Sugar Cane using process modeling” . Savannah Sugar Company, Numan,

Adamawa State was used as the research site.

3.2 The Savannah Sugar Company, Numan

Savannah Sugar Company Limited, Numan located in Adamawa State of North-

Eastern Nigeria was established in 1971 by then Federal Government of Nigeria. The

North eastern state government was accordingly saddled with the responsibility of land

acquisition, compensation payments and settlements of the affected communities. This

responsibility devolved the Gongola State government on creation of States in 1976. This

means that Savannah Sugar Company Limited was neither involded in land acquisition or

compensation. The Company is operating an integrated sugar farming and milling. It has

a mill capacity of 50,000 Mtpa and has the largest refinery in sub-Saharan Africa. The

transfer of its ownership to Dangote Sugar Company took place in 2003 and since then

there has been a joint ownership of the Sugar Company with Dangote possessing at least

75% of the partnership. Presently, the Company is cultivating a total landed area of

18,000 hectares and it is employing up to 20,000people made up of direct employees and

61
 62

farmer out growers. It is projected to produce 1million tons by 2015. The block diagram

of sugar processing of the Savannah Sugar Company, Numan is shown in figure 3.0

below.
 63

Juice Juice
Harvested clarificatio
cane Cane mills Purification Clear Syrup
Extracted chemical n, physical Evaporati
juice juice on
treatment separation (Thick juice)

(hydrated lime)

Mud (Filter cake)

1st commercial
Sugar Golden (WAX INTERNAL USES
Brown Sugar EXTRACT) - As fertilizer
commercial uses
- Shoe polish
- Candle

2nd
Commerci “Skiming Brown 1st
al Sugar
Fine off”
Brown
Brown Massecuites
Brown Sugar Commerc
Body Liquor Liquor melting sugar Centrifuging ial Sugar
(Chemical boiling
treatment)

water Molasse
s
Run Final
off
Distilation
Insecticide Yeast Animal
production production Feed
Local drinks
Perfume production

White Refined
White Centrifuging Market
Sugar
Massecuites

Key Note
Production products
By products
Return

Figure 3.0: Block Diagram of Sugar Manufacture process in Savannah Sugar

Company, Numan

3.2 Description of sugar production plant

Generally the organization is categorized into:

i) the milling department comprising of cane crushing and juice extraction unit; and

ii) processing department.

 64

3.2.1. Milling Department

This department is under the supervision of a Chief engineer and factory shift

assistants. The main objective of this department is to extract the maximum of juice from

the cane crushed, keeping losses of sucrose in bagasse to minimum. The staff of the

milling department is also responsible for the boilers, steam production, electricity

generation and the general maintenance repairs of all mechanical equipment such as

motors, mills workshop etc.

3.2.2 Processing Department

This department is under the control of a Process Manager and shift assistants.

The main objective is to extract and crystallize out the maximum amount of sucrose from

mixed juice received from the milling from the milling section. Main operations are :-

liming, juice heating, clarification and subsidation, mud filtration, evaporation, boiling in

vacuum pans, cooling in crystallisers and centrifuging of massecuites, drying of sugar.

3.2.3 Laboratory

The chemical and technical control of the factory – milling and processing – is

done by the laboratory under the supervision of a chief chemists assisted by shift

chemists and samplers working on a 24 hour basis. Sampling must be done at all the time

the factory is working so the laboratory work is organised accordingly.

Some products, such as bagasse, filter cake, massecuite, molasses, condensate water must

be sampled at fixed frequency when need arises.

3.3 Determination of Sugar Yield

The formula to determine sugar is complex and so does not depend on a single

equation however there are three measures of cane quality that are important, which will
 65

be briefly mentioned here. Brix is the percentage of dissolved solids on a weight per

weight basis and is measured by refractometer or density meter. Pol is a measure of the

passage of polarised light through the clarified juice. These two measures of juice quality

(corrected for fibre content of the stem) allow determination of the level of impurities in

the cane (ie. Brix minus Pol equals total impurities in the cane). Furthermore this allows

estimation of the sugar yield or commercial cane sugar (CCS) of a grower’s cane

(Mackintosh, 2000).

To calculate CCS it is assumed that three quarters of the impurities remain after the

juice is clarified. These impurities end up in the final molasses, which in turn consists

of ~40% non-recoverable sugar and 60% impurities. Therefore:

CCS = Pol of juice (corrected for fibre content of stem) – ¾ (impurities in cane x
40/60)
= Pol in cane - ½ (impurities in cane)
CCS is a measure of how much pure sucrose can be extracted from the cane. The final

return that the grower receives is determined by additional factors ( Mackintosh 2000).

3.4 Determination of Bagasse Yield

It consists of two types of fibre, which constitute 55% of bagasse dry weight.

These are the cellulose fibre of rind, vascular tissue and the pith of the cane stem.

Bagasse weight is therefore determined by integrating the concepts of Antoine(2000)

which states that every 1000kg 0f Cane there are between 350 – 750kg extractable

bagasse.
 66

3.5 Determination of Filter cake Yield

Filter cake weight in process juice is determined when impurities contained in the

juice are precipitated by treatment with lime and heat and after removal filtration they

form filter muds. It is integrated in the model using the relationship:

Fc = Mm + Ml

Where

Fc is filter cake,

Mm is mud mixture,

Ml is molasses fraction

3.6 Determination of Molasses Yield

Molasses is a residual syrup form which no crystalline sucrose can be obtained

following evaporation, crystallization and fugalling of the massecuite.

This is the residual syrup from which non crystalline sucrose can be obtained

following evaporation, crystallization and fugalling of massecuite. Between 27kg to 40kg

of molasses are produced per ton of cane. Its average composition is 20% water, 35%

sucrose, 20% reducing sugar, 15% sulphated ash and 10% others. Molasses is mainly

used as animal feed or transformed into rum; alcohol or ethanol fermentation and

distillation (Cock and Carlos, 1995; Antoine, 2000). Thus clarified sugar juice is boiled

and centrifuged the first time to produce ‘A’ sugar and ‘A’ molasses. ‘A’ molasses is then

boiled again to produce ‘B’ sugar and ‘B’ molasses. The ‘B’ molasses is boiled a third

time to produce ‘C’ sugar which is mixed with water and is used to seed the next round of

crystallisation. The ‘C’ molasses is referred to as ‘final’ or ‘blackstrap’ molasses

(Mackintosh, 2000).
 67

3.7 Development of Model

There are the various processes or methodologies that are being selected for the

development of the project depending on the project’s aims and goals. Many

development life cycle models have been developed to achieve different required

objectives. The models specify the various stages of the process and the order in which

they are carried out.

The selection of the model has a very high impact on the testing that is to be

carried out. Based on the software application required, the iterative model is selected

because it is the development model that suits the application in sugar processing. The

MATLAB model was relevantly developed for the purpose of predicting sugar yield from

cane sugar. The model as earlier stated in literature was derived from Thaval and Kent

2012a which served as the bases for the development of new set of equations. Details of

the model development procedure and equations are shown in appendix IV.

3.7.1 The MATLAB Simulation Model

The following model analysis is based on mass balance model comprehensively

represented in equation 2.34 above.

Assumptions

The efficiency of the MATLAB model determined to be 75% basically due to the

following assumptions:

• Clarification Temperature, T = 1020C

• Juice pH=7

• And Exhaust pressure, P=1.5kpa

• These global parameters are defined in the var.m .

 68

• All values were measured in metric tons.

3.7.2.1 Source Code Of The Model Developed For The MATLAB Simulation

(a) Source Code of ‘var.m’ MATLAB File

functionxVal=var(x)
%Constants and Variables for Prediction of Sugar

Eff = 0.75; %Milling efficiency of 75%

%GLOBAL PARAMETERS
T=102; % Clarification temp (between 102 and 105 degree
celcius)
pH=7; % (+-1) Juice pH
P=1.5; %(kpa) %%Exhaust pressure

if(strcmp('Eff',x)) %GLOBAL PARAMETERS

xVal(1)=Eff;
elseif(strcmp('Param',x))%GLOBAL PARAMETERS
xVal(1)=T;
xVal(2)=pH;
xVal(3)=P;
end
end

(b) Source Code of ‘Predictfxn.M’ Matlab File

function Pw=predictFxn(C,conv)
%Fetching list of Variables from var.m file
xV=var('Eff'); %Efficiency

Eff=xV(1);
Pw = zeros(size(C));

%Then computing for each component of the sugarcane

extracted in the mill
for i=1:length(C)
ifconv==1
C(i) = C(i) * 1000; %(Conversion from metric ton to kg)
end
Pw(i,1)=Eff * (48.76/100) * C(i); %Mass of Bagasse
extracted (kg)
Pw(i,2)=Eff * (3.94/100) * C(i); %Mass of Filter cake
extracted (kg). Contains dirt composition
Pw(i,3)=Eff * (5.2/100) * C(i); %Mass of Molasses extracted
(kg)
 69

Pw(i,4)=Eff * (13/100) * C(i); %Mass of Sucrose extracted

(kg)
Pw(i,5)=Eff * (24.4/100) * C(i); %Mass of Natural water
extracted (kg)
Pw(i,6)=Pw(i,1)+Pw(i,2)+Pw(i,3)+Pw(i,4)+Pw(i,5);
Pw(i,7)=C(i)-Pw(i,6);
end
end

3.7.2 Validation of the Models

Model validation as defined by Wikipedia(2015) is the substantiation that a

computerized model within its domain of applicability possesses a satisfactory rang of

accuracy consistent with the intended application of the model or Validation is the task of

demonstrating that the model is a reasonable representation of the actual system: that it

reproduces system behaviour with enough fidelity to satisfy analysis objectives.. A model

should be built for a specific purpose or set of objectives and its validity determined for

the purpose.

The model in this study was based on a sufficient amount of a data of ninety (90)

days each of Field and Theoretical simulation of four factors including sugar, bagasse,

molasses and filter cake. The data used here was obtained from the Savannah Sugar

Company, Numan. It was subjected to a statistical analysis of variance(ANOVA) and

comparing the means using least significant difference(F_LSD) to test the validity of the

MATLAB model developed.

 70

3.8 Milling Process

3.8.1 The Milling Train:

The milling process essentially involves the removal of juice from sugarcane by

squeezing the cane between pairs of large cylindrical rolls in a series of milling units

collectively called a milling train. After passing through a pair of rolls and expressing

juice, the remaining sugarcane material is known as bagasse.

To aid in the extraction of juice from the much drier bagasse (prepared cane),

water or diluted juice is added to the bagasse before it enters the milling unit in a process

called imbibition. Such water is called imbibition water. It is usually added to the bagasse

at a rate that is typically 200% to 300% of the rate of fibre passing through the milling

train.

After passing through a juice screen to remove most of the fibre in the juice, the juice

from the milling units called mixed juice, is taken away for processing into sugar.

3.9 Analysis of sugar prediction model, using MATLAB software

The modality of this research has more need for simulation software than a

proprietary software package (standalone software developed in Java, for instance).

Hence, the choice of MATLAB.

For the purpose of this simulation, the following functions and m-files are created:

SN FUNCTION M-FILE NAME DESCRIPTION

NAME
1 xVal var.m Contains constants and variables
2 Pw predictFxn.m Contains array of the equation model for
the prediction of sucrose content of
sugarcane and other by-products.
Description of the m-files
 71

The m-files are created in the MATLAB project directory.

1. Function xVal is a function that keeps the global variables and constants, and

supplies them to any caller once the right parameters are inserted.

2. For instance, a call to var with the parameter ‘Eff’ returns set efficiency of the

milling train. Thus the efficiency of the system can be varied for varying output,

without changing the model.

3. Other parameters include Temperature, pH of the juice and Pressure of the

exhaust.

4. The second m- file (predictFxn.m) contains the MATLAB model of the already

established equations for the prediction of Bagasse, Filter Cake, Molasses and

Sucrose (line 14 to 17). Though Mass of Natural water extracted is not required, it

is included in the analysis to ensure proper account of the mass balance equations

(line18).

5. Line 19 is the total mass of components whose values have been extracted into the

array Pw (Product masses).

6. Line 20 represents the difference between the supplied Cane and the sum total of

the products. This is a mass measure of the efficiency of the system.

7. To analyze a single sugarcane input mass

a. The following command is entered in the command window of MATLAB

i. PRED=predictFxn(x,y)

b. Where PRED is the variable that holds the result of the analysis

c. x is the mass of crushed cane (in kg or in metric tons) to be analyzed

 72

d. y is the condition for conversion from metric ton to kg (1 for conversion, 0

for non-conversion)

8. To analyze an array of sugarcane input mass

a. The sample data is read from excel file into the MATLAB command

window thus:

i. CANE_WEIGHT =xlsread('data.xlsx', 1, 'B2:B91'); %CANE

WEIGHT

b. Then the Prediction analysis is executed for all data sequentially using the

following command:

i. PRED_1=predictFxn(CANE_WEIGHT,y)

9. In either case, Columns 1 through 7 represent

i. Mass of Bagasse in cane

ii. Mass of Filter cane and other insoluble solids in cane
iii. Mass of Molasses in cane
iv. Mass of Sugar in cane
v. Mass of Natural water in cane
vi. Total mass of extracted components
vii. Mass difference between the crushed sugarcane and total mass of

extracted components in that order.

10. In this thesis, a sample data for three months were analyzed using the MATLAB

model thus developed.

11. The next section contains the result of a single prediction process generated from

the analysis.

12. The graphs generated compare the predicted values with the actual values got

from a milling industry.

 73

The raw data obtained from milling factory is stored in data.xlsx worksheet. It contains

the Crushed Cane Weight, Bagasse, Filter Cake, Molasses and Sugar extracts. All values

were measured in metric tons.

The following steps were taken to process the data for the required result:

1. The Crushed Cain Weight, Bagasse, Filter Cake, Molasses and Sugar extracts are

read into the command window using the following commands:

§ DAY_NUM =xlsread('data.xlsx', 1, 'A2:A91'); %DAY NUMBER

§ CANE_WEIGHT =xlsread('data.xlsx', 1, 'B2:B91'); %CANE WEIGHT

§ BAGASSE =xlsread('data.xlsx', 1, 'C2:C91'); %BAGASSE

§ FILTER_CAKE =xlsread('data.xlsx', 1, 'D2:D91'); %FILTER CAKE

§ MOLASSES =xlsread('data.xlsx', 1, 'E2:E91'); %MOLASSES

§ SUGAR =xlsread('data.xlsx', 1, 'F2:F91'); %SUGAR

2. Using the CANE_WEIGHT value from the read data as a parameter feed, an

instant of the prediction model function is then implied using the following

command line:

§ PRED_1=predictFxn(CANE_WEIGHT,0)

3. The result thus generated is a 90-day analysis of varying sugarcane weight. It

contains seven (7) columns as listed in description 9 above.

4. Thus the result thus generated is saved in excel format using the command line:

§ xlswrite('prediction_output',PRED_1);

5. This action saves the result in the excel file named prediction_output.xls which

can be accessed from the project directory.

 74

6. To compare the output Bagasse, Filter cake molasses and sugar Extracted using

the model, with their counterparts obtained from field data, the following

commands are used:

§ BAGASSE_PRED =xlsread('prediction_output.xls', 1, 'C2:C91');

%BAGASSE_EXTRACT_PREDICTED

§ FILTER_CAKE_PRED =xlsread('prediction_output.xls', 1, 'D2:D91');

%FILTER_CAKE_EXTRACT_PREDICTED

§ MOLASSES_PRED =xlsread('prediction_output.xls', 1, 'E2:E91');

%MOLASSES_EXTRACT_PREDICTED

§ SUGAR_PRED =xlsread('prediction_output.xls', 1, 'F2:F91');

%SUCROSE_EXTRACT_PREDICTED

7. The values for SUGAR and SUGAR_EXTRACT_PREDICTED are plotted

against DAY_NUM.
 75

CHAPTER FOUR

4.0 RESULTS AND DISCUSSION

4.1 RESULTS

The results obtained in this research included the following:

A) Source Code as presented in section 3.7.2.1a

B) The software for simulating sugar production;

C) Table of Simulated values from the MATLAB model/ software (shown below in

section 3.7.2.1b);

D) Graphical comparisons of Field versus Theoretical values of sugar and its by-

products comprising of bagasse, scum and molasses presented in Figures 4.1, 4.2,

4.3 and 4.4 respectively; and,

E) Table of Analysis of Variance (ANOVA) shown in table 4.1.

4.2 Discussion on Sugar

4.2.1 The comparative behavior of factory versus predicted sugar results

Figure 4.1 below represents the curves of sugar generated over a period of 90 days

(3months), a typical factory production results as against the sugar predicted for the same

period using the same quantity as input.

The values were obtained by mass balance calculations and the process did not

distinguish different categories of cane received (Vernom, 1996) such as variety, cycle

etc.

75
 76

Figure 4.1: Sugar comparison curves between Field and Model Predicted Values

Taking a critical look at the graphs, it was observed that the model predictions

and the factory-based curves were in agreement since the maintained the same pattern

throughout the range of 90 day production period. However some minor cases of slight

variations could be seen the two curves in the first twenty (20) days, between the fiftieth

to sixtieth (50 - 60) days and the 75th – the 90th day. In most cases, the model simulated

curve assuming a slightly peaks comparatively.

The most likely reasons these variation even though we may not expect the two curves to

be naturally the same could be ascribed to:

i) Efficiency factor since the Model has a design efficiency of 75%: the variations

in local factory conditions with respect to lower or higher efficiencies probably

due to aging machines could have been responsible for the differences. If the

field efficiency of the milling system is above the MATLAB model efficiency

of 75%, then, this may be responsible the observed trend of some slight curve
 77

heights. On the other hand, a higher efficiency of the model equally project

higher curves. Most machines in the factory have been operating for over thirty

(30) years at a highly reduced efficiency. This fact can be accepted as evidence

considering the rather relatively smaller variations in the compared values of the

bye products especially that of bagasse in figure 4.2 as well as appendices I and

II respectively. This is so in view of the fact that bagasse is a fibrous insoluble

solid matter (Brane, 1996), it is classical and its value change cannot be so

significant naturally. The cause may also be due to production error as these

declining values always occurs in sequence of days within the same interval.

ii) Imbibition is a factor that constitute the factory’s milling efficiency. Low

shredding/crushing of the cane at the respective mills may have resulted in more

imbibitions water at the expense of partially ruptured cane cells: the result of

this is that more water might have been added which some sucrose which could

have been extracted by the water might be conveyed away as part of bagase.

While the prospective sugar has been lost as sucrose in the bagasse, more

imbibition has on the other hand been generated which will require more steam

energy powering to extract through the evaporators in an effort to achieve the

required raw sugar (Thaval and Kent, 2012b).

iii) The outstanding values of sugar generated by the model compared to those of the

factory environment as reflected in the results may have also been caused by juice

heating below or above the optimum temperature since it is known in principle

that low temperatures often results in juice inversion or alcohol formation and

excessive temperature leads to carmilization of juice.

 78

iv) Doses of additives like lime, coagulants etc may have in some cases within the

investigation period been misapplied; for instance, phosphate requirements in

most cases is ≥ 200ppm (g/kg) and cold liming is PH of 4.5 while hot liming

occurs at 8+or -2pH to achieve an optimum of 7 ± 1 PH to account for the

property of clarified juice.

v) Brix entering the evaporator may have fallen outside the required range of 13-

16% or brix leaving the evaporator(s) may have exceeded 60-65%.

vi) Use of Module: some factories including the one within which this research work

was conducted instead of using models rather use modules for predictions of

sugar production. Modules work on the principle of Tons Cane per Tons Sugar

(TCTS) which is an assumption index. It provided for example that given an input

of 30,000tons of cane, 10tons of sugar could be expected. The empericallity of

this index is therefore so much so that another TCTS value can be adopted other

than 10 at some other time due to certain assumption process or systems. Hence

the model guarantees a precise figure which is constant at fixed efficiency.

4.3 Results and Discussion on Bye Products of Sugar

4.3.1 Bagasse

Bagasse may be referred to as a primary bye product of sugar production. It is the first

and only product that leaves the production line immediately after coming out from the

last mill, hence it does not go through the rigours and long process of production to have

fully retrieved but it comes out early and it is used to aid the process that produced it

thereby being used to complete the process that produced it by way of utilizing it to

power steam into the boilers, heaters, evaporators, centrifuging, and eventually
 79

crystallizing and dehydration sugar to the final production stage. Bagasse is represented

in figure 4.2 below.

Figure 4.2: Bagasse Comparison Curves between Field and Model Predicted Values

The curves comparing the amount of bagasse through a factory process with that

of a model developed in this work as presented were obtained from data from appendices

I and II respectively. The curves indicate a close agreement between the two

comparative conditions. Bagasse appears to be constant in its value but for some little

liquid that might always be left trapped in the cells of the fibres.

Bagasse as in plates in appendix IV is an essential raw material for the production

of paper and boards in addition to the former. The values observed in appendix I and II

agrees fully with Antoine (2000) and Vernom et al 1996 with regards to the value or

proportion of it can be expected from crushing 1000kg of cane.

 80

4.3.2 Filter cake (scum or mud)

Filter cake is the second bye product which is normally extracted after bagasse

and it is the smallest in quantity among the three major bye products of sugar. Filter cake

produced from field and the simulated values are shown in figure 4.3 comparatively. The

curves are both so low below 100 tons compared to those of bagasse and molasses. The

close relationship between the graphs and similarity in pattern connotes agreement

between them and suggests little or insignificant variations between the two curves.

There is however an over-bloated rise between day 10 to 15. The most likely cause of this

may be a case of higher production within that period or the possibility of some left over

molasses which was not accounted for the previous day, and it was wrongly weighted

along with the following days production.

Figure 4.3: Comparison curves of Filter cake Field and Model Predicted Values
 81

4.3.3 Molasses

Molasses is the final bye product of sugar that always quits the process last, but

after which is the sugar finally realized. It is a liquid which is known to possess a very

high proportion of water in it with some traces of un-extracted sugar and other minor

impurities. It is a valid raw material in the in the liquor production industry. It is

important to note that of all the bye products of sugar production, non is thrown away as

waste but are all utilized in one thing or the other.

Molasses comparative results between factory and model simulated results is

presented graphically in figure 4.4 below. The curves as can be seen demonstrate a close

agreement arising from the values obtained in appendix I and II. The values and the graph

are in such a pattern that agrees with the conventional pattern found in modern sugar

factories. This fact goes further to some valid assessment of the MATLAB model.

Figure 4.4: Comparison curves of Molasses Field and Model Predicted Values
 82

The relatively higher peaks observable in the pattern of the curves of the

MATLAB model is a likely indication of the model’s more precise ability to extract the

molasses fluid form the mixed juice.

4.4 Analysis of Variance (ANOVA) Discussions:

The mean values obtained from the field and using the developed models for

sugar production and the by-products which include baggase, filter cake and molasses

where analysed to determine any significant difference between the means. Analysis of

Variance (ANOVA) was carried out using GenStat Analytical Software (Discovery

Edition 3) at 1 % (p<0.01) probability level. The ANOVA table for bagasse, filter cake,

molasses and sugar products are shown below.

***** Analysis of variance *****

Variate: BAGASSE

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 361684. 361684. 3.33 0.070
Residual 178 19328528. 108587.
Total 179 19690212.

Variate: FILTER_CAKE

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 2457. 2457. 1.71 0.192
Residual 178 255391. 1435.
Total 179 257848.

Variate: MOLASSES

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 3183.0 3183.0 3.81 0.053
Residual 178 148755.4 835.7
Total 179 151938.3

Variate: SUGAR

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 5656.5 5656.5 6.19 0.014
Residual 178 162758.6 914.4
Total 179 168415.1
 83

From Table 4.1 showing the Least Significant Difference (LSD) at 1 % probability level

p<0.01), the mean value obtained for the bagasse from the field (735 tons) and from the

developed model (645 tons) were not significantly different at 1 % (p>0.01) probability

level. Similarly no significant differences were observed between the means obtained for

filter cake and molasses at the 1 % (p>0.01) probability level. For the sugar product, the

mean values obtained from the field and from the model were observed and means were

not significantly different (p>0.01) at 1 % probability level.

Table 4.1: Least Significant Difference Obtained from the ANOVA

Product Data source Mean value LSD
(tons) 1%
Bagasse Field 735 127.9 ns
Model 645
Filter cake Field 56.7 14.70 ns
Model 49.3
Molasses Field 56.7 11.22 ns
Model 65.1
Sugar Field 53.8 11.74 ns
Model 65.1
Mean values with LSD having the superscript ‘ns’ indicate ‘not significantly different’ at the given
probability level

Since the ANOVA presented in table 4.1 above shows no significant difference between

the sugar, bagasse, filter cake and molasses obtained from Savannah Sugar Factory and

the MATLAB model developed, the MATLAB model is therefore validated.

 84

CHAPTER FIVE

5.0 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS

5.1 SUMMARY

An existing model that relies on classical mass balance provided the platform for

an extension and application of several mathematical equations which provided for the

simulation of sugar production.

The major objective of this research is the prediction of Sugar Yield from Sugar

Cane. Sugar value is often not known or estimated until production is completed in the

factory at every given occasion. Sugar is therefore a variable which was expected to be

predicted. Other variables that accompany the production of sugar as major by-products

are bagasse, molasses, filter cake and waste water; hence the research “Prediction of

Sugar Yield from Sugar Cane” explored the corresponding values of these by-products as

well.

The development of MATLAB model was done in three stages. Firstly, an

enhanced milling unit extraction (mathematical) model was developed to access the mill

performance parameters and predict the extraction performance of the milling process.

New definitions for the milling performance parameters were developed and model was

tested and compared with factory data from Savannah Sugar Company, Numan.

Variations in the mill performance of the mill performance parameters were observed and

studied.

A soluble and insoluble solids model was developed using modeling theory and

experimental data to track the amount of sucrose, bagasse, molasses, filter cake and waste

water in the milling train through the cane supply.

84
 85

The MATLAB software is an already existing mathematical analysis and

simulation software, with vast subsidiary projections. The modality of this research has

more need for simulation software than a proprietary software package (standalone

software developed in Java, for instance), hence, the choice of MATLAB. Several

graphical representations of the outcome analysis were also projected from the MATLAB

simulation report.

Certain aspects of the sugar extraction process such as nucleation of crystals

present themselves as complex phenomena owing to the complexity of interacting

variables. Also certain effects like impurities, on the kinetics is nonlinear are

unpredictable (Lauret et al, 2000; Thaval, 2012). The modeled equations made use of the

input value of crushed cane as a decision variable in which sugar yield was obtained to be

279.57T/2,150.542 metric tons representing 13% of the input value; hence its bye

products obtained suggest the following proportions: 48%, 5.2%, 4.7% of bagasse,

molasses and filter cake per total mass of cane (metric tons) respectively. The results of

the MATLAB simulation, the data generated from it was subjected to Analysis of

Variance (ANOVA) to ascertain the validity of the developed model at 99% probability

level. Sucrose (Sugar), bagasse, filter cake and molasses values of the analysis of

variance showed no significant difference.

5.2 CONCLUSION

From the results of the studies the following conclusions were drawn:

1. The MATLAB model developed is C + I = Cj+ Mm + Imp. + B

Where, C = sugar cane; I = imbibition water; Cj = clarified juice, Mm = mud

mixture, Imp = impurities in the juice and, B = bagasse

 86

2.The MATLAB model is capable of predicting sugar yield from sugar cane, with

efficiency of 75%.

3. The predicted yield sugar yield and that of the field data were in agreement with

each other.

There was no significant difference (at 99% probability) between Sucrose

(Sugar), bagasse, filter cake and molasses values obtained from Savannah Sugar

Company and the values generated from the MATLAB model.

5.3 RECOMMENDATIONS

1. Improvements may be made on this model to enhance its usability. The model
used here is a Novel model within the Nigerian community done using linear
equations; further works should performed using non-linear equations for
increased accuracy of the systems.
2. Optimization indices may be introduced on this model as part of enhancement
preferably to promote sugar production in Nigeria especially now that more
new innovations are emerging on sugar production such as the new ABU
Zaria sugar biotechnology factory and the Lau sugar company in
Taraba State.

5.4 CONTRIBUTIONS TO THE LITERATURE/KNOWLEDGE

This study enriched the literature with a model for estimating the sugar, bagasse,

molasses and filter cake yield from sugar cane for the production of sugar at the

Savannah Sugar Company, Numan, Adamawa State, Nigeria. The model can be an

engineering management tool for assessing the production process showing poor

performing units requiring improvement without running the plant. This eliminates the

wastage and optimizes production in the factory. The study fosters the much talked about

marriage between research and the industry.

 87

REFERENCES

Alder M. D.(2001). An Introduction to Mathematical Modelling. Heavens for For

Books.com.Pp193-110.

Ale S. O.(2011). Matlab/Scilab For Data Analysis And Visualization: Platform For

National Development National Centre for Equipment Maintenance &

Development (NCEMD) Equip. Maintenance Course, Vol. 5, pp.11-16.

Antoine, R. (2000). Sugar Industry Training Center, Maurituis Pp4-12.

Arrascaete, A. and Friedman, P.(1987). Energy Management – Bagasse drying.

International Sugar Journal Vol. 89 No.1060 Pp68 -71.

Atiku, N.A.(1999). Strategies for Achieving Self-sufficiency in Sugar Production in

Nigeria, Focus on Savannah Sugar Company, Numan, Adamawa State. Pp1-10

Barnes, A. C. (1974). The Sugarcane: Leonard Hill Books

Bennett M. C.(1957). The Physical Nature and Behavior of Cane Sugar juice.
International Sugar Journal Pp 208-212

Bennett M. C.(1975). Fluctuation Technology in Sugar Manufacture. Sugar industry

Technologists Paper 381 Pp22-32

Billings S. A.(2013). Nonlinear System Identification NARMAX Methods in Time,

Frequency, and Sparto. Temporary Domains.
Bull, T. (2000). The Sugarcane Plant. In "Manual of cane growing", M Hogarth, P
Allsopp, eds. Bureau of Sugar Experimental Stations, Indooroopilly, Australia. pp
71-83.

Busari,L.D. and W.J. Wayagari(1999). Project Profile and Viability, Appraisal for NCRI
10 TCD Mini Brown Sugar Processing Plant Natural Cereal Research Institute
Badeggi. Pp 4- 9

Bureau of Sugar Experiment Stations (1984). The standard laboratory manual for
Australian sugar mills. (Vol. 1 Principles and Practices).

Bureau of Sugar Experiment Stations (2001). The standard laboratory manual for
Australian sugar mills. (Vol. 2 Analytical methods & tables).

Chen J. C. P and Chou C. C.(1993). Cane sugar Handbook: A Manual for Cane Sugar
Manufacturers and Their Chemists.12th edition.Pp 27-33, 113-166 and 604-614.
 88

Clarke,M.A. and M.A. Godshell(1987). Chemistry and Processing of Sugar beet and
Sugarcane: Proceedings of the Symposium and Chemistry and Processing of Sugar
beet, Denver Colorado. Pp175-178.

Cock, J.H. and Carlos A.L.G.(1995). Overview of Sugarcane Production in Colombia,

ISS CT XX11 Congress Features Pp406-407. International Sugar Journal Vol. 97.

Curriculum K-12 Directorate(undated). NSW, Department of Education and Training.

Dalal, S. R.(2003). Handbook of Reliability Engineering; Software Reliability models: A

Selective Survey And New Directions. Edited by Hoang Pam Springer-Verlag
London Limited. P201 retrieved 02/11/2014 http://www.springer.de/phys/

Dangote Sugar Refinery Plc: Unaudited Group Results for Nine Months Ended 30
September 2014 retrieved on 21/12/2015 http://dangote.com/downloads/DSR-9-
months-2014-results-release.pdf-

Earle, R.L. (2002). Unit Operations in Food Processing, , NZIFST

Engelman C.(1964). MATLAB: A program for on-line assistance in symbolic

computations. Proceedings, FJCC .Spartan Books, Washington D.C.

Enibe S. O.(2013). MATLAB/SCILAB FOR Scientific and Engineering Analysis, Pp1-7.

El-Belghi K., Rabhi Z. and E.Vorobiev(2005). Journal of the Science of Food and
Agriculture. Kinetic Model of Sugar Diffusion from Sugar beet tissue treated by
Pulsed Electric field. 85: 213-218.
Factfish Sugar cane, Production World Statistic and Data. Retrieved on 21/12/2015 from
http://www.factfish.com/country

Foster, D. H. (1956). Analytical methods used in assessing mill tandem performance.

Proceedings of LX International Society of Sugar Cane Technologists Congress.

Indian Sugar Mills Association (ISMA)2010 http://www.indiansugar.com/Sugar.aspx

17/12/2015

GAIN Report (2013). USDA Agricultural Services: Global Agricultural Information

Network, Pp 1- 6. Retrieved on 17/12/2015
http://gain.fas.usda.gov/Recent%20GAIN%20Publications/Sugar%20Annual_Lago
s_Nigeria_5-23-2013.pdf.

Gupta, P. K. and Hira, D. S. (2008). Operations Research, S Chad and Company Ltd,
New Delhi, India.
 89

Hahn B. H. and D. T. Valentine(2010). Essential MATLAB for Engineers and

Scientists,4th edition; Academic Press is an imprint of Elsevier. 30 Corporate Drive,
Suite 400, MA 01803, USA, Linacre House, Jordan Hill, OX2 DP, UK.

Hartel R. W.(2004). Food Engineering – Vol III. Crystallization. Pp1-9

Lauret, P.,Boyer, H. and Gatina, J. C.(2000). Hybrid Modelling of Sugar Boiling Process.
Control Engineering Practice, International Federation of Automatic Control 2000,
8(3), Pp299-310.<hal-00766248>

Liang, Y.D. (2004). Introduction to Java Programming; Comprehensive Version 6th

edition, Pp12- 88
Mackintosh, D. (2000). Sugar Milling. In "Manual of cane growing", M Hogarth, P
Allsopp, eds. Bureau of Sugar Experiment Stations, Indooroopilly, Australia. pp
369-377.

MURRY, C. R.(1959). Preliminary bagasse tests in the two roll experimental mill.
Proceedings of Australian Society of Sugarcane Technologists, 26, 67-72.

NSDC ,Federal Ministry of Industry, Trade and investment Federal Republic of Nigeria
retrieved on 17/12/2015.http://www.nsdcnigeria.org/

Okusanya, B.A. and O.M. Ajogbasile,(1980). Sugarcane Disease in Nigeria, A

Symposium of Sugarcane Industry in Nigeria. Pp1-9.

Patrick M. and Thomas O. (2013). Holland Graphics and GUIs with MATLAB .Chapman
& Hall/CRC. Third edition.

Pidduck, J. (1955). Physical properties of bagasse. Proceeding so f Queensland Society of

Sugar Cane Technologies, 22, 147-155.

Quarteroni A.(2009). Mathematical Models in Science and Engineering. Vol. 56

Number1. Pp 10-19.

Ramos, O.G.(1997). The Sugarcane Tradition in Colombia. XXII, International Congress

of Sugarcane Technologies, International Sugar Journal, Vol.XCVII issue No.1161,
Pp417-420.

Rein P (2007). Cane Sugar Engineering. Berlin: Bartens

Rediff.com (2015). Sugar-producing-countries-in-the-world/20130219.htm retrieved on

07/09/2015 http://www.rediff.com/business/slide- show/slide-show-1-25-biggest-

Russel, G. E. (1968). An investigation of the extraction performance of sugarcane

crushing trains. Ph.D Thesis, University of Queensland.
 90

Sharma, J.K. (2008). Operation Research. Theory and Applications; 3rd Edition.
Macmillian India Ltd, New Delhi, India

Smejkal, Q., Bagherzadeh, A. and R.Shcick(2008). Influence of Temperature on

Crystallization of Sugar. Technical University of Berlin, Published by Sugar
Industry, vol.133(5).Pp330-339

Sotudedeh-Gharegh et al (2009). Modeling and Optimization of the Sugar Extraction

Process. International Journal of Food Engineering, vol.5 Iss4. Art5; Berkely
Electronic Press. Pp1-9

Suarez et al(2008). Model Predictive Control Strategies for Batch Sugar Crystallization
Process.

Thaval ,O. P. and Kent, G. A.(2012a). Modeling the Flow of Juice Through a Mill.
International Sugar Journal 114 (1363), Pp484-489. Retrieved on 20 07/2015
http://eprints.qut.ed.au/58521/.

Thaval ,O. P. and Kent, G. A.(2012b). Modeling the flow of cane constituents in the
milling process. In Nel, Sanet, Peacock, Steve, Ramgareeb, Sumita, Sweby,
Deborah, Carslow, Dorothy, & Macdonald, Danile (Eds.) Proceedings of the 85th
Annual Congress of the South AfricanSugar Technologists’ Association, South
African Sugar Technologists’ Association, Durban, South Africa, pp. 435-453.

USDA Foreign Agricultural Service (http://www.whichcountry.co/which-country-is-the-

largest-producer-of-sugar-cane-in-the-world/)

Van der Pol C and Alexander JB (1955). Decomposition of sucrose in the milling
process. Proc S Afr Sug Technol Ass 29: 46-53..

Vermeulen, D.(2006). How Sugar is Made. Vil. 1 Sugar Pp1-8.

Vernom, R. A. (1996). Sugarcane: Printed in Great Britain by Printall Co:22

Ogmore Crescent, Bridgend Mid-Glamorgan. Published by International Media Ltd, P.O.

|Box 26 Port Talbolt, West Glamorgan, United kingdom.

Walford SM (1996). Composition of cane juice. Proc S Afr Sug Technol Ass 70: 265-266.

Wiley and Sons - Wayne C. Turner (2006). Energy Management Handbook,

Wikipedia (2012). Free encyclopedia retrieved 15/3/2012.

en.wikipedia.org/…/https://en.wikipedia.org/wiki/Verification_and_validation_of_c
omputer_simulation_models

Wikipedia (2015). Free encyclopedia retrieved on 18/12/15. Pp1 -3

 91

en.wikipedia.org/…/https://en.wikipedia.org/wiki/Verification_and_validation_of_compu
ter_simulation_models

Yates, R.A.(1996). Factors that Affect the Quality of Sugarcane. Issue No.2 Published by
International Media Ltd , West Glamorgan, United Kingdom. Pp1-10.
"http://www.appropedia.org/index.php?title=Sugar_Production_from_Cane_Sugar_
(Practical_Action_Technical_Brief)&oldid=131600"
 92

APPENDICES

APPENDIX I
Field Data of sugar Production and the bye products obtained for 90 Days
(all weighs are in metric tons)

DAY CANE WEIGHT BAGASSE FILTER CAKE MOLASSES SUGAR

1 1453.75 391.31 28.9 28.8 16
2 1412.55 999.01 67 73 61
3 1565.87 831.84 57.7 60.4 70
4 872.16 454.24 30.4 32.6 50
5 1838.15 1031.01 62.5 77 80
6 880 447 29.93 36.6 18
7 1579.24 918.47 71.1 66.2 39
8 1902.01 1120 79.9 79.7 94
9 203 12.4 0.8 6.9 27
10 1631.7 903.18 65.3 68.4 98
11 1690.33 969.53 65.9 70.8 80
12 445.33 250.68 153 18.7 30
13 1288.25 725.97 435 54 50
14 193.29 114.89 8.7 8.1 21
15 1066.9 594.4 40.5 44.7 67
16 1331.09 704.59 51.9 55 20
17 1440.22 784.99 49 60.3 74
18 1537.5 829.45 52.28 64.4 70
19 907.3 487.08 38.1 38 40
20 563.04 327.24 21.4 23.6 13
21 1596.8 817.2 63.9 84 70
22 2005.08 1055.58 78.2 84 80
23 101.54 52.35 3.5 4.3 52
24 1889.14 1051.04 64.26 79.2 84
25 1368.71 746 48.91 51.3 60
26 1875.93 1063.23 84.4 78.6 32.9
27 714.26 401.13 27.1 29.9 20
28 975.39 574.35 41 40.9 56
29 1606.09 947.86 72.3 100.9 37
30 1023.61 602 34 64.3 46
31 1611.02 904.53 54 101.3 46
32 1446.07 731.26 56.4 90.9 33
33 1575.66 866.21 63 66 68
34 376.04 217.3 14.3 15.8 20
 93

35 461.55 217.3 14.3 15.8 21

36 1689.59 970 76 70.8 43
37 1494.5 832.8 50.81 90.9 54
38 901.24 496.55 30.6 37.8 74.38
39 1870.08 1100.04 72.9 78.4 85
40 2196.48 1197.72 87.9 92 105
41 551.03 326.83 20.9 23.1 12
42 1509.63 797 63.2 63 70
43 2110.2 1169.29 95 88.4 59
44 1593.66 899.55 54.18 100.2 68
45 2150.93 1163.03 73 90.1 86
46 820.7 451.77 32 34.4 35
47 1914.16 1115.4 76.6 80.2 70
48 2004.48 1154.21 76.2 84 41
49 809.48 435.12 34 33.9 87
50 2120.64 1194.7 95.4 88.9 43
51 390.96 219.33 13.29 24.6 11
52 1928.84 1044.34 65.6 80.8 73
53 1901.43 1105.18 75 81.2 84
54 1314.24 722.75 50 52.4 73
55 912.47 625.25 50 52.4 73
56 223.51 1514.78 93.2 93 66
57 198.2 1297.37 92.2 85.8 28
58 2143.12 1180.62 72.39 89.2 54
59 1516.3 826.52 64.4 91.1 45
60 2048.16 1398.75 81.7 87.8 62
61 651.48 512.07 26.4 27.6 20
62 1169.55 744.41 48.1 48 72
63 2139.55 1297.56 96.4 89.8 107
64 757.9 398.56 26.34 32.4 81
65 1911.36 1040.4 64.4 91.1 118
66 2216.97 1316.45 84.9 91.2 65
67 378.72 235.85 15.9 16.6 21
68 259.67 151.07 10.1 10.9 0.5
69 622.87 368.87 24.9 26.1 25
70 258.01 171.71 25 27.6
71 1259.36 324.04 52.9 52.8
72 1474.4 853.11 50.13 61.8 18
73 1340.19 615.14 33.2 43.6 57
74 421 244.7 16.4 17.6 5
75 1051.16 603.06 47.3 44.1 67
 94

76 885 517.38 30.09 37.1 42

77 122.05 706.12 39.1 51.2 33
78 1051.16 603.06 47.3 44.1 67
79 2255.19 1319.19 84.4 44.2 126
80 1222.75 706.12 39.1 51.2 3
81 1550.62 837.22 60.5 65 4
82 925.76 528.52 37 38.8 43
83 911.42 540 38.3 38.8 43
84 1664.5 968.14 74.9 69.7 82
85 484.48 272.68 16.47 20.3 7
86 1220.75 695.65 62.62 51.1 88
87 1463.04 848.47 55.6 61.3 45
88 1027.22 586.68 43.1 43 59
89 1610.44 859.44 72.5 67.5 72
90 1555.14 889.5 52.87 65.2 100
 95

APPENDIX II

Theoretical Results (data) of sugar Production and the bye products obtained From MATLAB Simulation for 90 replications.

CANE FILTER
DAY BAGASSE SUGAR MOLASSES IMBIBITION TOTAL DIFFERENCE
WEIGHT CAKE
1 1453.75 749.698875 56.69625 57.27775 75.595 354.715 1293.982875 159.767125
2 1412.55 728.452035 55.08945 55.65447 73.4526 344.6622 1257.310755 155.239245
3 1565.87 807.519159 61.06893 61.695278 81.42524 382.07228 1393.780887 172.089113
4 872.16 449.772912 34.01424 34.363104 45.35232 212.80704 776.309616 95.850384
5 1838.15 947.933955 71.68785 72.42311 95.5838 448.5086 1636.137315 202.012685
6 880 453.816 34.32 34.672 45.76 214.72 783.288 96.712
7 1579.24 814.414068 61.59036 62.222056 82.12048 385.33456 1405.681524 173.558476
8 1902.01 980.866557 74.17839 74.939194 98.90452 464.09044 1692.979101 209.030899
9 203 104.6871 7.917 7.9982 10.556 49.532 180.6903 22.3097
10 1631.7 841.46769 63.6363 64.28898 84.8484 398.1348 1452.37617 179.32383
11 1690.33 871.703181 65.92287 66.599002 87.89716 412.44052 1504.562733 185.767267
12 445.33 229.656681 17.36787 17.546002 23.15716 108.66052 396.388233 48.941767
13 1288.25 664.350525 50.24175 50.75705 66.989 314.333 1146.671325 141.578675
14 193.29 99.679653 7.53831 7.615626 10.05108 47.16276 172.047429 21.242571
15 1066.9 550.20033 41.6091 42.03586 55.4788 260.3236 949.64769 117.25231
16 1331.09 686.443113 51.91251 52.444946 69.21668 324.78596 1184.803209 146.286791
17 1440.22 742.721454 56.16858 56.744668 74.89144 351.41368 1281.939822 158.280178
18 1537.5 792.88875 59.9625 60.5775 79.95 375.15 1368.52875 168.97125
19 907.3 467.89461 35.3847 35.74762 47.1796 221.3812 807.58773 99.71227
20 563.04 290.359728 21.95856 22.183776 29.27808 137.38176 501.161904 61.878096
 96

21 1596.8 823.46976 62.2752 62.91392 83.0336 389.6192 1421.31168 175.48832

22 2005.08 1034.019756 78.19812 79.000152 104.26416 489.23952 1784.721708 220.358292
23 101.54 52.364178 3.96006 4.000676 5.28008 24.77576 90.380754 11.159246
24 1889.14 974.229498 73.67646 74.432116 98.23528 460.95016 1681.523514 207.616486
25 1368.71 705.843747 53.37969 53.927174 71.17292 333.96524 1218.288771 150.421229
26 1875.93 967.417101 73.16127 73.911642 97.54836 457.72692 1669.765293 206.164707
27 714.26 368.343882 27.85614 28.141844 37.14152 174.27944 635.762826 78.497174
28 975.39 503.008623 38.04021 38.430366 50.72028 237.99516 868.194639 107.195361
29 1606.09 828.260613 62.63751 63.279946 83.51668 391.88596 1429.580709 176.509291
30 1023.61 527.875677 39.92079 40.330234 53.22772 249.76084 911.115261 112.494739
31 1611.02 830.803014 62.82978 63.474188 83.77304 393.08888 1433.968902 177.051098
32 1446.07 745.738299 56.39673 56.975158 75.19564 352.84108 1287.146907 158.923093
33 1575.66 812.567862 61.45074 62.081004 81.93432 384.46104 1402.494966 173.165034
34 376.04 193.923828 14.66556 14.815976 19.55408 91.75376 334.713204 41.326796
35 461.55 238.021335 18.00045 18.18507 24.0006 112.6182 410.825655 50.724345
36 1689.59 871.321563 65.89401 66.569846 87.85868 412.25996 1503.904059 185.685941
37 1494.5 770.71365 58.2855 58.8833 77.714 364.658 1330.25445 164.24555
38 901.24 464.769468 35.14836 35.508856 46.86448 219.90256 802.193724 99.046276
39 1870.08 964.400256 72.93312 73.681152 97.24416 456.29952 1664.558208 205.521792
40 2196.48 1132.724736 85.66272 86.541312 114.21696 535.94112 1955.086848 241.393152
41 551.03 284.166171 21.49017 21.710582 28.65356 134.45132 490.471803 60.558197
42 1509.63 778.516191 58.87557 59.479422 78.50076 368.34972 1343.721663 165.908337
43 2110.2 1088.23014 82.2978 83.14188 109.7304 514.8888 1878.28902 231.91098
44 1593.66 821.850462 62.15274 62.790204 82.87032 388.85304 1418.516766 175.143234
45 2150.93 1109.234601 83.88627 84.746642 111.84836 524.82692 1914.542793 236.387207
 97

46 820.7 423.23499 32.0073 32.33558 42.6764 200.2508 730.50507 90.19493

47 1914.16 987.132312 74.65224 75.417904 99.53632 467.05504 1703.793816 210.366184
48 2004.48 1033.710336 78.17472 78.976512 104.23296 489.09312 1784.187648 220.292352
49 809.48 417.448836 31.56972 31.893512 42.09296 197.51312 720.518148 88.961852
50 2120.64 1093.614048 82.70496 83.553216 110.27328 517.43616 1887.581664 233.058336
51 390.96 201.618072 15.24744 15.403824 20.32992 95.39424 347.993496 42.966504
52 1928.84 994.702788 75.22476 75.996296 100.29968 470.63696 1716.860484 211.979516
53 1901.43 980.567451 74.15577 74.916342 98.87436 463.94892 1692.462843 208.967157
54 1314.24 677.753568 51.25536 51.781056 68.34048 320.67456 1169.805024 144.434976
55 912.47 470.560779 35.58633 35.951318 47.44844 222.64268 812.189547 100.280453
56 223.51 115.264107 8.71689 8.806294 11.62252 54.53644 198.946251 24.563749
57 198.2 102.21174 7.7298 7.80908 10.3064 48.3608 176.41782 21.78218
58 2143.12 1105.206984 83.58168 84.438928 111.44224 522.92128 1907.591112 235.528888
59 1516.3 781.95591 59.1357 59.74222 78.8476 369.9772 1349.65863 166.64137
60 2048.16 1056.236112 79.87824 80.697504 106.50432 499.75104 1823.067216 225.092784
61 651.48 335.968236 25.40772 25.668312 33.87696 158.96112 579.882348 71.597652
62 1169.55 603.136935 45.61245 46.08027 60.8166 285.3702 1041.016455 128.533545
63 2139.55 1103.365935 83.44245 84.29827 111.2566 522.0502 1904.413455 235.136545
64 757.9 390.84903 29.5581 29.86126 39.4108 184.9276 674.60679 83.29321
65 1911.36 985.688352 74.54304 75.307584 99.39072 466.37184 1701.301536 210.058464
66 2216.97 1143.291429 86.46183 87.348618 115.28244 540.94068 1973.324997 243.645003
67 378.72 195.305904 14.77008 14.921568 19.69344 92.40768 337.098672 41.621328
68 259.67 133.911819 10.12713 10.230998 13.50284 63.35948 231.132267 28.537733
69 622.87 321.214059 24.29193 24.541078 32.38924 151.98028 554.416587 68.453413
70 258.01 133.055757 10.06239 10.165594 13.41652 62.95444 229.654701 28.355299
 98

71 1259.36 649.451952 49.11504 49.618784 65.48672 307.28384 1120.956336 138.403664

72 1474.4 760.34808 57.5016 58.09136 76.6688 359.7536 1312.36344 162.03656
73 1340.19 691.135983 52.26741 52.803486 69.68988 327.00636 1192.903119 147.286881
74 421 217.1097 16.419 16.5874 21.892 102.724 374.7321 46.2679
75 1051.16 542.083212 40.99524 41.415704 54.66032 256.48304 935.637516 115.522484
76 885 456.3945 34.515 34.869 46.02 215.94 787.7385 97.2615
77 122.05 62.941185 4.75995 4.80877 6.3466 29.7802 108.636705 13.413295
78 1051.16 542.083212 40.99524 41.415704 54.66032 256.48304 935.637516 115.522484
79 2255.19 1163.001483 87.95241 88.854486 117.26988 550.26636 2007.344619 247.845381
80 1222.75 630.572175 47.68725 48.17635 63.583 298.351 1088.369775 134.380225
81 1550.62 799.654734 60.47418 61.094428 80.63224 378.35128 1380.206862 170.413138
82 925.76 477.414432 36.10464 36.474944 48.13952 225.88544 824.018976 101.741024
83 911.42 470.019294 35.54538 35.909948 47.39384 222.38648 811.254942 100.165058
84 1664.5 858.38265 64.9155 65.5813 86.554 406.138 1481.57145 182.92855
85 484.48 249.846336 18.89472 19.088512 25.19296 118.21312 431.235648 53.244352
86 1220.75 629.540775 47.60925 48.09755 63.479 297.863 1086.589575 134.160425
87 1463.04 754.489728 57.05856 57.643776 76.07808 356.98176 1302.251904 160.788096
88 1027.22 529.737354 40.06158 40.472468 53.41544 250.64168 914.328522 112.891478
89 1610.44 830.503908 62.80716 63.451336 83.74288 392.94736 1433.452644 176.987356
90 1555.14 801.985698 60.65046 61.272516 80.86728 379.45416 1384.230114 170.909886
 99

APPENDIX III: ANALYSIS OF VARIANCE (ANOVA) CALCULATIONS

GenStat Release 7.22 DE (PC/Windows) 18 December 2015 13:04:59

Copyright 2015, VSN International Ltd

The GenStat Discovery Edition can be used for educational or not-for profit
research purposes in qualifying countries. A list of qualifying countries can
be viewed at http://discovery.genstat.co.uk. Commercial use of the
GenStat Discovery Edition is strictly prohibited.

________________________________________

GenStat Discovery Edition 3

GenStat Procedure Library Release PL15.2
________________________________________

105.............................................................................

***** Analysis of variance *****

Variate: BAGASSE

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 361684. 361684. 3.33 0.070
Residual 178 19328528. 108587.
Total 179 19690212.

***** Tables of means *****

Variate: BAGASSE

Grand mean 690.

Factor A B
735. 645.
 100

*** Standard errors of differences of means ***

Table Factor
rep. 90
d.f. 178
s.e.d. 49.1

*** Least significant differences of means (5% level) ***

Table Factor
rep. 90
d.f. 178
l.s.d. 96.9

110.............................................................................

***** Analysis of variance *****

Variate: BAGASSE

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 361684. 361684. 3.33 0.070
Residual 178 19328528. 108587.
Total 179 19690212.

***** Tables of means *****

Variate: BAGASSE

Grand mean 690.

Factor A B
735. 645.

*** Standard errors of differences of means ***

 101

Table Factor
rep. 90
d.f. 178
s.e.d. 49.1

*** Least significant differences of means (1% level) ***

Table Factor
rep. 90
d.f. 178
l.s.d. 127.9

115.............................................................................

***** Analysis of variance *****

Variate: FILTER_CAKE

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 2457. 2457. 1.71 0.192
Residual 178 255391. 1435.
Total 179 257848.

***** Tables of means *****

Variate: FILTER_CAKE

Grand mean 53.0

Factor A B
56.7 49.3

*** Standard errors of differences of means ***

Table Factor
rep. 90
d.f. 178
 102

s.e.d. 5.65

*** Least significant differences of means (5% level) ***

Table Factor
rep. 90
d.f. 178
l.s.d. 11.14

120.............................................................................

***** Analysis of variance *****

Variate: FILTER_CAKE

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 2457. 2457. 1.71 0.192
Residual 178 255391. 1435.
Total 179 257848.

***** Tables of means *****

Variate: FILTER_CAKE

Grand mean 53.0

Factor A B
56.7 49.3
*** Standard errors of differences of means ***

Table Factor
rep. 90
d.f. 178
s.e.d. 5.65

*** Least significant differences of means (1% level) ***

 103

Table Factor
rep. 90
d.f. 178
l.s.d. 14.70

125.............................................................................

***** Analysis of variance *****

Variate: MOLASSES

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 3183.0 3183.0 3.81 0.053
Residual 178 148755.4 835.7
Total 179 151938.3

***** Tables of means *****

Variate: MOLASSES

Grand mean 60.9

Factor A B
56.7 65.1

*** Standard errors of differences of means ***

Table Factor
rep. 90
d.f. 178
s.e.d. 4.31

*** Least significant differences of means (5% level) ***

Table Factor
rep. 90
 104

d.f. 178
l.s.d. 8.50

130.............................................................................

***** Analysis of variance *****

Variate: MOLASSES

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 3183.0 3183.0 3.81 0.053
Residual 178 148755.4 835.7
Total 179 151938.3

***** Tables of means *****

Variate: MOLASSES

Grand mean 60.9

Factor A B
56.7 65.1

*** Standard errors of differences of means ***

Table Factor
rep. 90
d.f. 178
s.e.d. 4.31

*** Least significant differences of means (1% level) ***

Table Factor
rep. 90
d.f. 178
l.s.d. 11.22
 105

135.............................................................................

***** Analysis of variance *****

Variate: SUGAR

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 5656.5 5656.5 6.19 0.014
Residual 178 162758.6 914.4
Total 179 168415.1

***** Tables of means *****

Variate: SUGAR

Grand mean 59.5

Factor A B
53.8 65.1

*** Standard errors of differences of means ***

Table Factor
rep. 90
d.f. 178
s.e.d. 4.51

*** Least significant differences of means (5% level) ***

Table Factor
rep. 90
d.f. 178
l.s.d. 8.90

140.............................................................................
 106

***** Analysis of variance *****

Variate: SUGAR

Source of variation d.f. s.s. m.s. v.r. F pr.

Factor 1 5656.5 5656.5 6.19 0.014
Residual 178 162758.6 914.4
Total 179 168415.1

***** Tables of means *****

Variate: SUGAR

Grand mean 59.5

Factor A B
53.8 65.1

*** Standard errors of differences of means ***

Table Factor
rep. 90
d.f. 178
s.e.d. 4.51

*** Least significant differences of means (1% level) ***

Table Factor
rep. 90
d.f. 178
l.s.d. 11.74
 107

APPENDIX IV:EQUATIONS FOR THE MATHEMATICAL MODEL

•

• The Model in Equation 2.7 as postulated by Thaval (2012b) is written thus:

• ṁc + ṁI = ṁJm + ṁB

• so

• The process model becomes:

• C + I = Mj + B

• where (3.1)

• B = Bagasse

• Mj = Mixed juice

• C = Cane

• I = Imbibition water

• But Mj = A + Imp

Where

• A = S + Nw + Mm (3.2)

• A = absolute juice

• Imp = Impurities in the juice

• S = Sucrose (sugar)

• Nw = Natural water in the juice

• So:

• Mj = S + Nw + Mm + Imp (3.3)

• The Model in Equation 2.7 as postulated by Thaval (2012b) is written thus:

• ṁc + ṁI = ṁJm + ṁB

• so
 108

• The process model becomes:

• C + I = Mj + B

• where (3.1)

• B = Bagasse

• Mj = Mixed juice

• C = Cane

• I = Imbibition water

• But Mj = A + Imp

• Where

• A = S + Nw + Mm (3.2)

• A = absolute juice

• Imp = Impurities in the juice

• S = Sucrose (sugar)

• Nw = Natural water in the juice

• So:

• Mj = S + Nw + Mm + Imp (3.3)

According to Antoine (2000), for every 1000kg of cane crushed, Bagasse is 488kg

: . B = 1, 049.46T

From (3.1)

Mj = C-B +I (3.13)

75
Nw = X weight of cane (Brane 1974)
100

From (3.9)
 109

S + Imp = C-Nw-Mm + I = 997.862T

9.9
Mm = approx x wt Cane (Based on Production parameter; Chen and Chou, 1993 and
100

Brane 1995)

But Mm = Fc + Ml

So, S + Imp = 997.862 (3.14)

From (3.10)

Cj + Imp = C – Mm + I

Cj + Imp = 2610.762T (3.15)

From equation (3.14); Imp = 997.862 - S

From equation (3.15); Imp = 2610.762 - Cj

NB: Imp Þ 997.862 –S = 2610.762 - Cj

Sucrose(S )= 279.57

Note that;

Mud mixture Mm = the fraction of yet to be extracted quantities of Molasse and filter

cake in the absolute juice with the emergence of equation.

Mm = Fc + Mc (3.17)

Molasses, (M) according to Chen and Chou 1993, asserted that there 40-52kg of

Molasses in every one ton of crushed cane)

: . Now converting all known weights given above to percentages:

Mjw
Mixed juice, Mjp percent = Mjp x 100 (3.18)
Cw

: . Mjp = 82.54%
 110

Also imbibition water added in percentages

Iw
Ip = x 100 (3.19)
Cw

Percent weights of Bagasse

Bw
BP = x 100 (3.20)
Cw

Determining the percentage of filter cake (Fc):

Fcw
Fc p = x100 (3.21)
Cw
And also for molasses percentage
Mw
(Mp) = x 100
Cw

Natural water (Nw) contained in the crushed weighed 1612.9T

: . Percentage of the water, NWp:

The following weights of of sugar(C), Bagasse(B), Molasses(M) and Filter cake(Fc) were

converted to their corresponding percentages of the input cane value respectively as

follows:

S = 13%C Þ0.13C

Nwp = 75% Þ 0.75C

Bp = 48.76%C Þ0.488C

Mp = 5.2%C Þ0.052C

Fcp = 4.7% Þ 0.047C

It is worthy of note that:

 111

ii) the mass percentage of sugar produced is in conformity with the table of values of

sugarcane composition by Barnes (1974).

iii) the equations and the MATLAB model that culminated in the resulting output

were derivatives of the mass balance model as proposed by Thaval (2012b).

Hence, to account for the reabsorption factor and filling ratio of the milling train,

efficiency of the machine must be taken into consideration for each component of the

sugarcane. For instance,

(3.22)

Where Se = Extracted Sucrose (kg)

= Efficiency of the milling train.

 112

APPENDIX V: WORLD PRODUCTION OF SUGAR CANE CONT’D FROM

TABLE 1.2

Table 1.2 World Sugar cane Production Quantity (tons) 2013-2015 cont’d

Country Rank Value as number

Venezuela 21 7.339.634

El Salvador 22 7.162.995

Ecuador 23 7.158.265

Nicaragua 24 7.041.333

Sudan 25 6.797.900

Iran 26 6.200.000

Honduras 27 6.082.000

Kenya 28 5.900.000

Paraguay 29 5.544.797

Swaziland 30 5.450.000

Costa Rica 31 4.850.000

Dominican Republic 32 4.771.211

Bangladesh 33 4.434.000

Zambia 34 4.000.000

Zimbabwe 35 3.880.000

Mauritius 36 3.815.782

Moambique 37 3.800.000

Uganda 38 3.350.000

Madagascar 39 3.250.000
 113

Tanzania 40 2.992.177

Nepal 41 2.930.000

Malawi 42 2.900.000

Ethiopia 43 2.750.000

Panama 44 2.482.183

Guyana 45 2.460.171

Congo, Democratic Republic of the 46 2.000.000

Cote d’lvoire 47 1.968.917

Fiji 48 1.600.000

Nigeria 49 1.455.000

Jamaica 50 1.402.600

Cameroon 51 1.200.000

Haiti 52 1.200.000

Japan 53 1.191.000

Laos 54 1.180.000

Balize 55 1.078.000

Sri Lanka 56 960.270

Senegal 57 890.000

Trinidad and Tobago 58 810.000

Congo, Republic of the 59 650.000

Morocco 60 619.561

Cambodia 61 600.000

Angola 62 510.000
 114

Burkina Fso 63 480.000

Chad 64 390.000

Uruguay 65 375.000

Mali 67 345.000

Papua New Guinea 68 300.000

Gabon 69 280.000

Barbados 70 280.000

Liberia 71 265.000

Burundi 72 223.000

Somalia 73 220.000

Malaysia 74 213.978

Niger 75 190.000

Virgin Islands, U.S. 76 150.000

Ghana 77 150.000

Suriname 78 120.100

Rwanda 79 115.000

Central African Repulic 80 110.000

Saint Kitts and Neis 81 100.000

Afghanistan 82 89.880

Sierra Leone 83 77.000

Bahamas 84 57.500

Benin 85 52.000

Puerto Rico 86 40.000

 115

Cape Verde 87 28.375

Saint Vincent and the Grenadines 88 18.100

Bhutan 89 15.000

Saint Lucia 90 10.000

Iraq 91 8.168

Grenada 92 7.200

Guinea-Bissau 93 6.350

Antigua and Barbuda 94 5.365

Portugal 95 5.350

Dominica 96 4.800

Lebanon 97 4.700

Anguilla 98 3.562

French Polynesia 99 3.500

Oman 100 1.250

Yemen 101 500

Spain 102 100

Syria 103 55

Djibouti 104 52

Singapore 105 50

American Samoa 106` 29

Wallis and Futuna 107 21

Samoa 108 12

Source: Fatfish Sugar cane, 2015 http://www.factfish.com/country

 116

APPENDIX VI

PLATES OF BYE PRODUCTS TAKEN AT THE SAVANNAH SUGAR COMPANY, NUMAN

PLATE 2a : Bagasse at savannah sugar company, Numan

 117

Plate 2b: Bagacillo savannah sugar company, Numan

 118

Plate 2c:Filter cake taken at savannah sugar company, Numan

 119

Plate 2d : sugar produced at the bagging room of savannah sugar company, Numan