Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE THESIS MASTERS IN TECHNOLOGY (BIOTECHNOLOGY & BIOPROCESS ENGINEERING) OPTIMISATION Qf TOMATO PASTE PRODUCTION, STORAGE AND USE C.F Campbell BY CLAIRE FRANCES CAMPBELL SUPERVISORS: Dr J E Bronlund &DrL VChong 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE ABSTRACT The manufacture of tomato products is one of the key activities of Heinz-Watties. The optimisation of tomato paste in terms of it's production, storage and use is important to understand as it is the base ingredient to many of the 21 major food brands Heinz-Watties supply in New Zealand. This thesis describes the work carried out to characterise tomato paste products in terms of the quality and quantity, correlations that exist between variables, the process variability at Heinz-Watties Hastings and formulation using tomato paste. This thesis has identified that tomato paste needs to be accurately characterised in terms of both quality and quantity. Characteristics which influence the quality of tomato paste are factors such as the flavour, sweetness and viscosity to which the 0 Brix, total solids and insoluble or soluble solids must be known. 0 Brix measurement is affected by insoluble solids (which scatter light) resulting in an inaccurate reading when employing the refractometer 0 Brix method. To solve this, the tomato paste must be centrifuged in order to separate the insoluble from soluble portion within the tomato paste. Colour, acidity and microbial stability are also important quality characteristics of tomato paste. These are measured by specific tomato paste colorimetric methods, pH and Howard mould count methodologies respectively. The quantity of tomato paste produced or the amount to use within recipes has been identified in this thesis to be accurately measured by measuring the tomato solids, insoluble and soluble solids and 0 Brix. This is due to the tomato paste being made up of total tomato solids and water which can be further broken down into solids which are insoluble (fibrous) & solids which are soluble. Several different methods were identified within the thesis for measuring total solids. The best method in terms of repeatability was identified to be the vacuum oven method, however due to the twelve hour wait prior to obtaining results coupled with the large amount of equipment needed this would be restricted to a lab environment. Within the factory processing environment the best total solids method to employ would be that of the microwave oven method. Further testing beyond the scope of this thesis would need to be completed to perfect the methodology. C.F Campbell 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE 11 Correlations between variables were also explored within this thesis, to save time and equipment usage within the processing environment or to give an indication of variables during production. For example specific gravity can be measured instead of total solids ( as long as the insoluble solids portion is known) and with the use of correlations the other parameters can be predicted. Further experimental validation beyond this thesis, using Heinz-W atties tomato paste does need to occur prior to use. ' Other correlations investigations were that of the variables during dilution. The 0 Brix levels and viscosity were found to not change linearly during dilution. Therefore, by constructing simple mechanistic models on the interactions of the 0 Brix levels and viscosity on the proportions of insoluble and soluble solids simple equations have been devised within this thesis to allow the prediction of these parameters after dilution. The current process of producing tomato paste at Heinz-W atties was characterised to identify the extent and cause of process variability. This work showed that although total solids levels are well controlled, the ratio of insoluble solids to total solids is not. The cause of this was most likely due to poor control over the break process and the extent of enzymatic pectin hydrolysis that occurs. Some suggestions on online measurement options to enable better control of this were explored within this thesis, such as, measuring specific gravity using an online densitometer (Coriolis mass flow meter) and to use previously mentioned correlations to determine total solids. Or alternatively online viscosity by the use of a tubular viscometer or refractive index meter. Further work should be carried out beyond this thesis to investigate how tomato ripeness and break processing conditions could be controlled to ensure reduced variability in the ratio of insoluble solids to total solids. This is the key to good control of tomato paste and diluted tomato paste viscosity. C.F Campbell 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE lll ACKNOWLEDGEMENTS I would like to thank and acknowledge the assistance and guidance of a number of inspirational and literally amazing people of which I would have never have completed this project. Firstly a big thanks to my Massey supervisors, Dr John Bronlund and Dr Li Chong (and their respective partners) who have been my mentors and inspiration throughout the tomato paste events. Next, to all the staff of Heinz-Watties, King Street, Hastings, for their support throughout the project and with an unlimited willingness to assist and supply tomato paste and more tomato paste. In particular at Heinz-Watties, Mr Daya Vithanage with an un-limiting level of assistance and support, Mr Colin Watson, Mr Paul Brizzle, Mr A vin Krishnan, Mr Danny Eagleton, Mrs Sandra Chambers, Mrs Dale Cooley, Ms Martha Flynn whose experience and expertise was invaluable. Many, many thanks go out to all my friends whose help (from the joys of stirring tomato paste to making me dinner, absent of the tomato paste, thankfully), encouragement and understanding proved invaluable and to which I am sincerely grateful for. Also a big thanks to my boyfriend (Stuart), Uncle Roy and to Mrs Wendy and Mr Dave Thomas, whose kind hearts and support at the crucial stage of writing up were invaluable. Many thanks goes to my many work colleagues at whose encouragement and understanding through writing up while working full time is cherished. And most importantly to my Mum (Kay), my Dad (Joe), my brother (Peter) whose help, love, guidance and inspiration throughout was unfathomable and of which I cannot express in words but can only say it was such a powerful force that keep me striving for the goal ahead. I dedicate this piece of work to you three super-special people who each in your own pursuits have carved, and continue to carve everyday, a special report of skills and understanding that is simply indescribable and to which I have the utmost respect and admiration for. C.F Campbell 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE CONTENTS PAGE Chapter. 1 2 DESCRIPTION Abstract Acknowledgements Contents Page List of Figures List of Tables Project Overview Review of Literature 2.1 2.2 2.3 Introduction Definitions Composition 2.3.1 Composition of tomatoes Pg. i ii iv viii xi 1 2 2 4 5 5 IV 2.3 .1.1 Structure and major constituents of the 2.4 C.F Campbell tomato fruit 5 2.3.1.2 Water 7 2.3 .1.3 Carbohydrates 7 2.3.l.4 Protein and fat 9 2.3 .1.5 Nutrients, vitamins & acids 9 2.3.1.6 Carotenoids & chlorophylls 11 2.3.2 Composition of tomato paste 12 2.3.2.1 Water 13 2.3.2.2 Carbohydrates 13 2.3.2.3 Protein 14 2.3.2.4 Nutrients 14 2.3.2.5 Vitamins & acids 15 2.3.2.6 Carotenoids & chlorophylls 15 2.3.2.6.1 Lycopene 15 Chemical Changes in Tomato Systems 2.4.1 Changes during ripening 2.4.2 Pectin & enzyme activity 2.4.3 Changes during processing 17 17 18 19 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE 3 C.F Campbell 2.5 2.6 2.7 Physical Properties & Measurements 2.5.1 Variety 2.5.2 Soluble solids 2.5.3 Total Solids 2.5.4 Colour 2.5.5 Density Predictions 2.5.6 Rheology 21 21 22 23 25 27 28 2.5.6.l Experimental data and rheological model fitting 29 2.5.6.2 Viscosity 41 Production Process 43 2.6.1 Tomatoes are received from the field 43 2.6.2 Preparation of tomatoes for processing 44 2.6.3 Hot break process 45 2.6.4 Warm break process 45 2.6.5 Packaging of the tomato paste for later use 45 2.6.6 Flow chart 47 Quality 2. 7.1 Quality within the field 2.7.1.1 Soil-Borne Micro-organisms 2.7.2 Quality of the field 2. 7 .2.1 Field Diseases 2. 7.3 Quality in processing 48 49 49 50 50 51 Characterisation of Tomato Products 54 3.1 Introduction 54 3.2 Identification of key properties for measurement 55 3.3 Separation of the pulp and serum fractions of tomato paste 65 3.4 Soluble solids measurement 66 3.4.1 Brix measurement principles 66 3.4.1.1 Operation 66 3.4.1.2 Calibration 67 3.4.1.3 Mechanism 67 3.4.1.4 Time & temperature 68 3.4.2 Initial investigations 69 3.4.3 Effect of time on °Brix readings 71 3.4.4 Effect of temperature on "Brix readings 72 3.4.4.1 Methodology 72 3.4.4.2 Results 72 3.4.5 Effect of insoluble solids on °Brix measurements 74 3.5 Total solids Measurement 78 3.5.1 Methods 79 3.5.1.1 FAO 79 3.5.1.2 MOS 79 3.5.1.3 VOS 79 3.5.1.4 Freeze 80 16/08/04 V OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE vi 4 5 C.F Campbell 3.6 Insoluble solids measurement 83 3.7 pH measurement 85 3.8 Colour measurement 86 3.9 Mould count measurement 87 3.10 Conclusions 88 Correlations Between Variables 90 91 91 4.1 SGand TS 4 .2 4 .3 4.4 4.1.1 Literature 4.1.1.1 4.1.1.2 4.1.1.3 4.1.2 Modelling 4.1.3 Temperature 4.1.4 Comparisons Dr Andy Crawford's correlation 91 Tomato Bulletin data 92 Heinz Watties Analytical Bench Manual 92 94 . 98 99 4.1.5 Experimental data 4.1.6 Inline mass-flow meter 100 102 103 0 Brix, SS, IS and TS 4.2.1 Modelling 4.2.2 Experimental Viscosity 4.3.1 4.3.2 Modelling Experimental 4.3.2.1 4.3.2.2 Method Results 4.3.2.2.1 4.3.2.2.2 4.3.2.2.3 4.3.3 Viscosity Conclusion Correlation Conclusions 4.4.1 SG vs TS 4.4.2 °Brix 4.4.3 Viscosity 103 105 106 106 110 110 111 Power law 111 Behaviour index 112 Consistency index 113 116 117 117 117 118 Process Variability 120 5 .1 Long term variability over a tomato paste production season 120 5.1.1 °Brix 121 5.1.2 Total solids 122 5.1.3 Specific gravity 123 5.1.4 Bostwick consistency 125 5.1.5 More detailed examination of in-plant variability 127 16/08/04 6 7 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE vu 5.2 Preliminary investigation on causes on variability in insoluble 5.3 5.4 solid levels 130 5.2.1 Pectin levels in initial fruit 130 5.2.2 Ripeness of fruit 130 5.2.3 Proportion of fruit ripeness 132 5.2.4 Break tank operation 133 5.2.5 Fruit variety variability through variety differences 133 5.2.6 Post production storage 134 On-line measurement of process variability 135 Conclusions 140 Formulation 141 6.1 Introduction 141 6.2 Derivation of equations 142 6.2.1 Dilution of total, soluble and insoluble solids 142 6.2.2 Dilution with respect to °Brix 143 6.2.3 Effect of dilution on specific gravity (density) 146 6.2.4 Effect of dilution on apparent viscosity 148 6.3 Experimental validation of dilution equations 150 6.3.1 Dilution of total, soluble and insoluble solids 150 6.3.2 Dilution with respect to °Brix 152 6.3.3 Effect of dilution on specific gravity (density) 153 6.3.4 Effect of dilution on apparent viscosity 154 6.4 Conclusions 156 Conclusions & Recommendations 159 References 168 Appendices 176 Appendix One. Varieties of tomatoes used by Heinz W atties Ltd 17 6 Appendix Two. Tomato paste process flow chart 177 C.F Campbell 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE viii UST OF FIGURES Figure. 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 C.F Campbell Description Pg. Cross-Section of a Tomato Fruit 5 Shear rate verses apparent viscosity for 25% and 30% total solids 29 hot break tomato paste at 32°C Graph of shear stress vs temperature for 25% and 30% total solids (%TS) for hot break tomato paste at a constant shear rate of 5oos·1 (Temperature Range 32.3-82.2°C) K values for hot break (HB) and cold break (CB) tomato paste vs total solids (%TS) using exponential and power equations to determine K (consistency index). Apparent viscosity vs shear rate for hot break and cold break tomato paste using the simple power law as per Rao et al. ( 1981) and Harper & El Sahrigi (1965) Yield stress vs total solids for hot break (HB), cold break (CB) and all concentrates of tom~to paste Apparent viscosity vs shear rate for hot break (HB) and cold break (CB) tomato paste using the Casson model Shear stress vs shear rate for cold break tomato paste using the Herschel-Bulkley model (TS=30%) Flow chart of tomato paste processing Mould count to the %minimum rot (by weight) in tomato paste samples as per Goose & Binsted (1973) Munsell colour system Production of hexenal products· Refractometer working principle 0 Brix vs time for whole warm break tomato paste 0 Brix vs g insoluble solids/g tomato paste Quality parameter vs g insoluble solids/g tomato paste 0 Brix vs quality Progressive decrease in °Brix of the wash waters as the insoluble solids fraction is washed and re-centrifuged in order to isolate the insoluble solids fraction 30 33 34 37 38 40 47 52 57 61 68 71 75 76 77 83 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE ix 4.1 Specific gravity (SG) vs %total solids (%TS) using Dr 93 Crawford, Tomato Bulletin and Heinz Watties analytical manual data correlations at 25°C 4.2 Specific gravity (SG) vs % total solids (%TS) at differing 96 insoluble solids to total solids ratios (ISffS) at 25°C using Choi & Okos (1986) model with literature data overlaid. 4.3 Specific gravity (SG) vs % total solids (%TS) using 98 Choi & Okos (1986) density functions at different temperatures 4.4 Specific gravity (SG) vs % total solids (%TS) at different ratios 100 of insoluble solids to total solids (ISffS) for warm break tomato paste 4.5 Specific gravity (SG) vs %total solids (%TS) of experimental 101/124 values with Choi & Okos (1986) model (using CRC Handbook 1988) density values) and literature data at differing ISffS ratios overlaid ( also repeated in Chapter 5) 4.6 0 Brix vs fraction of total solids (XTs) over a range of Xrs!XTS 104 typical of that of warm break tomato paste at 30% total solids 4.7 0 Brix vs total solids fraction (XTs) at differing XrslXTs 105 fractions of experimental data and model data overlaid 4.8 Apparent viscosity vs total solids (%TS) 109 4.9 Log apparent viscosity vs log shear rate 111 4.10 Flow behaviour index (n) vs insoluble solids content (%Xrs) 112 4.11 Consistency index (k) vs insoluble solids (Xrs) 113 4.12 Log insoluble solids content (Xrs) vs Log consistency index (k) 114 4 .13 Arrhenius plot (ln(viscosity) vs lff(°K) 115 4.14 Prediction of apparent viscosity vs % insoluble solids content 116 for tomato paste at 5oos·1 and actual experimental data overlaid 5.1 0 Brix vs samples over the production period of 20/1/01 121 to 25/4/01 5.2 Total solids (%TS) vs samples over the production 122 period of 20/1/01 to 25/4/01 5.3 Specific gravity (SG) vs samples over the production 123 period of 20/1/01 to 25/4/01 C.F Campbell 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE X 5.4 Bostwick consistency vs samples over the production 125 period of 20/1/01 to 25/4/01 5.5 Bostwick consistency vs specific gravity (SG) over the 126 production period of 20/1/01 to 25/4/01 5.6 Solids content (X1s, Xss and XTs) vs time of processing 128 5.7 Fraction of insoluble solids to total solids (Xis/XTs) vs 128 time of processing 5.8 Total solids (%TS) vs time of processing 129 5.9 Enzyme activity vs ripeness of tomatoes entering the 131 tomato paste process 5.10 Green and breaker tomatoes proportion within truckloads 132 vs time from 17/4/00 0:05 to 21/4/00 15:11 (approximately 3-4 truckloads per hour) 6.1 0 Brix(undiluted)/0 Brix(diluted) vs water added (W) 144 6.2 0 Brix(u)/0 Brix(d) ratio vs amount of water added (W) 145 with different insoluble solids contents (r = XIScu/XTS(u)) of the original paste changing 6.3 Specific gravity predicted (SG) vs amount of water added (W) 147 with different insoluble solids contents CX1slXTs) 6.4 Apparent viscosity ratio vs dilution 149 6.5 Total solids (%TS) vs water added(%) with average model 150 overlaid 6.6 0 Brix vs water added(%) with average model overlaid 152 6.7 Specific gravity (SG) vs water added 153 6.8 Apparent viscosity ratio of diluted tomato paste/undiluted 154 tomato paste vs water added C.F Campbell 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE xi UST OF TABLES Table No. Description Pg. 2.1 Composition of a tomato 6 2.2 Daily dietary allowance of nutrients for USA and NZ 10 2.3 Composition of tomato paste 12 2.4 Polysaccharide content within an average tomato and tomato 22 paste 2.5 Simple power law experimental values used in Figure 2.2 31 2.6 Constants for consistency index equations for hot break and 32 cold break tomato paste at 25°c 2.7 Coefficient values for detennining yield stress for the Casson 37 model (eqtn 2.5) 2.8 Variables for the Casson model as used in Figure 2.7 38 2.9 Variables for Herschel-Bulkley model used in Figure 2.8 39 3.1 Repeatability of 0 Brix measurements 69 3.2 Summary of 0 Brix measurements over a range of temperatures 72 3.3 Total solid method comparison 78 3.4 Summary of total solid measurement results 80 3.5 Advantages and disadvantages of the four total solid methods 81 3.6 Repeatability of the pH measurement 85 3.7 Precision of colour measurements using the Minolta digital 86 colorimeter 3.8 Precision of the Howard mould count 87 4.1 Variables used in Rao et. al. ( 1981) 109 5.1 Specifications for Endress and Hauser coriolis mass flowmeter; 136 model Promass 631 C.F Campbell 16/08/04 OPTIMISATION OF TOMATO PASTE PRODUCTION, STORAGE AND USE 1.0 PROJECT OVERVIEW Tomato paste is used as a base ingredient to many of the 21 major food brands Heinz­ Watties supply and it is estimated at having an input value of $4.8 million per annum in factories around New Zealand. At the Hastings site of Heinz-Watties, tomato paste is produced during the months of November to March ( depending on harvesting of the tomatoes) and used all year round for products such as spaghetti, baked beans, soups, meat and pasta sauces. During the time tomato paste is not being used, it is stored aseptically in 200L drums outside, within the factory premises. The optimal utilisation of the tomato paste is reliant on the following issues: i) The adequate control of the paste quality during paste manufacture. ii) The accurate recording of stock levels in terms of quantity and quality of the stored product iii) The appropriate calculation of the amount of paste used during the formulation of the final products. Each of the aspects listed above are dependent on the accurate characterisation of the tomato paste that is produced. This requires appropriate measurement methods which are being used that are repeatable, have the accuracy needed and are also suitable for use in an industrial environment. If such methods are available then the stock levels would accurately reflect the amount of tomato paste ingredient is in storage and the manufacturing process could be better controlled to produce more consistent paste quality. In addition to this the most appropriate measure of the paste composition could be used to calculate the amount of paste needed to achieve the required formulated product functionality (in terms of such aspects as flavour and colour). The specific aims of this project were to firstly determine the most appropriate measurement methods for complete tomato paste characterisation. These methods were then used to assess the degree of variability of tomato paste production during the 2000 season at Heinz-Watties Australasia, King Street, Hastings. The project also investigated · how physical and compositional properties of tomato paste are correlated and how they can be related to the functionality of formulated products. C.F. Campbell 16/08/04