The present study was aimed at extending our knowledge of the antioxidative properties of the milk protein products, whey protein isolate (WPI) and sodium caseinate (NaCas), in oil-in-water (O/W) emulsions rich in polyunsaturated fatty acids (PUFAs). In particular, the objective was to contribute to our understanding of the compositional and processing factors that influence the oxidative stability of protein-stabilised O/W emulsions. Linoleic acid (approximately 60 %) was used as the lipid for the oil phase (10.6 %). The emulsion samples were usually incubated at 50 °C to accelerate lipid oxidation. Lipid oxidation indicators were lipid hydroperoxides and headspace hexanal, determined by solid phase microextraction (SPME) combined with gas chromatography (GC). WPI- or NaCas-stabilised emulsions were prepared using a wide range of protein concentrations (0.5, 1.0, 2.0, 3.0, 4.0, 7.0 or 10.0 %) at two droplet sizes (d32 = 0.31 and 0.65 µm). In general, higher lipid oxidation levels were found for the larger droplet size. Increasing protein concentration led to a decrease in the lipid oxidation rate. The greatest decrease in lipid hydroperoxide levels (values after 4 h) occurred at up to 4.0 % protein concentration. The greatest decrease in hexanal levels (values after 24 h) occurred at up to 4.0 % protein concentration in WPI emulsions (0.31 µm). The hexanal levels were more independent of the protein concentration in the other emulsion types. The hexanal level decreased at protein concentrations > 4.0 % in NaCas emulsions (0.31 and 0.65 µm) and at protein concentrations > 7.0 % in WPI emulsions (0.65 µm). The difference between lipid hydroperoxide generation in emulsions with small and large droplet sizes decreased with increasing protein concentration. This effect was more pronounced in NaCas emulsions. In general, NaCas was a better inhibitor of lipid oxidation than WPI, but WPI appeared to be the better antioxidant at some droplet size/protein concentration combinations. The protein in the continuous phase, i.e. the unadsorbed protein, played an important role in lipid oxidation. In principal, the lipid hydroperoxide and hexanal levels showed the same development over the continuous phase protein concentration as over the protein concentration in WPI and NaCas emulsions (d32 = 0.31 µm). A low NaCas level in the continuous phase already led to a relatively low hexanal level, whereas a higher WPI level was required. When NaCas solution was added to a WPI emulsion or WPI solution was added to a NaCas emulsion, a synergistic antioxidative effect was observed. The high molecular weight fractions (molecular weight = 12000-14000) of WPI and NaCas contained pro-oxidative metal ions that contributed to lipid oxidation in the emulsions. An enrichment of NaCas emulsions with the low molecular weight fraction of NaCas (with a molecular weight = 12000-14000) notably inhibited lipid oxidation. An enrichment of WPI emulsions with the low molecular weight fraction of WPI (with a molecular weight = 12000-14000) also seemed to inhibit lipid oxidation, but the effect was not significant. The protein solutions were enriched with these fractions before emulsion preparation. Pure WPI solution or mixed WPI/NaCas (1:1, weight/weight) solution with 1.12 or 2.24 % protein concentration was heated at 84 °C for up to 40 min, cooled and then used to prepare emulsions. Lipid oxidation was generally not affected by the heat treatment or the degree of whey protein denaturation. However, at the lower WPI concentration, more hexanal was produced for the longer heating times (20, 30 and 40 min) and this appeared to be connected with the physical instability of the emulsions. Greater oxidative stability was found at the higher protein concentration and when the proteins were mixed, pointing to a possible synergistic antioxidative effect of WPI and NaCas. The addition of the free radical source 2,2’-azobis(2-amidinopropane) dihydrochloride (AAPH) greatly increased the oxygen uptake and the generation of lipid hydroperoxides in the emulsions. The oxidative stability increased with increasing protein concentration (1.0, 4.0 and 7.0 %). NaCas had a greater antioxidative effect than WPI. The inhibition of oxygen uptake appeared to be largely influenced by the free-radical-scavenging activity of the system, determined by the protein type and the protein concentration, as the radicals were produced linearly over time and oxygen was consumed linearly over time. It can therefore be concluded that free-radical-scavenging activity represents a major antioxidative mechanism of the milk proteins. Oxygen was consumed much faster in emulsions than in protein solutions when the same level of AAPH was incorporated. In a WPI (1.0 % protein) emulsion, much lower levels of protein hydroperoxides than of lipid hydroperoxides developed. This pointed to a much greater reactivity of linoleic acid than of the milk proteins with oxygen. In contrast, the exposure of WPI to oxidising linoleic acid in an emulsion (1.0 % protein) or to AAPH in aqueous solution led to oxidative damage of the whey proteins, indicated by the loss of amino acids. The loss of specific amino acids was different for proteins in the continuous phase or cream phase of an emulsion or in WPI solution. The present study confirms the antioxidative potential of WPI and NaCas and gives new insights into their functionality as oxidative stabilisers in O/W emulsions.

Yeasts are traditionally used as part of the surface microflora in surface-ripened cheeses, where they contribute positively to the flavour of the cheese. The primary objective of this study was to investigate the potential of three dairy yeasts to provide attributes as adjuncts in anaerobically ripened cheeses. Geotrichum candidum (B9001), Yarrowia lipolytica (B9014) and Candida kefyr (B9006), obtained from the Fonterra Co-operative Group Ltd, Palmerston North, New Zealand, were studied. They showed diverse metabolic activities in laboratory media, which were influenced by the growth conditions. The metabolic activities of special interest were the lipase and proteinase activities and the production of volatile compounds, as these are important for cheese ripening and flavour development. Lipase activity (p-nitrophenyl butyrate assay) and proteinase activity (fluorescein isothiocyanate β-casein assay) were determined in three fractions prepared from yeast cultures and designated as extracellular fraction, washed-cell fraction and intracellular fraction. Lipase activity of G. candidum was detected only in the extracellular fraction and increased five fold when induced by safflower oil in a shake culture (0.16 µM/min/mL supernatant at 24 h). Lipase expression was delayed in static cultures. Y. lipolytica showed lipase activity in extracellular, washed-cell and intracellular fractions under all conditions. Static cultures in both glucose and safflower oil media showed higher lipase activity than shake cultures. The lipase activity of Y. lipolytica was higher in the late stationary phase than in the log phase under all conditions tested. The highest lipase activity was detected in a 192 h static culture grown in safflower oil medium (0.13 µM/min/mg dry cell weight, 0.3 µM/min/mg dry cell weight and 4.29 µM/min/mL supernatant in the intracellular, washed-cell and extracellular fractions respectively). C. kefyr did not show any lipase activity (< 0.03 µM/min/mL culture) under any of the growth conditions tested. Proteinase activity was detected in the intracellular fraction of 72 h shake cultures of G. candidum grown in both glucose medium and safflower oil medium (154 and 122 RFU/min/mg dry cell weight respectively) but was not detected in static cultures. Proteinase activity was absent in the Y. lipolytica cultures under all conditions tested (< 10 RFU/min/mL culture). C kefyr showed low proteinase activity (12-74 RFU/min/mL supernatant) in the extracellular fraction only in shake cultures grown in glucose medium. Volatile compounds of the headspace were sampled and analysed using solid phase microextraction (SPME) and gas chromatography-mass spectrometry (GC-MS). The concentrations of volatile compounds were highest in shake cultures grown in glucose medium for all three yeasts. All yeasts produced several alcohols. Several esters were also detected in the G. candidum and C. kefyr cultures whereas aldehydes were detected only in the G. candidum cultures. G. candidum and Y. lipolytica were selected for cheese production trials because of their active cheese ripening enzymes. These yeasts, grown under different conditions, were added to Cheddar cheese (10 L vat). The yeast adjuncts influenced the cheese ripening by lipolysis [in terms of the production of free fatty acids (FFAs) analysed by gas chromatography-flame ionisation detector (GC-FID)] and the production of volatile compounds (SPME-GC-MS), whereas proteolysis (analysed by size-exclusion high performance liquid chromatography) by yeast enzymes was not obvious. The influence of Y. lipolytica as an anaerobic adjunct to cheese ripening was dependent on the growth conditions used during its propagation in laboratory media. The concentration of total FFAs was very high (37.1 mg/g cheese at 6 months) when a 192 h Y. lipolytica culture grown in safflower oil medium was added to a cheese make, whereas the cultures grown in glucose medium did not have any detectable effect. Addition of G. candidum culture to the cheese curd was more effective than its addition to the cheese milk. Both G. candidum and Y. lipolytica lipase(s) selectively hydrolysed the long-chain unsaturated fatty acids from the milk triglyceride in the cheese environment. Also, Y. lipolytica lipase exhibited some selectivity towards hydrolysis of butyric acid from the milk fat in the cheese. 2-Heptanone, 3-methyl-2-butanone and 2-nonanone were detected (1-10 x 106 relative peak area) only in the cheeses with yeast adjuncts but not in the control cheese. Enhancement of the production of both conjugated linoleic acid (CLA) and ethyl esters in a washed-curd, dry-salted cheese (375 L vat), made with G. candidum, Y. lipolytica, Propionibacterium freudenreichii ssp. shermanii, Lactobacillus fermentum and Lb. rhamnosus, was only partially successful. Higher concentrations of ethyl esters (> five fold; analysed by SPME-GC-MS) were produced in the cheeses made with yeast adjuncts. However, the concentration of total CLA (free plus esterified; analysed by GC-FID) did not increase although a higher concentration of free linoleic acid (> 10 fold), the substrate for CLA synthesis, was produced in the cheeses made with yeast adjuncts. A study of the formation of aromatic volatile compounds by C. kefyr in a medium containing L-phenylalanine (L-phe) showed that the yeast's ability to produce phenyl ethanol, phenyl ethyl acetate and benzaldehydc (analysed by SPME-GC-MS) was enhanced with an increase in the initial L-phe concentration (in the experimental range; analysed by enzymatic assay using phenylalanine ammonia lyase), but the yield was very low (20-27%). The initial concentration of glucose (in the experimental range; analysed by enzymatic assay using Peridochrom glucose reagent) did not affect the production of these aromatic volatile compounds. This study successfully showed that the yeasts G. candidum and Y. lipolytica, when used as anaerobic adjuncts, can influence the ripening and flavour development in Cheddar and washed-curd, dry-salted cheeses. The study also showed the capability of C. kefyr to produce aromatic volatile compounds from amino acid fermentation but the yields need to be increased by further manipulation of the medium components and the culture conditions before this capability can be used commercially.

Richmond Ltd. has recently undergone a change in strategy, away from the traditional commodity based meat industry, towards the modern food business. To do this, opportunities to add value to their current product range must be identified. This involves the conversion of traditionally low value commodity based products into products that demand a premium. An example of this is converting muscles that are currently used for grinding meat into a further processed convenience food (i.e. ready meals). Another method is to add further value to premium products by making them more appealing to consumers (i.e. nutritional information on labels). This work details investigations into the functional properties of selected beef forequarter muscles (low value commodity products) and the nutritional properties of selected export lamb products (premium products). The functional properties of a number of beef forequarter muscles were measured to identify which had the best potential for further processing applications with respect to ready meals. The functional properties of tenderness, cook loss and shrinkage were measured for the Latissimus Dorsi, Pectorialis Profundus (Point End Brisket), Infraspinatus (Cross Cut Blade), Triceps Brachi Longhead (Main muscle in Bolar Shoulder Clod), Supraspinatus (Chuck Tender), Serratus Ventralis and Triceps Brachi Medialhead (Muscle in Bolar Shoulder Clod. From the tests conducted the Infraspinatus and the Triceps Brachi Longhead have been identified as having the best functional properties with respect to further processing for ready meal applications. As well as conducting tests to identify the forequarter muscles with the best potential for further processing applications, investigations were carried out to identify cooking regimes that would optimise the functional properties. This work confirmed that there are three major chemical reactions, which determine the resultant functional properties of cooked meat. They are the denaturation and aggregation of the myofibrillar proteins and the denaturation and solubilisation of connective tissue (collagen). At around 50°C myosin (45% to 50% of the myofibrillar proteins) denatures, which results in a substantial increase in cook loss and reduction in water holding capacity. At around 60°C collagen (main connective tissue protein) denatures, which results in a substantial increase in tenderness and increase in cook loss. This is because as the collagen denatures it loses it mechanical strength (increase in tenderness) and can no longer support its own structure, and causes it to contract. This contraction causes fluid within the meat and cook loss caused by the denaturation of myosin to be expelled from the meat by compressive forces (squeezed out). At around 70°C actomyosin (22% of the myofibrillar proteins) denatures. This results in a substantial increase in the cook loss and firming of the meat. The increase in cook loss or decrease in water holding capacity that occurs with myofibrillar protein denaturation is due to the fact that when these proteins denature and aggregate their ability to bind water is greatly reduced. From the results of the cooking regime trials it is recommended that for functional property considerations that during the cooking of further processed meat products (i.e. ready meal applications) a meat temperature of 62°C should be aimed for, for the slowest heating region during cooking (usually the centre). This is because it has been identified that a cooking temperature of 65°C should not be exceeded otherwise detrimental effects can occur to the functional properties of the cooked meat. For health concerns a 7D bacterial death reduction has to be achieved. This means that for a cooking temperature of 62°C the meat has to be held at this temperature for at least 5 minutes. Therefore the total cooking time would be the time needed to heat all the meat to 62°C plus 5 minutes to ensure a safe product. The heating or cooking system employed should also ensure that a minimal amount of the meat is heated above 65°C. This can be easily achieved by minimising the external cooking temperature, but long cooking times will result. An industrial cooking process will be a compromise between the cost associated with longer residence time and product functionality. As mentioned earlier another way to add value is to supply nutritional information for selected cuts. Consequentially one of the objectives of this project was to provide some nutritional information for selected meat cuts. Though the primary objective of this part of the project was to develop a method for producing the needed information, so that Richmond N.Z. Ltd. can develop further information on an as needs basis. The nutritional characteristics of a number of export lamb cuts from the saddle region has also been investigated and a method devised to allow further characterisation of other cuts. The method involves breaking down a standard cut into its constituent components (e.g. Frenched rack consists of loin eye, fat cap, intercostals and fatty tissue). The constituent components are tested for their nutritional properties. The frenched rack nutritional properties are calculated from the nutritional properties of the constituents components and the yield data (percentage of each constituent component within a frenched rack) for frenched racks. This method allowed the identification of the main sources of variation for nutritional characteristics. These differences were found to be caused by the lean to fat ratio, not nutritional differences in lean tissue from the same region of lamb (i.e. loin eye and tenderloin very similar nutritionally). The difference in lean to fat ration also accounts for the variation between grades (i.e. PX grade lamb cuts have a higher fat content than YX grade lamb cuts due to PX grade cuts having a higher percentage fat tissue in their cuts). The cuts characterised were the shortloin section (whole section or chop), rack section (whole section or chop), 75mm racks frenched 25mm, boneless loin and tenderloin for both PX and YX grade lamb. The method will be applicable to other regions of lamb (i.e. hindquarter and forequarter) for which nutritional information already exists, but for which yielding data will have to be collected. The method would also be applicable to other species such as beef and venison, but both nutritional data for constituent components and yielding data would have to be collected.

Richmond Ltd. has recently undergone a change in strategy, away from the traditional commodity based meat industry, towards the modern food business. To do this, opportunities to add value to their current product range must be identified. This involves the conversion of traditionally low value commodity based products into products that demand a premium. An example of this is converting muscles that are currently used for grinding meat into a further processed convenience food (i.e. ready meals). Another method is to add further value to premium products by making them more appealing to consumers (i.e. nutritional information on labels). This work details investigations into the functional properties of selected beef forequarter muscles (low value commodity products) and the nutritional properties of selected export lamb products (premium products). The functional properties of a number of beef forequarter muscles were measured to identify which had the best potential for further processing applications with respect to ready meals. The functional properties of tenderness, cook loss and shrinkage were measured for the Latissimus Dorsi, Pectorialis Profundus (Point End Brisket), Infraspinatus (Cross Cut Blade), Triceps Brachi Longhead (Main muscle in Bolar Shoulder Clod), Supraspinatus (Chuck Tender), Serratus Ventralis and Triceps Brachi Medialhead (Muscle in Bolar Shoulder Clod. From the tests conducted the Infraspinatus and the Triceps Brachi Longhead have been identified as having the best functional properties with respect to further processing for ready meal applications. As well as conducting tests to identify the forequarter muscles with the best potential for further processing applications, investigations were carried out to identify cooking regimes that would optimise the functional properties. This work confirmed that there are three major chemical reactions, which determine the resultant functional properties of cooked meat. They are the denaturation and aggregation of the myofibrillar proteins and the denaturation and solubilisation of connective tissue (collagen). At around 50°C myosin (45% to 50% of the myofibrillar proteins) denatures, which results in a substantial increase in cook loss and reduction in water holding capacity. At around 60°C collagen (main connective tissue protein) denatures, which results in a substantial increase in tenderness and increase in cook loss. This is because as the collagen denatures it loses it mechanical strength (increase in tenderness) and can no longer support its own structure, and causes it to contract. This contraction causes fluid within the meat and cook loss caused by the denaturation of myosin to be expelled from the meat by compressive forces (squeezed out). At around 70°C actomyosin (22% of the myofibrillar proteins) denatures. This results in a substantial increase in the cook loss and firming of the meat. The increase in cook loss or decrease in water holding capacity that occurs with myofibrillar protein denaturation is due to the fact that when these proteins denature and aggregate their ability to bind water is greatly reduced. From the results of the cooking regime trials it is recommended that for functional property considerations that during the cooking of further processed meat products (i.e. ready meal applications) a meat temperature of 62°C should be aimed for, for the slowest heating region during cooking (usually the centre). This is because it has been identified that a cooking temperature of 65°C should not be exceeded otherwise detrimental effects can occur to the functional properties of the cooked meat. For health concerns a 7D bacterial death reduction has to be achieved. This means that for a cooking temperature of 62°C the meat has to be held at this temperature for at least 5 minutes. Therefore the total cooking time would be the time needed to heat all the meat to 62°C plus 5 minutes to ensure a safe product. The heating or cooking system employed should also ensure that a minimal amount of the meat is heated above 65°C. This can be easily achieved by minimising the external cooking temperature, but long cooking times will result. An industrial cooking process will be a compromise between the cost associated with longer residence time and product functionality. As mentioned earlier another way to add value is to supply nutritional information for selected cuts. Consequentially one of the objectives of this project was to provide some nutritional information for selected meat cuts. Though the primary objective of this part of the project was to develop a method for producing the needed information, so that Richmond N.Z. Ltd. can develop further information on an as needs basis. The nutritional characteristics of a number of export lamb cuts from the saddle region has also been investigated and a method devised to allow further characterisation of other cuts. The method involves breaking down a standard cut into its constituent components (e.g. Frenched rack consists of loin eye, fat cap, intercostals and fatty tissue). The constituent components are tested for their nutritional properties. The frenched rack nutritional properties are calculated from the nutritional properties of the constituents components and the yield data (percentage of each constituent component within a frenched rack) for frenched racks. This method allowed the identification of the main sources of variation for nutritional characteristics. These differences were found to be caused by the lean to fat ratio, not nutritional differences in lean tissue from the same region of lamb (i.e. loin eye and tenderloin very similar nutritionally). The difference in lean to fat ration also accounts for the variation between grades (i.e. PX grade lamb cuts have a higher fat content than YX grade lamb cuts due to PX grade cuts having a higher percentage fat tissue in their cuts). The cuts characterised were the shortloin section (whole section or chop), rack section (whole section or chop), 75mm racks frenched 25mm, boneless loin and tenderloin for both PX and YX grade lamb. The method will be applicable to other regions of lamb (i.e. hindquarter and forequarter) for which nutritional information already exists, but for which yielding data will have to be collected. The method would also be applicable to other species such as beef and venison, but both nutritional data for constituent components and yielding data would have to be collected.

The objective of this study was to investigate the effects of preheat treatments on fouling by fresh whole milk (FWM), recombined whole milk (RCB) and reconstituted whole milk (Recon) in the high-temperature heater of indirect UHT plants. Various preheat treatments prior to evaporation during milk powder manufacture were applied to skim milk powder (SMP, 75 °C 2 s, 85 °C, 155 s and 95 °C, 155 s) and whole milk powder (WMP, 95 °C, 33 s). These preheat treatments were so-called “evaporator preheat treatments”. Skim milk powder (SMP) and whole milk powder (WMP) were derived from the same original batch of pasteurised FWM to remove the effects of the variation in milk composition between different milk batches. These SMPs were recombined with anhydrous milk fat and water to prepare RCB, and WMPs were reconstituted with water to prepare Recon. Then, (homogenized) FWM, RCB and Recon were subjected to various preheat treatments (75 °C, 11 s, 85 °C, 147 s and 95 °C, 147 s) prior to UHT processing. These preheat treatments were so-called “UHT preheat treatments”. Temperature difference (hot water inlet temperature – milk outlet temperature) was taken as a measure of the extent of fouling in the high-temperature heater. The slope of the linear regression of temperature difference versus time (for two hours of UHT processing) was taken as fouling rate (°C/h). Increasing both evaporator and UHT preheat treatments resulted in increasing fouling rate and total deposit weight for all three whole milk types for several milk batches. In the case of FWM, there was no reduction in fouling rate with increasing UHT preheat treatment whether FWM was homogenized then preheated, preheated then homogenized or not homogenized at all. These findings, which are wholly consistent and well replicated, are in apparent conflict with the results of most previous comparable studies. Possible reasons for this are explained. Further investigations of the effects of homogenization relating to the role of whey protein on the surface of the fat globules showed that whey protein associated with the membrane covering the surface of fat globules for homogenized then preheated FWM, RCB and Recon and that association increased with increasing heating process stage. The increasing association of whey protein with the milk fat globules membrane with increasing severity of heating process stage became faster when preheat treatment was more severe: the association of whey protein plateaued on intermediate temperature heating when the milks were preheated at 75°C, 11 s and on preheating when the milks were preheated at 95°C, 147 s. In the case of FWM, the thickness of the membrane covering the surface of fat globules for homogenized then preheated FWM, which increased with the severity of heating process stage, was greater than the thickness of the membrane in preheated then homogenized FWM. Preheating then homogenization resulted in the greater interfacial spreading of small molecules on the surface of fat globules, i.e. whey protein or small molecules from the disintegration of casein micelles during preheating. Possible basic mechanisms for UHT fouling in the high-temperature heater include: the reduction in the solubility of calcium phosphate and the deposition of protein as fat-bound protein and non-fat-bound protein. When non-fat-bound protein in milk plasma deposited, it could be a carrier for the deposition of mineral, such as, the precipitate of calcium phosphate in the casein micelles or the deposition of complexes between whey protein and casein micelles.

Richmond Ltd. has recently undergone a change in strategy, away from the traditional commodity based meat industry, towards the modern food business. To do this, opportunities to add value to their current product range must be identified. This involves the conversion of traditionally low value commodity based products into products that demand a premium. An example of this is converting muscles that are currently used for grinding meat into a further processed convenience food (i.e. ready meals). Another method is to add further value to premium products by making them more appealing to consumers (i.e. nutritional information on labels). This work details investigations into the functional properties of selected beef forequarter muscles (low value commodity products) and the nutritional properties of selected export lamb products (premium products). The functional properties of a number of beef forequarter muscles were measured to identify which had the best potential for further processing applications with respect to ready meals. The functional properties of tenderness, cook loss and shrinkage were measured for the Latissimus Dorsi, Pectorialis Profundus (Point End Brisket), Infraspinatus (Cross Cut Blade), Triceps Brachi Longhead (Main muscle in Bolar Shoulder Clod), Supraspinatus (Chuck Tender), Serratus Ventralis and Triceps Brachi Medialhead (Muscle in Bolar Shoulder Clod. From the tests conducted the Infraspinatus and the Triceps Brachi Longhead have been identified as having the best functional properties with respect to further processing for ready meal applications. As well as conducting tests to identify the forequarter muscles with the best potential for further processing applications, investigations were carried out to identify cooking regimes that would optimise the functional properties. This work confirmed that there are three major chemical reactions, which determine the resultant functional properties of cooked meat. They are the denaturation and aggregation of the myofibrillar proteins and the denaturation and solubilisation of connective tissue (collagen). At around 50°C myosin (45% to 50% of the myofibrillar proteins) denatures, which results in a substantial increase in cook loss and reduction in water holding capacity. At around 60°C collagen (main connective tissue protein) denatures, which results in a substantial increase in tenderness and increase in cook loss. This is because as the collagen denatures it loses it mechanical strength (increase in tenderness) and can no longer support its own structure, and causes it to contract. This contraction causes fluid within the meat and cook loss caused by the denaturation of myosin to be expelled from the meat by compressive forces (squeezed out). At around 70°C actomyosin (22% of the myofibrillar proteins) denatures. This results in a substantial increase in the cook loss and firming of the meat. The increase in cook loss or decrease in water holding capacity that occurs with myofibrillar protein denaturation is due to the fact that when these proteins denature and aggregate their ability to bind water is greatly reduced. From the results of the cooking regime trials it is recommended that for functional property considerations that during the cooking of further processed meat products (i.e. ready meal applications) a meat temperature of 62°C should be aimed for, for the slowest heating region during cooking (usually the centre). This is because it has been identified that a cooking temperature of 65°C should not be exceeded otherwise detrimental effects can occur to the functional properties of the cooked meat. For health concerns a 7D bacterial death reduction has to be achieved. This means that for a cooking temperature of 62°C the meat has to be held at this temperature for at least 5 minutes. Therefore the total cooking time would be the time needed to heat all the meat to 62°C plus 5 minutes to ensure a safe product. The heating or cooking system employed should also ensure that a minimal amount of the meat is heated above 65°C. This can be easily achieved by minimising the external cooking temperature, but long cooking times will result. An industrial cooking process will be a compromise between the cost associated with longer residence time and product functionality. As mentioned earlier another way to add value is to supply nutritional information for selected cuts. Consequentially one of the objectives of this project was to provide some nutritional information for selected meat cuts. Though the primary objective of this part of the project was to develop a method for producing the needed information, so that Richmond N.Z. Ltd. can develop further information on an as needs basis. The nutritional characteristics of a number of export lamb cuts from the saddle region has also been investigated and a method devised to allow further characterisation of other cuts. The method involves breaking down a standard cut into its constituent components (e.g. Frenched rack consists of loin eye, fat cap, intercostals and fatty tissue). The constituent components are tested for their nutritional properties. The frenched rack nutritional properties are calculated from the nutritional properties of the constituents components and the yield data (percentage of each constituent component within a frenched rack) for frenched racks. This method allowed the identification of the main sources of variation for nutritional characteristics. These differences were found to be caused by the lean to fat ratio, not nutritional differences in lean tissue from the same region of lamb (i.e. loin eye and tenderloin very similar nutritionally). The difference in lean to fat ration also accounts for the variation between grades (i.e. PX grade lamb cuts have a higher fat content than YX grade lamb cuts due to PX grade cuts having a higher percentage fat tissue in their cuts). The cuts characterised were the shortloin section (whole section or chop), rack section (whole section or chop), 75mm racks frenched 25mm, boneless loin and tenderloin for both PX and YX grade lamb. The method will be applicable to other regions of lamb (i.e. hindquarter and forequarter) for which nutritional information already exists, but for which yielding data will have to be collected. The method would also be applicable to other species such as beef and venison, but both nutritional data for constituent components and yielding data would have to be collected.

The increasing price of oil and the destabilisation of the world’s climate are urging governments, businesses and individuals to constantly investigate energy-efficient technologies and methodologies and pursue the adoption of energy efficiency programmes in a global effort to reduce energy consumption, greenhouse gas emissions and ultimately energy costs. In New Zealand, one of the biggest industrial energy efficiency projects was started in 2002 by a multinational dairy company, the Fonterra Co-operative Group, in partnership with the energy service company Demand Response Ltd; the project currently aims at reducing by 15% the energy costs at all Fonterra’s major production sites throughout the country. This thesis, undertaken as part of the above project, examines the development and implementation of a structured and integrated energy monitoring and targeting methodology (M&T) for the most efficient operation of all Fonterra’s chilled water systems, with an initial focus on the ones installed at Clandeboye, one of the Fonterra’s sites involved in the energy saving project. A data collection system (Insite) was already in place at Clandeboye to enable storage and analysis of some of the site’s utility metering data. After identification of key chilled water system components and definition of data requirements for M&T purposes, an analysis of past energy consumption trends (based on multiple regression calculations) was carried out to develop an historical benchmark of the energy used, compare it with current energy performance and thus identify opportunities for future improvements. The creation of an M&T reporting system for presenting findings to operators and management was the last essential part of the thesis development. The study has highlighted that the robustness of the proposed regression model was badly affected by the unreliability of the existing data collection system and the uncertainty associated with poorly documented changes to operating conditions/plant configuration that had occurred over time. The conclusion is that, while the developed M&T methodology is theoretically valid and readily applicable, further developments are necessary (and recommended) to make it suitable for other similar systems.

Richmond Ltd. has recently undergone a change in strategy, away from the traditional commodity based meat industry, towards the modern food business. To do this, opportunities to add value to their current product range must be identified. This involves the conversion of traditionally low value commodity based products into products that demand a premium. An example of this is converting muscles that are currently used for grinding meat into a further processed convenience food (i.e. ready meals). Another method is to add further value to premium products by making them more appealing to consumers (i.e. nutritional information on labels). This work details investigations into the functional properties of selected beef forequarter muscles (low value commodity products) and the nutritional properties of selected export lamb products (premium products). The functional properties of a number of beef forequarter muscles were measured to identify which had the best potential for further processing applications with respect to ready meals. The functional properties of tenderness, cook loss and shrinkage were measured for the Latissimus Dorsi, Pectorialis Profundus (Point End Brisket), Infraspinatus (Cross Cut Blade), Triceps Brachi Longhead (Main muscle in Bolar Shoulder Clod), Supraspinatus (Chuck Tender), Serratus Ventralis and Triceps Brachi Medialhead (Muscle in Bolar Shoulder Clod. From the tests conducted the Infraspinatus and the Triceps Brachi Longhead have been identified as having the best functional properties with respect to further processing for ready meal applications. As well as conducting tests to identify the forequarter muscles with the best potential for further processing applications, investigations were carried out to identify cooking regimes that would optimise the functional properties. This work confirmed that there are three major chemical reactions, which determine the resultant functional properties of cooked meat. They are the denaturation and aggregation of the myofibrillar proteins and the denaturation and solubilisation of connective tissue (collagen). At around 50°C myosin (45% to 50% of the myofibrillar proteins) denatures, which results in a substantial increase in cook loss and reduction in water holding capacity. At around 60°C collagen (main connective tissue protein) denatures, which results in a substantial increase in tenderness and increase in cook loss. This is because as the collagen denatures it loses it mechanical strength (increase in tenderness) and can no longer support its own structure, and causes it to contract. This contraction causes fluid within the meat and cook loss caused by the denaturation of myosin to be expelled from the meat by compressive forces (squeezed out). At around 70°C actomyosin (22% of the myofibrillar proteins) denatures. This results in a substantial increase in the cook loss and firming of the meat. The increase in cook loss or decrease in water holding capacity that occurs with myofibrillar protein denaturation is due to the fact that when these proteins denature and aggregate their ability to bind water is greatly reduced. From the results of the cooking regime trials it is recommended that for functional property considerations that during the cooking of further processed meat products (i.e. ready meal applications) a meat temperature of 62°C should be aimed for, for the slowest heating region during cooking (usually the centre). This is because it has been identified that a cooking temperature of 65°C should not be exceeded otherwise detrimental effects can occur to the functional properties of the cooked meat. For health concerns a 7D bacterial death reduction has to be achieved. This means that for a cooking temperature of 62°C the meat has to be held at this temperature for at least 5 minutes. Therefore the total cooking time would be the time needed to heat all the meat to 62°C plus 5 minutes to ensure a safe product. The heating or cooking system employed should also ensure that a minimal amount of the meat is heated above 65°C. This can be easily achieved by minimising the external cooking temperature, but long cooking times will result. An industrial cooking process will be a compromise between the cost associated with longer residence time and product functionality. As mentioned earlier another way to add value is to supply nutritional information for selected cuts. Consequentially one of the objectives of this project was to provide some nutritional information for selected meat cuts. Though the primary objective of this part of the project was to develop a method for producing the needed information, so that Richmond N.Z. Ltd. can develop further information on an as needs basis. The nutritional characteristics of a number of export lamb cuts from the saddle region has also been investigated and a method devised to allow further characterisation of other cuts. The method involves breaking down a standard cut into its constituent components (e.g. Frenched rack consists of loin eye, fat cap, intercostals and fatty tissue). The constituent components are tested for their nutritional properties. The frenched rack nutritional properties are calculated from the nutritional properties of the constituents components and the yield data (percentage of each constituent component within a frenched rack) for frenched racks. This method allowed the identification of the main sources of variation for nutritional characteristics. These differences were found to be caused by the lean to fat ratio, not nutritional differences in lean tissue from the same region of lamb (i.e. loin eye and tenderloin very similar nutritionally). The difference in lean to fat ration also accounts for the variation between grades (i.e. PX grade lamb cuts have a higher fat content than YX grade lamb cuts due to PX grade cuts having a higher percentage fat tissue in their cuts). The cuts characterised were the shortloin section (whole section or chop), rack section (whole section or chop), 75mm racks frenched 25mm, boneless loin and tenderloin for both PX and YX grade lamb. The method will be applicable to other regions of lamb (i.e. hindquarter and forequarter) for which nutritional information already exists, but for which yielding data will have to be collected. The method would also be applicable to other species such as beef and venison, but both nutritional data for constituent components and yielding data would have to be collected.

Agglomerated milk powders are produced to give improved properties such as flowability, dispersibility, reduced dustiness and decreased bulk density. A key function of these powders is to dissolve "instantly" upon addition to water and because of this they are also called "instant milk powders". They are produced by agglomerating the undersized fines that are returned to the top of the spray drier with milk concentrate droplet spray. Interaction occurs in a collision zone, often with multiple sprays and fines return lines. Agglomeration can be a difficult process to control and operators find it hard to fine tune the process to produce specific powder properties. This work aimed to understand the effects of key droplet and fines properties on the extent of agglomeration to allow a mechanistic understanding of the process. Three scales of spray drier were investigated in this study with different rates of evaporation; a small scale drier (0.5 - 7 kg water h-1), a pilot scale drier (80 kg water h-1) and a range of commercial production scale driers (4 - 15 000 kg water h-1). A survey of operators of commercial scale driers showed that control of instant milk powder production to influence bulk density is highly intuitive. Fines recycle rates were expected to be important in control of agglomeration processes and were estimated on a specific plant by using the pressure drop measured in the fines return line. A model based on pressure drop along a pneumatic pipeline under-predicted the experimental values for pressure drop due to solids, which means a calibration curve should be generated for each specific drier. Fines recycle rates were predicted to be significantly higher at 95 to 130 % of production rates compared to those expected by operators of 50%. Experimental measurements agreed with existing models for the effect of temperature on the density and viscosity of milk concentrates. Experimental results showed that the surface tensions of concentrated milks were within the same range as literature values for standard milks below 60°C, but were significantly higher for milk above 60°C. This is thought to be linked to the mechanism of skin formation due to disulphide cross linking at high temperatures and concentrations. Powder properties were also established for selected products produced on the commercial scale driers. These powders were then used in experiments on the two smaller driers. Because collision frequency depends on the velocity and droplet size of sprays; these properties were measured for the small scale drier and estimated, where possible, for the pilot and commercial driers. The small scale agglomerating spray drier was configured to alter droplet and particle properties when interacting a vertical fines particle curtain with a horizontal spray sheet. An extensive design and improvement process was carried out to ensure the system consistently delivered these streams in a controllable manner. The processes of collision and adhesion occur very quickly inside the spray drier. In order to assess the extent of agglomeration that has occurred, the feed streams must be compared to the final product stream. An ideal way to do this is to use an agglomeration index which compares the particle size distributions of the feed (fines recycle and spray streams) and the particle size distribution of the product stream (the agglomerated powder). The index described changes between these steams across the particle size distribution and is called an agglomeration efficiency, ξg. However, it was found that the presence of fines in the product of the one-pass design obscured the agglomerates formed. The agglomeration efficiency, ξg, was modified to become ξh which subtracted the fines stream from the agglomerated product distribution. In this way ξh models industrial operation where the fines are recycled, by effectively just comparing the spray and product streams entering and leaving the process. The small scale drier was used for an experimental study on natural and forced agglomeration, where the drier was operated with spray only, then with spray and fines. For natural agglomeration, SEM images of the product powder indicated that little agglomeration occurred between spray droplets. The product yield was unacceptably low (~ 40%) due to adhesion of spray droplets to the drying chamber wall opposing the horizontal spray. When the fines curtain was introduced in the forced agglomeration experiments, product yield increased above 50% because the fines acted as collectors for the spray droplets. However, the agglomeration performance of the modified spray drier was lower than expected. The equipment design was then optimised by considering three key issues; fines dispersion, droplet dispersion and stickiness, and agglomerate breakdown. Final experiments studied agglomeration at low fines to spray mass flux ratios and showed that increasing the fines size had a positive effect on agglomeration efficiency,ξh. The agglomeration study at pilot scale identified the effect of key variables, total solids, concentrate and fines flow rate, and fines size on the agglomeration efficiency. A dimensionless flux approach was used to explain the experimental results. The fines to spray mass flux ratio and the projected area flux ratio (at constant concentrate flow rate) were found to be the most suitable to represent the physical processes during agglomeration. Experimental results showed that a higher dimensionless flux resulted in more agglomeration and as well as small fines size and atomising low solids concentrate. The critical Stokes number highlighted the importance of particle size and collision velocity on the outcome of the collision as well as the importance of stickiness on adherence following the collision. A statistical analysis established a relational model for predicting the agglomeration efficiency based on fines size, total solids and the fines to spray mass flux ratio. This thesis has gained insight into agglomeration processes during spray drying and knowledge about how to define the extent of agglomeration. Practical findings from this research can have a significant impact on successful spray drying operation for instant powders. There are some practical steps to be taken industrially to promote the control of agglomerating spray driers. The first step is to measure and control the flow of fines recycled to the top of the spray drier. The next step is to validate the findings at industrial scale and link the agglomeration index to the bulk powder properties. However, there are many challenges that remain to be tackled in the area of milk powder agglomeration. Milk powder agglomeration at the top of the spray drier is a complex process involving many different variables. A more detailed study of the micro processes that occur during agglomeration will give increased understanding of the relationships between key operating variables and agglomerate properties.

<div>The overall goal of this thesis was to study the performance of two related portable multipurpose solar dryers, DehytrayTM and DehymeleonTM, in comparison to open-air sun drying by drying tomatoes, apples and mint under West Lafayette, Indiana weather conditions. Thin layer drying tests were conducted on tomato slices, apples slices and mint leaves, with three temperatures [24°C (75°F), 35°C (95°F) and 54 °C (130°F)], and an airflow velocity of 1 m/s to determine the drying kinetics of these products during diurnal drying cycles typical for solar and/or open-air sun drying. Subsequently, field drying tests were conducted for tomatoes slices, apples slices and mint leaves with the two solar drying technologies (DehymeleonTM and DehytrayTM) and open-air sun drying using uncovered Dehytrays as the control. The average temperatures achieved for these technologies were 45°C (113°F), 60°C (140 °F) and 27°C (80.6 °F) for the DehymeleonTM, DehytrayTM and open-air sun drying, respectively. Moisture diffusivity were in the order of 10-4 to 10-9 (m2/s) for the different methods, depending directly on the product, temperatures and air flow inside the drying chamber.</div><div><br></div><div>Quality attributes (color, vitamin C and microbial growth) were measured before and after the field drying tests. Color difference (ΔE) for DehymeleonTM solar dryer showed the least variation compared with the fresh products. However, for the DehytrayTM ΔE increased due to the impact of its higher temperature and direct sunlight exposure that led to Maillard reactions and caramelization in the case of tomatoes and apples slices. Additionally, vitamin C (Ascorbic acid) content for tomatoes and apples slices was affected for the high ranges of temperatures reached inside the Dehytray™. Denaturing of vitamin C was less observed for DehymeleonTM, maintaining values of 166 mg/100 g dm for tomatoes, and 104.2mg/100g for apples slices. There was no significant difference (α = 0.05) in the microbial growth for the DehytrayTM and open-air drying compared to the fresh product, however, there was significant difference for the DehymeleonTM when drying tomatoes and apples slices, without up one log reduction on the original microbial population. In the case of mint, DehymeleonTM had a 2.3 log reduction, which is similar to L-lactic acid sanitizer achieved by another study in the literature, compared with 0.4 log obtained by the DehytrayTM and 0.47 log obtained by open-air sun drying. The differences in microbial growth were observed because the temperatures inside the drying chamber of the DehymeleonTM was low and product moisture content was above the safe equilibrium moisture content (EMC) for both tomatoes and apples during the early critical hours at the onset of the drying process, which was favorable to mold growth. The lack of a fan to intermittently or constantly flush out humid air released from the crop dried in the DehytrayTM negatively affected its performance. The insufficient airflow in the drying chamber of the DehymeleonTM and its inability to achieve the high temperatures observed in the DehytrayTM negatively affected its performance. Both solar dryers, DehymeleonTM and DehytrayTM achieved high hygienic condition during drying due to their enclosed chambers than protected the crop from contaminant in the environments. Their portability and design for large-scale manufacturing and deployment are a positive development that would be helpful to small and mid-size growers, as well as households (home gardens). Areas for further research were highlighted.</div><div><br></div><div><br></div>