Doctoral Degrees (Viticulture and Oenology)

Permanent URI for this collection

https://scholar.sun.ac.za/handle/10019.1/575

Browse

Browsing Doctoral Degrees (Viticulture and Oenology) by Subject "Alcoholic beverages -- Flavor and odor -- South Africa"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    The influence of base wine composition and wood maturation on the quality of South African brandy
    (Stellenbosch : Stellenbosch University, 2004-03) Snyman, Caroline Lucie Charlotte; Lambrechts, M. G.; Du Toit, M.; Stellenbosch University. Faculty of AgriSciences. Dept. of Viticulture and Oenology. Institute for Wine Biotechnology.
    ENGLISH ABSTRACT: Brandy production is a multi-step process that involves grape harvesting, base wine fermentation, distillation, wood maturation and blending. Within each of these production process steps there are a number of factors that can influence the composition and resultant quality of the base wine, unaged and wood matured distillates. These factors include geographic and climatic features of the origin of grapes used, viticultural practices, grape maturity, grape variety, vintage variation, vinification techniques, storage of the base wine prior to distillation, distillation technique, age and origin of oak wood used for maturation and barrel toasting levels. The composition of flavour is extremely complex in wine and distilled beverages such as brandy. A multitude of compounds can take part in the formation of flavour, and it is rare that a particular compound, that is solely responsible for nuances of a specific flavour, is identified. Thus, taking brandy production factors and the nature of aroma and flavour into account, it is clear that thorough understanding of the complexities affecting the perception of quality in brandy is difficult. In commercial brandy production, standard operating procedures do not allow for the separate distillation of brandy base wines from different producers and mixing of base wines from different regions and producers thus occurs. This makes it difficult to determine whether the quality determination of the base wine is in fact an indication of the quality of the resultant distillate after the required period of wood maturation. Therefore, the aim of this study is to determine whether there is any merit in storing and distilling brandy base wines from different producers and regions separately. This was done by firstly determining the demographic and production factors that influence brandy base wine composition. Secondly, the influence of brandy base wine composition on the quality of the respective unaged potstill distillates as well as the ultimate style and quality of the threeyear old wood matured potstill distillates was determined. For this purpose, four potstills with a capacity of 2000 L each were isolated and used for commercial scale distillations of 33 and 25 brandy base wines in 1999 and 2000, respectively, at the DisteIl distillery in Worcester. The experimental outlay used in the study closely emulated Distell's standard operating procedures for commercial brandy production. Chenin blanc and Colombar are the two most popular grape varietals used in the making of brandy base wine in South Africa. Due to the complexity of the brandy production process, the first part of the study only focussed on South African young Chenin blanc wines, with a predominantly fermentation derived aroma. The concentration of iso-amyl acetate, hexyl acetate, ethyl caprylate, ethyl caprate, 2-phenethyl acetate and octanoic acid was significantly higher in wines awarded gold and silver medals and decreased significantly with subsequent decreases in quality categories. Ethyl lactate exhibited the opposite pattern. A quality predictor model based on the behaviour of 21 volatile compounds quantified in these wines had a resonable prediction accuracy when having to predict the quality of wines made in the same vintages as it had been trained on. It was significantly poorer when tested on a completely different vintage of wines to the ones it had been trained on. The number of vintages, number of wines per vintage and the number of compounds quantified for the model influence its prediction accuracy. Twenty-seven volatile compounds were quantified in 33 and 25 brandy base wines, their unaged and three year old distillates from 1999 and 2000, respectively. ANOVA and CART analysis showed that vintage, region, harvest time, choice of cultivar and yeast strain can have a significant influence on the volatile compound composition of brandy base wines, their unaged and three year old distillates. These factors as well as the volatile compound composition were also found to influence the sensory quality of these products. Base wines, unaged and three year old distillates originating from the De Dooms region, which predominantly cultivates table grapes, were of significantly lower quality than those from the remaining regions. Products made from grapes harvested early in the season were of significantly higher quality. The volatile aroma compound composition was found to differ significantly between the 1999 and 2000 base wines and distillates, irrespective of the exclusion of those samples that had undergone partial or complete malolactic fermentation. Consequently, quality indicating compounds may vary from vintage to vintage. The relationship between the quality of brandy base wines and the concentration of n-butanol, iso-amyl acetate, ethyl lactate, ethyl caprylate, octanoic- and decanoic acid was the same as that reported in young Chen in blanc wines in this study. In unaged distillates, increased levels of ethyl lactate also exert a negative influence on distillate quality. Iso-amyl acetate, hexyl acetate, ethyl caproate, ethyl caprylate, n-butanol, octanoic acid, ethyl caprate and decanoic acid showed some positive correlation, whilst iso-butanol, ethyl lactate, acetic acid, acetaldehyde and ethyl acetate showed a significant negative correlation to three-year old distillate quality. Sensory descriptive analysis on selected good, average and poor quality distillates using the South African brandy aroma wheel showed that there are small differences in profile between the good and average distillates, there were however significant differences between the good and poor quality distillate profiles throughout maturation. After three years of wood maturation, the aroma profile of poor quality distillates can be characterised by prominent herbaceous and woody aromas, which are more intense than the fruity aromas. Good quality distillates contained characteristically intense fruity aromas. Volatile compound concentration differences were noted during the course of and after three years of wood maturation and in barrels of varying ages. Distillates matured in new block barrels exhibited significant differences in volatile and wood compound composition after three years when compared to remaining barrels used. The style classification of the three-year old potstill distillate was influenced by demographic and production factors and volatile compound composition, but not by the sensory quality of the distillates. In summary, vintage, region, cultivar, harvest time and choice of yeast strain have a significant influence on the volatile composition of brandy base wines, their unaged and three year old potstill distillates, which in turn affects the sensory quality of these products. These effects cannot be viewed in isolation as they jointly exert an influence on the composition and quality of these products. From a commercial perspective, this study has provided a valuable indication as to which production and demographic factors can influence the quality and style of potstill brandy. Thus, future brandy base wine intake should, as far as possible, take place in such a manner to allow base wines originating from the same cultivar or region or harvest time or combination thereof (and to a lesser extent yeast strain) to be received simultaneously at the distillery for distillation.