Thesis (MSc (Wine Biotechnology))--University of Stellenbosch, 2009. / ENGLISH ABSTRACT: Fermentation is a complex process in which raw materials are transformed into high-value
products, in this case, grape juice into wine. In this modern and economically competitive
society, it is increasingly important to consistently produce wine to definable specifications and
styles. Process management throughout the production stage is therefore crucial to achieve
effective control over the process and consistent wine quality. Problematic wine fermentations
directly impact on cellar productivity and the quality of wine. Anticipating stuck or sluggish
fermentations, or simply being able to foresee the progress of a given fermentation, would be
extremely useful for an enologist or winemaker, who could then take suitable corrective steps
where necessary, and ensure that vinifications conclude successfully. Conventional methods of
fermentation monitoring are time consuming, sometimes unreliable, and the information limited
to a few parameters only. The current effectiveness of fermentation monitoring in industrial wine
production can be much improved. Winemakers currently lack the tools to identify early signs of
undesirable fermentation behaviour and to take preventive actions.
This study investigated the application of Fourier transform mid infrared (FT-IR)
spectroscopy in transmission mode, for the quantitative and qualitative monitoring of alcoholic
fermentation during industrial wine production. The major research objectives were firstly to
establish a portfolio of quantitative calibration models suitable for quantification of the major
quality determining parameters in fermenting must. The second major research objective
focused on a pilot study aimed at exploring the use of off-line batch multivariate statistical
process control (MSPC) charts for actively fermenting must. This approach used FT-IR spectra
only, for the purpose of qualitative monitoring of alcoholic fermentation in industrial wine
production. Towards these objectives, a total of 284 industrial-scale, individual, actively
fermenting tanks of the seven major white cultivars and blends, and nine major red cultivars, of
Namaqua Wines, Vredendal, South Africa, were sampled and analysed with FT-IR
spectroscopy and appropriate reference methods during vintages 2007 to 2009.
For the quantitative strategy, partial least squares regression (PLS1) calibration models for
determination of the classic wine parameters ethanol, pH, volatile acidity (VA), titratable acidity
(TA) and the total content of glucose plus fructose, were redeveloped to provide a better fit to
local South African samples. New PLS1 models were developed for the must components
glucose, fructose and yeast assimilable nitrogen (YAN), all of which are frequently implicated in
problem fermentations. The regression statistics, that included the standard error of prediction
(SEP), coefficient of determination (R2) and bias, were used to evaluate the performance of the
redeveloped calibration models on local South African samples. Ethanol (SEP = 0.15 %v/v, R2 =
0.999, bias = 0.04 %v/v) showed very good prediction and with a residual predictive deviation
(RPD) of 30, rendered an excellent model for quantitative purposes in fermenting must. The
models for pH (SEP = 0.04, R2 = 0.923, bias = -0.01) and VA (SEP = 0.07 g/L, R2 = 0.894, bias
= -0.01 g/L) with RPD values of 4 and 3 respectively, showed that the models were suitable for
screening purposes. The calibration model for TA (SEP = 0.35 g/L, R2 = 0.797, bias = -0.004
g/L) with a RPD of 2, proved unsatisfactory for quantification purposes, but reasonable for
screening purposes. The calibration model for the total content of glucose plus fructose (SEP =
0.6.19 g/L, R2 = 0.993, bias = 0.02 g/L) with a RPD of 13, showed very good prediction and can
be used to quantify total glucose plus fructose content in fermenting must. The newly developed
calibration models for glucose (SEP = 4.88 g/L, R2 = 0.985, bias = -0.31 g/L) and fructose (SEP
= 4.14 g/L, R2 = 0.989, bias = 0.64 g/L) with RPD values of 8 and 10 respectively, also proved fit
for quantification of these important parameters. The new calibration models of ethanol, total
glucose plus fructose; and glucose and fructose individually, showed an excellent relation to
local South African samples and can be easily implemented by the wider wine industry.
Two calibration models were developed to determine YAN in fermenting must by using
different reference methods, namely the enzyme-linked spectrophotometric assay and Formol
titration method, respectively. The results showed that enzyme-linked assays provided a good
quantitative model for white fermenting must (SEP = 14.10 mg/L, R2 = 0.909, bias = -2.55 mg/L,
RPD = 6), but the regression statistics for predicting YAN in red fermenting must, were less
satisfactory (data not shown). The Formol titration method could be used successfully in both
red- and white fermenting must (SEP = 16.37 mg/L, R2 = 0.912, bias = -1.01 mg/L, RPD = 4). A
minor, but very important finding was made with respect to the storage of must samples that
were taken from tanks, but that could not immediately be analysed with FT-IR spectroscopy or
reference values. Principal component analysis (PCA) of frozen samples showed that must
samples could be stored frozen for up to 3 months and still be used to expand the calibration
sample sets when needed. Therefore, samples can be kept frozen to a later stage if immediate
analyses are not possible.
For the purpose of the pilot study that focused on the use of FT-IR spectroscopy for
qualitative off-line monitoring of alcoholic fermentation, a total of 21 industrial-scale fermentation
tanks were monitored at 8- or 12-hourly intervals, from the onset of fermentation to complete
consumption of the grape sugars. This part of the work excluded quantitative data, and only
used FT-IR spectra. MSPC charts were constructed on the PLS scores of all the FT-IR spectra
taken at the various time intervals of the different batches, using time as the y-variable. The
primary aim of this research objective was to evaluate if the PLS batch models could be used to
discriminate between normal and problem alcoholic fermentations. The models that were
constructed clearly showed the variations in patterns over time, between red- and white wine
alcoholic fermentations. One Colombar tank that was fermented at very low temperature in
order to achieve a specific wine style, was characterised by a fermentation pattern that clearly
differed form the rest of the Colombar fermentations. This atypical fermentation was identified
by the batch models constructed in this study. PLS batch models over all the Colombar
fermentations clearly identified the normal and problem fermentations.
The results obtained in this study showed that FT-IR spectroscopy showed great potential
for effective quantitative and qualitative monitoring of alcoholic fermentation during industrial
wine production. The work done in this project resulted in the development of a portfolio of
calibration models for the most important quality determining parameters in fermenting must.
The quantitative models were subjected to extensive independent test set validation, and have
subsequently been implemented for industrial use at Namaqua Wines. Multivariate batch
monitoring models were established that show good discriminatory power to detect problem
fermentations. This is a very useful diagnostic tool that can be further developed by monitoring
more normal and problem fermentations. Future work in this regard, will focus on further
optimisation and expansion of the quantitative and qualitative calibration models and
implementation of these in the respective wineries of Namaqua Wines. / AFRIKAANSE OPSOMMING: Fermentasie is ‘n komplekse proses waartydens rou material getransformeer word na produkte
van hoë waarde, in hierdie geval, druiwesap na wyn. In die huidige ekonomies-kompeterende
samelewing, is dit al hoe meer belangrik om volhoubaar wyn te produseer wat voldoen aan
definieerbare spesifikasies en style. Goeie prosesbestuur tydens die wynproduksie stadium is
baie belangrik om herhaalbaarheid en gehaltebeheer te verseker. Problematiese
wynfermentasies het ’n direkte impak op beide kelderproduktiwiteit en wynkwaliteit. Die
voorkoming van slepende- of steekfermentasies, of selfs net om probleme te voorsien, sou
uiters bruikbaar wees vir ‘n wynkundige of wynmaker, wat dan die toepaslike regstellende
stappe kan neem waar nodig, om te verseker dat die wynbereiding suksesvol voltooi word.
Konvensionele metodes van monitering van alkoholiese fermentasie is tydrowend, soms
onbetroubaar en die inligting beperk tot ‘n paar parameters. Die huidige effektiwiteit van
fermentasie monitering in industriële wynproduksie kan heelwat verbeter word. Wynmakers
ervaar tans ’n behoete aan tegnologië wat die vroeë tekens van ongunstige fermentasiepatrone
kan identifiseer, en hul doeltreffendheid om moontlike regstellende aksies te neem, is dus
beperk.
Hierdie studie het die toepassing van Fourier transformasie mid-infrarooi (FT-IR)
spektroskopie in transmissie, ondersoek met die oog op kwantitatiewe en kwalitatiewe
monitering van alkoholiese gisting tydens industriële wynproduksie. Die vernaamste
navorsingsdoelwitte was eerstens om ’n portefeulje van kwantitatiewe kalibrasiemodelle te
vestig, wat geskik is om die belangrikste kwaliteitsbepalende parameters in gistende mos te
kwantifiseer. Die tweede hoofnavorsingsdoelwit was ’n loodsstudie wat ondersoek ingestel het
na die opstel van multiveranderlike statistiese proseskontrole grafieke van aktief-gistende mos,
met die oog op aflyn-kwalitatiewe monitering van alkoholiese gisting in industriële
wynproduksie. Hiervoor is slegs FT-IR spektra gebruik. Vir die doel van hierdie studie is
monsters van ’n totaal van 284 individuele, aktief-gistende tenke van die sewe hoof wit kultivars
en hul versnydings en nege hoof rooi kultivars van Namaqua Wyne, Vredendal, Suid Afrika,
geneem. Al die monsters is met toepaslike chemiese metodes en FT-IR spektroskopie analiseer
tydens die parsseisoene van 2007 tot 2009.
Vir die kwantitatiewe strategie is parsiële kleinste kwadraat (PKK1) kalibrasiemodelle vir die
bepaling van die klassieke wynparameters etanol, pH, vlugtige suur (VS), titreerbare suur (TS)
en die totale konsentrasie van glukose plus fruktose herontwikkel, om beter te pas op plaaslike
Suid-Afrikaanse monsters. Nuwe PKK1 kalibrasiemodelle is ontwikkel vir die komponente
glukose, fruktose en gis-assimileerbare stikstof, aangesien hierdie komponente gereelde
aanduidings van probleemgisting is. Die regressiestatistieke het die standaardvoorspellingsfout
(SVF), bepalingskoëffisiënt (R2) en sydigheid ingesluit en was gebruik om die prestasie van die
herontwikkelde kalibrasiemodelle vir plaaslike Suid-Afrikaanse monsters te evalueer. Etanol
(SVF = 0.15 %v/v, R2 = 0.999, sydigheid = 0.04 %v/v) het baie goeie regressiestatistiek getoon
en met ‘n relatiewe voorspellingsafwyking (RVA) van 30, was dit ‘n uitstekende model vir
kwantifisering in gistende mos. Die modelle vir pH en VS met RVA waardes van 4 en 3
onderskeidelik, is geskik vir semi-kwantitatiewe toepassings. Die kalibrasiemodel vir TS met ‘n
RVA waarde van 2, was nie geskik vir akkurate kwantifisering nie, maar wel vir semikwantitatiewe
analises. Die kalibrasiemodel vir die totale glukose plus fruktose inhoud in
gistende mos, met ‘n RVA waarde van 13, het uitstekende regressiestatistiek gegee en is
geskik vir akkurate kwantifiseringsdoeleindes. Die nuut-ontwikkelde kalibrasiemodelle vir
glukose en fruktose, met RVA waardes van onderskeidelik 8 en 10, is geskik vir akkurate
kwantifisering van hierdie belangrike parameters. Die kalibrasiemodelle vir etanol, totale
glukose plus fruktose, en glukose en fruktose afsonderlik, het uitstekende korrelasies getoon
met plaaslike Suid-Afrikaanse monsters en is gereed om toepassing te vind in die wyer
wynindustrie.
Twee kalibrasiemodelle is ontwikkel om gis-assimileerbare stikstof in gistende mos te
bepaal, deur gebruik te maak van verskillende verwysingsmetodes van analise; hierdie metodes
was ‘n ensiem-gekoppelde spektrofotometriese toets en die Formoltitrasie metode. Resultate
het getoon dat goeie regressiestatistiek vir FT-IR spektroskopie-gebaseerde kalibrasiemodelle
waar data wat met die ensiem-gekoppelde toetse verkry is, as verwysingwaardes gebruik is, in
wit gistende mos (SVP = 14.10 mg/L, R2 = 0.909, sydigheid = -2.55 mg/L, RVA = 6), maar nie in
rooi gistende mos nie. Die Formoltitrasie metode as verwysingsmetode, was geskik vir die
ontwikkeling van goeie kalibrasiemodelle in beide rooi- en wit gistende mos (SVP = 16.37 mg/L,
R2 = 0.912, sydigheid = -1.01 mg/L, RVA = 4). ’n Sekondêre, maar baie belangrike bevinding is
gemaak met betrekking tot die stoor van mosmonsters wat geneem is van tenke, maar wat nie
dadelik met die verwysingsmetodes en FT-IR spektroskopie analiseer kon word nie.
Multiveranderlike hoofkomponentanalise op vars en gevriesde sapmonsters het getoon dat
gevriesde monsters gebruik kan word om die kalibrasie datastel uit te brei, wanneer benodig.
Dus, sapmonsters kan gevries word tot ’n later stadium as onmiddelike analises nie moontlik is
nie.
Vir die doel van die tweede navorsingsdoelwit van die studie, naamlik kwalitatiewe af-lyn
monitering van alkoholiese fermentasie met FT-IR spektroskopie, is ‘n totaal van 21 industriëlegrootte
fermentasietenks ge-monitor deur sapmonsters met 8- tot 12-uurlikse intervalle te trek,
vanaf die begin van fermentasie, totdat al die druifsuiker gemetaboliseer is. Vir hierdie deel van
die werk is die kwantitatiewe data nie gebruik nie; slegs die FT-IR spektra. Multiveranderlike
statistiese proseskontrole grafieke is opgestel op grond van die PKK tellings wat bereken is op
al die FT-IR spektra wat gemeet is by die verskillende tydsintervalle. Vir hierdie analise is tyd as
y-veranderlike gebruik. Die vernaamste doel van hierdie ondersoek was om te evalueer of die
PKK-gebaseerde modelle kon onderskei tussen normale en slepende gistings. Die modelle wat
verkry is, het die variasie oor tyd in die fermentasiepatrone tussen wit- en rooiwyn fermentasies
tydens alkoholiese gisting, duidelik uitgewys. Een Colombar tenk wat teen baie lae temperatuur
gefermenteer is om ‘n spesifieke wynstyl te verkry, se fermentasiepatroon het aansienlik verskil
van die ander Colombar tenks wat gemonitor is, en hierdie atipiese patroon is ook deur die
kwalitatiewe modelle identifiseer. ‘n PKK model oor al die Colombar fermentasies kon duidelik
tussen normale en slepende gistings onderskei.
Die resultate wat in hierdie studie verkry is, het getoon dat FT-IR spektroskopie baie goeie
potensiaal toon vir die aanwending van kwantitatiewe en kwalitatiewe monitering van
alkoholiese fermentasie tydens industriële wynproduksie. Die werk wat in hierdie projek gedoen
is, het gelei tot die vestiging van ‘n portefeulje van kalibrasiemodelle vir die belangrikste
kwaliteitsbepalende parameters in fermenterende mos. Die kwantitatiewe modelle is baie
deeglik getoets met onafhanlike toets datastelle, en daarna is die kalibrasiemodelle geimplementeer
vir industriële gebruik by Namaqua Wyne. Multiveranderlike statistiese
proseskontrole grafieke wat baseer is op data wat vanaf 21 verskillende fermentasietenks
verkry is, het baie goeie potensiaal getoon om probleemfermentasies vroeg te identifiseer. Dié
grafieke is ‘n baie nuttige diagnostiese hulpmiddel wat verder ontwikkel kan word om
verskillende tipes probleemfermentasies te monitor. Toekomstige navorsing in hierdie konteks,
sal toegespits word op die optimisering en uitbreiding van die kwantitatiewe en kwalitatiewe
modelle, sowel as toepassing van die tegnieke in die onderskeie kelders van Namaqua Wyne.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:sun/oai:scholar.sun.ac.za:10019.1/2754 |
| Date | 12 1900 |
| Creators | Magerman, Cynthia M |
| Contributors | Nieuwoudt, Helene H., University of Stellenbosch. Faculty of Agrisciences. Dept. of Viticulture and Oenology. Institute for Wine Biotechnology. |
| Publisher | Stellenbosch : University of Stellenbosch |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis |
| Rights | University of Stellenbosch |
Page generated in 0.0029 seconds