Prozessentwicklung zur Traubenverarbeitung und Mostgewinnung im Weinberg Development of a method for grape processing and juicing in the vineyard Hühn, T. (1) ; Häfele, M. (1) ; Erbach, M. (2) ; Hamatschek, J. (3) ; Köper, I. (2) ; Bernath, K. (1) ; Pecoroni, S. (3) ; Petry, W. (2) ; Brähler, F. (2) ; Lipps, M. (5) ; Walg, O. (5) ; Hamm, U. (5) ; Schauz, F. (3) ; Horstkötter, L. (4) ; Schmitt, I. (3) ; Dietrich, H. (6) ; Bamberger, U. (5) (1) Zürcher Hochschule für angewandte Wissenschaften, Departement Life Sciences und Facility Management, Institut für Lebensmittel- und Getränkeinnovation, Zentrum für Getränke- und Aromaforschung, Grüental, CH-8820 Wädenswil, tilo.huehn@zhaw.ch; www.beverages.ch; Tel. +41447899705; Fax +41447899950 (2) ERO Gerätebau GmbH, Niederkumbd (3) Westfalia Separator Food Tec GmbH, Oelde (4) Westfalia Separator Industry GmbH, Oelde (5) Dienstleistungszentrum Ländlicher Raum Rheinhessen-Nahe-Hunsrück, Bad Kreuznach (6) Forschungsanstalt Geisenheim, Fachgebiet Weinanalytik und Getränkeforschung, Geisenheim Zusammenfassung Bei der Weinbereitung spielt der Faktor Zeit aus technologischen, biochemischen und ökonomischen Gründen nach wie vor eine bedeutende Rolle. Für die verschiedenen oenologischen Prozesse von der Traubenernte bis hin zum gärfähigen Most müssen im Hinblick auf die erwünschte Endproduktqualität optimale Bedingungen und Zeitabläufe geschaffen werden. In diesem Zeitraum liegen zudem einige Risiken für die angestrebte Produktqualität, wie zum Beispiel unkontrollierte Enzymaktivitäten, die Vermehrung von unerwünschten Mikroorganismen, sowie die mechanische Belastung der Trauben durch die einzelnen Verarbeitungsschritte. Durch die rationelle Ernte und die unmittelbare Entsaftung der Maische im Weinberg werden die Prozesszeiten erheblich verkürzt. Daraus resultieren verschiedene Vorteile gegenüber dem konventionellen Traubenvollerntereinsatz: - verbesserte Kontrolle mikrobiologischer und enzymatischer Prozesse - verminderte Extraktion von unerwünschten Inhaltsstoffen aus Pflanzenbestandteilen - direkter Verbleib von Trester- und Trubbestandteilen im Weinberg Die Realisierung dieser Vorteile erfordert ein leistungsfähiges, kontinuierliches Entsaftungssystem, das mit der heutigen Vollerntertechnik kombiniert werden kann. Die erforderliche Maschinenkombination eines Traubenernteentsafters wird seit 2005 in Deutschland, Chile und Frankreich getestet und im Herbst 2009 weiterentwickelt. Summary For technological, biochemical and economic reasons, the time factor still plays a significant role in winemaking. With an eye to the desired end-product quality, optimum conditions and timeframes must be created for the various oenological processes from the grape harvest to fermentable must. There are also several risks affecting the desired product quality apparent within this timeframe, such as uncontrolled enzyme activity, the development of unwanted micro-organisms and the mechanical strain on the grapes during the individual processing stages. A rational harvest and immediate juicing of the must at the vineyard considerably shortens the process times. This yields various advantages in comparison to conventional mechanical harvesters:
- easier control of microbiological and enzymatic processes - reduced extraction of unwanted plant matter - grape marc and solids stay at the vineyard Realising these advantages requires a powerful and continuous de-juicing system which can be combined with modern harvesting technology. The necessary machine combination has been tested since 2005 in Germany, Chile and France and will be developed further in the 2009 harvest. 1 Effect of continuous juicing in the vineyard on must and wine quality 1.1 Must density, turbidity, total acidity and ph The process-related qualitative impact on the final product of using a combination of harvester/decanter technologies for grape de-juicing in the vineyard must be analysed. The musts obtained here, using hydrodynamic de-juicing, differ in particular from those harvested by hand or with a conventional harvesting machine, followed by de-juicing using a membrane press, in terms of turbidity in the colloidal range. No significant differences detected with respect to the standard parameters such as must density, ph and total acidity. The total polyphenol content measured in the wine from the Juiceliner is between 0 and 30% higher than from harvesting by hand followed by mash pressing, irrespective of the grape variety. Polyphenol concentrations, such as occur for mash vatting times in excess of 2-4 hours, were not exceeded. Depending on the health of the grapes and the harvest achieved, there were significant differences in the turbidity of the must prior to clarification.especially in the case of severe Botrytis infestations of the grapes (> 40% infection intensity) and harvesting capacities close to the process maximum (a flow of 9000 kg / h), centrifugal deposits of 10% to 15% (v/v) were found, depending on the grape variety. A comparison of colloidal turbidity shows that the musts obtained by means of the decanter exhibited a higher turbidity, before clarification, than the whole grape cluster press. The turbidity values are comparable with those of the mash press variants. A trend, dependent on the variety and health of the grapes, can be identified in a shift in particle size distribution within the turbidity spectrum, towards smaller particle sizes of <1mm.This Q?Feinsttrub?Q includes value-determining constituents from the grape skin, which find their way into the must under certain extraction conditions. After primary sedimentation, the turbidity levels of the Juiceliner variants tend to be lower than in the other variants (data not shown).
yield (kg/kg) [%] 1.2 Yield The comparison of different methods shows a varying picture with regard to the yields achieved (kg/kg), depending on grape variety (Fig. 1). The whole cluster and mash pressed variants were harvested by hand. For the mechanically harvested variants, the grapes were harvested using conventional harvesting equipment. All three variants were de-juiced using membrane presses. Approximately 1,500 m2 were harvested for each variant. The Grapeliner () and mash press variants were destemmed and squeezed before pressing. In 2005, the yield achieved (kg/kg) tended to be 3-10% higher than the comparison variants, or of a similar level (Riesling). In 2006, however, the yield achieved with the Juiceliner () was 3-10% lower than the highest yield for each of the other variants, or of a similar level (Müller-Thurgau). 100 90 80 70 60 50 40 30 20 10 0 Müller- Thurgau Sylvaner Riesling Müller- Thurgau Sylvaner 2005 2006 Riesling Figure 1 Yield realised (kg/kg) for the variants studied in Bad Kreuznach for 2005 and 2006. Grapeliner (), Juiceliner (), whole cluster press () mash press () In 2007, on a test area of 4 hectares in Rheingau, half the grapes were harvested using the Juiceliner and the other half using a conventional harvester and then de-juiced in a membrane press. The rows were harvested alternately. In this experiment, the Juiceliner yielded 15% (w/w) below the 80.5% (w/w) yield of the reference variants. 1.3 Aroma analysis 1.3.1 Fatty acid ethyl esters Fatty acid ethyl esters result from the esterification of the respective fatty acids with ethanol, and are enzymatically formed by the yeast during fermentation. These esters increase the impression of the wine's fruity odour (Moreno-Arribas and Polo 2009). The fatty acid ethyl ester contents of the 2005 Müller-Thurgau Juiceliner and 2005 Riesling Juiceliner variants were higher than for the other variants (Fig. 2). The cause of this difference is mainly due to the different turbidity values. The fatty acid ethyl ester formation in the 2005 vintage correlates strongly with the turbidity levels in the pre-
clarified must. The clearer the juice, the higher the synthetic capacity of the yeast (Ribéreau-Gayon 2006). The data generated from the 2006 vintage is unrepresentative, due to the severe Botrytis infestation of the grapes (> 80%). The esterases formed by Botrytis cinerea cause uncontrolled destruction of esters (fatty acid ethyl esters and acetate esters), making an assessment of the fermentation ester content impossible (Ribéreau-Gayon 2006). Fatty acid ethyl esters [mg/l] 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Müller-Thurgau Sylvaner Riesling Figure 2. Sum of the fatty acid ethyl esters (hexane, octane and decane fatty acid ethyl esters) after alcoholic fermentation in the 2005 vintage. Grape harvester (), Juiceliner (), whole cluster pressing (), mash pressing () 1.3.2 Linalool Figure 3 shows the Linalool content of the terpene-containing Müller-Thurgau and Riesling varieties. Due to the very low levels (<0.01 mg / l), both before and after fermentation, the data for the Sylvaner varieties is not shown. The differences between the variants used in the experiment are low. With regard to the extraction or release of terpenes during fermentation, the continuous de-juicing with the decanter and without vatting time showed no disadvantages resulting from the increased extraction of the grape skin.
Hexylacetat [mg/l] Linalool [mg/l] 0.10 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00 GP GP Müller Thurgau Riesling Figure 3 Linalool content of the 2005 vintage before ( ) and after ( ) alcoholic fermentation. Grape harvester (), Juiceliner (), whole cluster pressing (), mash pressing () 1.3.3 Hexanol Hexanol is formed enzymatically from multiple polyunsaturated fatty acids with 18 carbon atoms. The unsaturated fatty acids are mainly located in the grape skin and are released by long vatting times or mechanical action (Cardonnier and Bayonove 1981). Therefore, for whole grape cluster pressing, fewer fatty acids are available to be converted by enzymes into Hexanol. For the other variants, especially the Juiceliner variants, the extraction is greater and thus the hexanol levels are also higher (Fig. 4). 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 Müller-Thurgau Sylvaner Riesling Figure 4 2005 vintage hexyl acetate content after alcoholic fermentation. Grape harvester (), Juiceliner (), whole cluster pressing ( ), Mash pressing () During fermentation, hexanol is partially esterified into hexyl acetate by the yeast (Saerens 2006). Based on the hexyl acetate formation, this could be shown, before and during fermentation, with and without the addition of 5 mg / l hexanol. For this, a pasteurized Müller-Thurgau grape must at 20 C was fermented with VL3 Zymaflor yeast and analyzed. The results are shown in Figure 5.
Hexyl acetate content [mg/l] 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 Addition without Addition prior Addition in the middle No Addition fermentation fermentation of fermentation With fermentation Figure 5 Influence of Hexanol content on the formation of hexyl acetate during fermentation 2 Variations in process reliability The problems present in the results of fluctuating process reliability in terms of harvest performance near maximum throughput, high sublimate deposits in cases of grapes that had been severely putrified by botrytis, and insufficient yields were further studied in Germany and France between 2007 and 2009. Fluctuations in process reliability are caused by critical operating conditions. One of the constant determining factors at the time of harvest is the health of the grapes. For grapes that were subject to a severe Botrytis infestation (>40% infection intensity), it was shown that the decanter less flexible than the membrane press with regard to extraction variability. Fungal maceration of the grape skins usually leads to increased turbidity of the musts. While extraction diminishing whole cluster pressing takes these circumstances into account, possibilities to respond to such a scenario in continuous de-juicing remain limited. Since the hydrodynamic extraction from the berries often results in the skins being turned inside out and increased skin extraction, higher quantities of Botrytis cinerea - metabolites lead to clarification problems in the juice. Increased extraction of glucans has a negative effect on the filterability of the young wines. Experiments in 2006 (Rheinhessen) and 2008 (Rheingau) with disease incidence and intensity > 80% showed, however, that centrifugal deposits could be reduced to an acceptable level in the wine cellar by tailored turbidity management, using appropriate enzymes. The advantages of continuous de-juicing remain unaffected with regard to the short process duration and thus improved control of undesired enzymatic and microbial processes. A further critical operating condition arises from an accumulation of unwanted, green plant matter such as stems and the like during continuous de-juicing in the decanter. As a consequence of leaf senescence, this problem occurs increasingly in the second half of the harvest period, depending on weather conditions, other vintage-related differences and the nutritional state of the vines. The Accumulated leaf stalks and other plant matter of below-average specific weight lead to a reduction in separation efficiency during de-juicing, resulting in a higher centrifugal deposit in the must, and a reduced yield. This process occurs irrespective of the feed capacity and can only be rectified by shutting down the decanter, sometimes in conjunction with a cleaning cycle. Productivity lowering downtime is the result.
centrifugal sediment [% vol.] Feed rate [kg/h] This effect was observed several times in experiments conducted early and late in the harvest, over a five year period, including the experience from France, where senescence had not started because of the early harvest in August and September, (data not shown ). In the absence of this effect, even with feed capacities close to the process maximum, turbidity values, permitting immediate fermentation without prior clarification, can be achieved. Field trials in Rheingau in 2009 were able to clearly demonstrate the negative effect that could result from accumulated leaf stalks, etc. On an experimental plot of 4 ha, every second row of Riesling was completely stripped of leaves and stalks. As a result of this measure, the proportion of green plant matter in these rows was <5% (w / w). The Rheingau trials were conducted on 5th and 6th October this year, in the middle of the harvest period. Leaf senescence was moderately advanced. Disease incidence and strength in the berries as a result of Botrytis cinerea was approximately 10% (infection intensity). It should be noted that there was an increased proportion of dried berries caused by an early attack of downy mildew. With identical crop parameters (speed, vibration frequency and feed capacity) the rows without foliage and then the rows with foliage were harvested. The decanter was shut down between the row types, drained and rinsed. Based on the harvest volume, the feed rate for the test was limited to the middle third of the performance range at 5400 kg / h. The harvesting speed for a total feed of 5400 kg was 5.5 km / h. The feed rate and centrifugal deposits are shown in Fig.6. 14 12 10 8 6 4 2 6000 5000 4000 3000 2000 1000 0 0 0 0.5 1 1.5 2 2.5 Time [h] Figure 6 Centrifugal deposits during the harvest. Rows with foliage ( ( ) feed rate [kg / h] ( ) ), rows without foliage The centrifugal deposit values of the rows with leaves show a typical progression during the harvest reaching the critical operating conditions. After a short running in period, at a low feed rate, centrifugal deposits continuously increase with increasing feed rates. In contrast, the turbidity content from the rows without foliage shows, after a short increase in turbidity, a decrease to a level of 2% (v / v). The analysis of the pomace discharge, after shutting down the decanter, showed a significantly increased proportion of leaf stalks and other plant matter from the harvesting of the rows with foliage.
3 Optimization of process reliability - prevention of critical operating conditions Against the background of what has been described, of the critical operating conditions that have a negative effect, avoidance of the accumulation of leaf stalks and other unwanted plant matter is the focus for the optimization of process reliability. Two approaches were developed and tested last year, in 2009. 3.1 Improvement of destemmer sorting ability A solution for avoiding a critical operating state is to minimize the quantity leaf stalks and other unwanted plant matter, which make their way into the de-juicing process. For this purpose, a new destemmer and crusher was developed with a downstream sorting band added. Various perforated cylinders and spiked rollers were tested. With highspeed imaging each separation result was recorded and analyzed. Up until the next harvest, the geometry of the cylinder and holes will be revised and further tested. The separation results from the sorting band have, under the conditions tested to date, been positive. 3.2 Use of a macerator before the decanter Assuming that stems and other unwanted plant matter can never, even with increased sorting, be completely removed before the de-juicing process, the second approach aims to optimise the process reliability by shredding the harvested grapes to a size that is no longer a problem for the continuous de-juicing process. For this purpose, an upstream macerator was added to a stationary decanter. The decanter used here had a barrel diameter of 20 cm, in contrast to that of the Juiceliner with 50 cm. The grape material used, from the Muller-Thurgau variety, was harvested using a towed machine in a minimum cut vineyard and processed by means of a stationary destemmer. The proportion of unwanted plant matter after destemming was between 2-3% (m / m). The grape quality was extremely heterogeneous. In contrast to the continuous de-juicing, small 20 kg trials were carried out using the same grape material, where samples were left to stand for different lengths of time (2 and 6 hours) before being de-juiced using a membrane press. All unwanted plant matter was removed from these samples and then a defined volume of 2% re-added (w / w). The influence of shredding on the polyphenol index and the content of C 6 compounds (hexanal, (E)-2-hexenal, 1-hexanol, (E)-3-hexenol, (E)-2-hexenol) is shown in Figures 7 and 8.
Sum of C6 compounds [mg/l] Polyphenolindex DO 280 2.50 2.00 1.50 1.00 0.50 0.00 Macerator prior to Dekanter Decanter Mash vatting time 2h Mash vatting time 6h Figure 7 Polyphenol index after alcoholic fermentation; vatting times with 2% (m / m) green plant matter 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 Macerator prior to Decanter Decanter Mash vatting time 2h Mash vatting time 6h Figure 8 Sum of C 6 compounds after alcoholic fermentation, vatting times with 2% (m / m) green plant matter. In contrast to the decanter variant without macerator, a slight increase in the extraction of phenolic compounds can be identified resulting from the shredding. The value achieved is, however, at the same level as that of the small-scale tests with vatting times and de-juicing using a membrane press. The content of C 6 -compounds is likewise, as a result of the use of the macerator, slightly increased, and in a similar range to the samples that underwent vatting periods. The difference compared to continuous de-juicing without a macerator is +5%. All values are below the odour threshold in the wine matrix of 4 mg / l (Swiegers 2005). Due to the short processing time between macerator and decanter <1 minute, the risk of over-extraction of unwanted substances is limited.
4 Conclusion The first trials of the presented solutions show that critical operating conditions, caused by the accumulation of leaf stalks and other unwanted plant matter, can be avoided. Using a combined array of destemmer with sorting band or a macerator, there is the possibility of ensuring process reliability under the conditions caused by senescence. Further trials using a decanter with a bowl diameter of 50 cm and sensory evaluation of the wines will follow. By stabilizing the separation process, the current unsatisfactory yields under changing operational conditions will be improved. Building on the experience of recent years in mid-range performance, trials in the upper third of the system's perfomance range, over 7000 kg / h, will be carried out.. After further testing and new construction within the terms the project objectives, the chances for successful operation of the system in practice can be viewed as positive. Literatur CARDONNIER R. and BAYONOVE C. (1981). Etude de la phase prèfermentaire de la vinification, extraction et formation de certains composès de l arome : cas des terpenols, des aldèhydes et des alcool C 6. Connaissance Vigne Vin, 15, 269-286. MORENO-ARRIBAS and POLO (2009). Wine Chemistry and Biochemistry. Springer Science + Business Media, New York. RIBÉREAU-GAYON, P., et al. (2006): Handbook of Enology. Volume 1. The Microbiology of Wine and Vinifications. 2nd Edition. John Wiley & Sons Ltd., West Sussex. SAERENS, S. M. G., et al. (2006): The Saccharomyces cerevisiae EHT1 and EEB1 Genes Encode Novel Enzymes with Medium-chain Fatty Acid Ethyl Ester Synthesis and Hydrolysis Capacity. The Journal of Biological Chemistry, 281, 7, S.4446-4456. SWIEGERS, J. H., et al. (2005): Yeast and bacterial modulation of wine aroma and flavour. Australian Journal of Grape and Wine Research, 11, S.139-173