In the first phase of the Bacchus research, we discovered that perhaps Bacchus is not as similar as Sauvignon blanc as we first thought. This was due to the abundance of terpenes and trace presence of the thiols that are synonymous with the tropical and grapefruit aromas of New Zealand Sauvignon. We also noted that Bacchus is an incredibly complex wine, showing a relatively high amount of long-chain esters that are very fruity and aromatic.

During the 2016 vintage I isolated some Bacchus juice that had been reductively pressed under nitrogen, in order to carry out some trial runs for the second stage of my project. The dissolved oxygen (DO) was kept low and measured at <1 % after racking. I partitioned the juice into six 18-litre stainless steel fermentation kegs and then hyper-oxidised three of them. I did this by vigorously moving the juice around in contact with air until it turned brown. I’d hoped to be able to quantitatively measure the amount of DO that was added to the juice but this proved very tricky. The oxygen was consumed faster than I could add it meaning it was impossible to measure with my DO meter. I’ve since spoken to a number of wine researchers who specialise in analysing oxygen and they’ve given me some tips on how to do this for next year. It is a tricky procedure to get right but can be done.

So, in the end I had two variables in triplicate ready to ferment: three kegs of reductively pressed juice and three kegs of juice that was saturated with oxygen. They were both then inoculated with the same yeast strain and fermented at a temperature between 12 and 15 C (except for towards the end of fermentation and during the middle of the reductive batch when temperature was increased to encourage the fermentation to complete). An organic nitrogen addition at 25 g/hL was added 1/3 of the way through the fermentation.


The fermentation kinetics were distinctly different between the two types of juice. The oxidised juice raced away and dropped 30 oeschle in the first two days of fermentation. The reductive juice took a long time to get going and took 12 days to make the same sugar drop! It never really caught up and took just over 3 weeks to complete with the oxidative batch taking just over 2 weeks.

Sensory observations

During fermentation, the reductive batches showed some signs of reductive characteristics with some possible signs of slight hydrogen sulphide. This blew off by the end of fermentation without the need for aerating or ammoniacal nitrogen additions.

Post-fermentations the reductive fermentations were slightly tighter on the nose and palate, showing more signs of citrus lime and herbaceous aromas. They seemed more textured in the mouth and the oxidised-juice fermentations slightly more open.

After racking and sulphite addition the wines were re-tasted following a month of natural settling. There was some variation across the batches but overall the reductively made wines showed a greater expression of fruit and a more complex nose. If anything the wines made from oxidised juice were a little more approachable overall, with a rounder palate and nose, although less complex.

It should be noted that the tastings were not carried out blind or intended to be true sensory trials or to demonstrate any potential statistical significance. They were simply for observation.

Aroma analysis

Each of the batches of three wines were blended and a small sample was sent for Gas Chromatography / Mass Spectroscopy (GC/MS) analysis at Campden BRI. Geoff Taylor, who heads up the operation there, has developed a full scan, which can qualitatively measure the levels of different aroma compounds in a wine.

Interestingly there was little difference in the aromatic profiles of the wines. The wine made oxidatively actually showed an additional compound (1-undecanol, found in apples and bananas and that has a creamy taste), which I need to further investigate.

Both wines showed very high levels of aromatic esters and terpenes as well as the ubiquitously present methoxy-phenyl oxime that was found in all Bacchus wines from stage one. A new terpene (linalool, a floral aroma found in Muscat) was seen in abundance, which had not appeared in the other English Bacchus wines. This is positive news for the wines at Flint Vineyard and also reinforces the fact that Bacchus is perhaps not as similar to Sauvignon blanc as we thought. It also confirmed that further research is warranted into understanding the importance of the complex esters that are unique to Bacchus. I also have a hunch that the oxime compound may also be of importance to this grape variety that can be difficult to work with and intriguing and complex in the aromas it produce. Although this may just be a hunch and completely false.

Stage two of the project will be resumed this vintage (2017) when we will hopefully start delving a bit more and quantitatively determining the volumes of Bacchus aromas that are produced under different fermentation conditions.