Issue link: http://winesandvines.uberflip.com/i/776679
February 2017 WINES&VINES 43 PRACTICAL WINERY & VINEYARD WINEMAKING Two-year trials in a Bordeaux first-growth château The trial was conducted using Bordeaux bar- rels at a first-growth chateau in France's Pauillac appellation during the 2013 and 2014 vintages of the estate's flagship wine (majority Cabernet Sauvignon). For both years, the same wine was put into different trial barrels after malolactic fermentation (January after the harvest), and each trial was duplicated. The experiment included 16 bar- rels (3,600 liters) in 2013 and 32 barrels (7,200 liters) in 2014. Élevage was conducted according to the château's usual procedure, including racking and returning to the same barrel every three months. Wine from each barrel was analyzed after 15 months of élevage, following fining with egg white and removal of sediment. Wines from duplicate barrels were blended in equal quanti- ties for sensory analysis. Geographic origin of the wood and experimental parameters The Vicard Group's wood buyer was respon- sible for ensuring traceability. The trees were felled in winter, when the sap level was low. Stave wood for barrel production was selected after 30 months of natural seasoning at the Vicard Cooperage woodlot in Cognac. Every barrel in each protocol was built by assembling wood from 30 trees, using one stave per tree for the body. The two heads were built from 14 pieces, each taken from a different tree from among the 30 sourced for the staves. For the 2013 vintage trial, a ranking by tannin potential (TP) was performed for six geographic origins. For the 2014 vintage trial, the barrels were initially sorted by grain and then ranked by TP. Sorting by grain and tannin Two types of grain were selected: fine grain (f) and medium grain (m), corresponding to growth rings less than 2 mm wide and be- tween 2 and 3 mm, respectively. The oak was sorted by grain prior to TP analysis. After machining, the untoasted staves were analyzed by near-infrared spectrom- etry using an acousto-optic tunable filter crystal detection system. Only two TP levels were selected for the experiment: low TP (LTP), or less than 4,000 µg/g, and high TP (HTP), between 6,000 and 8,000 µg/g (val- ues expressed in µg ellagic acid equivalent/g dry wood). The 2014 trial involved logs from the Bercé forest. GEOGRAPHIC ORIGIN OF OAK STAVES USED 2013 Trial: Fontainbleau, Loches, Tronçais and Orléans forests 2014 Trial: Bercé, Darney, Loches and Tronçais forests