The Corks for Graham’s Ports

Readers of the blog will be familiar with our passion for quality and our attention to all the details of the viticulture and winemaking behind Graham’s Ports.  That same attention and passion is applied to every step and every component part of the process which delivers our wines to your tables.

Take for example, the cork in your bottle of vintage port:  a piece of natural material typically 49 millimetres long and 24 millimetres wide.  How detailed and passionate can you get about that?

Imagine yourselves in our shoes:  we worked for a year in the vineyards to grow and tend those grapes (we won’t even try to count the years of work that have gone into the vineyards), we knocked ourselves out working 24/7 during the harvest to vinify them in the best possible way, we nursed the wines in our cellars for two years for the Vintage wines (and even longer for other port styles), constantly tasting, blending, storing carefully, monitoring the quality of the wine itself and ensuring that every step in the handling of that wine from cask to bottle to you maintains that quality.  There’s nothing more satisfying than cleanly pulling out the cork and then settling down to enjoy a well-aged vintage port with family and friends.  Need we say it?  Being let down by 49 mm x 24 mm of cork absolutely breaks our hearts.

So, yes, we do get extremely passionate and detailed about trying to ensure the quality of the corks we use in Graham’s ports.

From every batch of corks delivered to us, we remove a significant sample quantity and run a half dozen different tests to assure ourselves of the quality of those corks before they are released for use in our bottling plants.  The technicians in our quality assurance laboratory have full power to reject an entire shipment of corks if the test samples fail to meet our standards for any of our tests.

First is a test specifically for TCA.  What’s that?  Careful what you ask:  it’s 2,4,6-trichloroanisole, or, for the layman, it’s the by-product of a chemical reaction which creates that smell commonly referred to as “cork taint.”  To try to ensure the corks will not cause this type of reaction, we select sample corks and immerse them in a special alcoholic solution for 24 hours.  After 24 hours we smell the solution – if the lab team detect any whiff of TCA we will reject the entire batch of corks, and not even bother to test further.

Next, we pour out some port into a petri dish, and set five corks into the dish for 24 hours, to check the rate of absorption of wine by the corks.

Lab technician João Magalhães sets the pressure to one bar in a test bottleneck containing azul metallica which is fixed to the tray below his hand

Tests are also performed to determine how much pressure is required to extract a cork (ideally between 20 and 40 kg), and how much to force the cork down fully inside the bottle (not less than 15 kg), to ensure the cork will be snug within the neck and create a good seal.

Finally, two more tests are made to determine the absorption of liquid by the cork.  In one test, bottles are filled with azul metallica, an intensely blue-dyed water, corked, and laid on their side for a week.  The other test fills a “bottle neck” with the azul metallica, corks it, and then creates a pressure of 1 bar inside the neck.  After a week we extract and examine the corks from both the bottles and the test-necks to determine if any blue dye has been absorbed into crevices in the cork.

Only after all these tests have produced satisfactory results will our Quality Control lab approve a batch of corks for use in bottling.  But we don’t stop analysing the performance of the corks with pre-bottling tests.

30 bottles of Graham's Crusted 2002 called into the lab from the lodge for a random test of cork quality after some years ageing in bottle

On a routine basis we will call in 30 bottles of a vintage port from storage at the lodge, extract the corks, and promptly re-cork the bottles.  We then examine the original corks to monitor their performance, and take any action we feel necessary to ensure the continued quality of our stored wines.

Possibly the most critical thing about cork is its compression inside the bottle neck – but we are so passionate and detailed about that, it will take a dedicated article to explain it properly.  Stay tuned.

Share this post

4 thoughts on “The Corks for Graham’s Ports

  1. What do you think is an acceptable “failure rate” for Port corks? As a rough estimate, I think about 8-10% of Vintage Port I have drunk has been tainted in some way, usually due to being corked or by some other reason connected to the cork (e.g. seepage). I wonder whether this accords with what you experience? That equates to about one bottle per 12-bottle case which seems disappointingly high considering that Vintage Port, if not expensive for a fine wine, is not cheap either. The failure rate is lower for cheaper Ports but then they, of course, do not tend to get cellared for such a length of time.

    I think many Port drinkers would be reassured if they knew that any properly-cellared Port they buy from, say, 2000 onwards is likely to have a much lower failure rate than before due to the new measures you are taking. Whilst it must break your hearts being let down by a piece of tree bark, I think it’s probably worse when you have just drawn the cork on an £80 bottle, looking forward to a glass of first-rate, perfectly mature, Vintage Port to find that smell of cardboard coming out of the decanting funnel!

    1. Hi Jacob,

      There is no failure rate that’s “acceptable” – no target failure rate we aim for. If we said one in a million, and you were the person to open that one millionth bottle – would that be acceptable to you? Probably not, and we feel the same. Fundamentally, we can only do our very very best to review and monitor the quality of the corks supplied to us, and reject the entire batch – as many as 100,000 corks – if even one of the test samples causes us any concern, as described above.

Leave a Reply

Your email address will not be published. Required fields are marked *