Screwcaps for Wine – the start of a revolution
Ever since we started making wine in Kumeu in 1944, effectively all of our
wines have been sealed with cork closures. In the mid 1980’s we started to
notice problems with cork taint in some of our wines, and this came to a
head with the 1989 vintage Kumeu River Chardonnay, when we experienced a
disproportionately large number of tainted bottles. At this stage we
investigated other sources of corks, including direct purchase from European
suppliers and then air-freighting the corks as we needed them. Our corks
usually come from two local suppliers who import the corks in large bales of
10000, then treat and print them before repackaging in bags of 1000. These
European cork suppliers proved to be no better in terms of cork taint, plus
there was the added difficulty of having to order corks at least 3 months in
advance, so we returned to our local suppliers.
We also looked at synthetic corks and trialled the Cellucork brand with some
of our 1993 Chardonnay. Then in 1994 we tried Supremecorq. The Cellucork
worked very well for the first year, giving no taint at all. However, after
twelve months these closures lost their elasticity and started to leak, and
they were also slowly disintegrating at the surface in contact with the
wine. The Supremecorq looked good for about six months, but after that the
wines started to oxidise badly and also showed a distinctive “plastic”
It was decided to persevere with natural cork, and to put up with the
inconvenience of what we knew was between 3% and 5% cork taint. At that
level the problem was not major, although we were starting to get a
significant number of complaints and bottles returned. We felt this was an
unavoidable consequence of being in the wine business, and accepted it with
good grace. What was difficult to gauge, however, was the damage that faulty
corks was doing to our overall reputation as a producer of quality wine. We
are sure that many customers did not identify cork taint, but just thought
that the wine was perhaps not very good. We have no way of knowing how many
good customers were lost through this process.
The 1998 vintage threw up some more problems. We used some batches of cork
that were clearly much worse than normal, and in some isolated cases we had
as much as 50% cork taint. In addition to this there were many examples of
oxidation and other negative flavour modifications caused by the cork. So,
in 1999 we initiated a program of cork testing to try and eliminate bad
batches, and thereby minimise the effect of cork taint on our wines.
The cork testing procedure we finally settled on involved firstly taking 50
corks at random from each bale of 10000 corks that were available from our
cork supplier. Each of these 50 corks was immersed in a vial containing
approximately 70 mL of neutral dry white wine, and then sealed with a lid.
In this way the entire cork surface was in contact with the wine. These
vials were stored in a heated room at 25º C for 24 hours, and then the wine
from each vial was decanted into a tasting glass and evaluated for signs of
cork taint. Our criterion for acceptance of a cork batch was a maximum of 2
glasses in 50 showing signs of cork taint, which represents 4% taint.
Out of 63 batches of cork tested in this way during 1999, only 22 batches
were accepted for use. The other 41 batches were apparently sold to other
wine companies. In many of these rejected batches the taint that we detected
was greater than 10%, and in one case it was greater than 30%. Through this
testing process we were also becoming aware of faults in the wine other than
the typical TCA (2,4,6-trichloroanisole) cork taint. These taints included
“woody” and “phenolic” characters that were not evident in the control
sample, as well as “flat” or “dull” characters and excessive colour pickup.
A shining example was shown us in 2000 by a brave group of winemakers in the
Clare Valley of South Australia. They had experienced similar frustration
with cork quality, and at the end of 1999 were looking forward to the
results from a new cork treatment process which was supposed to cure the
cork taint problem. Unfortunately the results of the trial of microwave
treated corks were so poor they were left wondering what could possibly be
done to fix the problem. Instead of tried to rectify an inherently flawed
system, they instead looked to a tried and true alternative closure: Stelvin.
The big problem they faced initially was the availability of a suitable
bottle, as most screwcap bottles looked “cheap”. They worked closely with
SaverGlass in France to produce a tall Riesling bottle with a BVS (Bague
Verre Stelvin) finish, and a 30mm fill height. Most 750 mL bottles have a
55mm fill height (that is, from the top of the glass finish to the wine) to
allow for the insertion of a 45mm cork. The 30mm fill height was necessary
to reduce to a minimum the ullage space in the bottle, to avoid too much air
exposure after sealing.
This group of Clare Valley producers decided to bottle all or part of their
2000 vintage Riesling using the Stelvin closure and SaverGlass bottle, and
their example stirred the interest of ourselves and a number of New Zealand
wine producers who had similar frustrations with the poor quality of cork
closures. A group of Marlborough Winemakers held a meeting in February 2001
to also discuss possible alternatives to cork closures. Similar to the Clare
Valley group, they identified long-skirted screwcaps as the most promising
closure to replace cork, and a technical subcommittee was formed to
investigate the viability of screwcap closures. Before long there was great
interest from other parts of New Zealand, and a national organisation was
formed. The New Zealand Screwcap Wine Seal Initiative has more than 30
members from all over New Zealand, and its functions include technical
assistance to its members in correctly applying screwcaps, and also
promoting the screwcap to consumers as a quality closure for all types of
wine. Michael Brajkovich is the current Chairman of the Screwcap Initiative.
The Screwcap is also known as ROTE, or roll-on tamper-evident. “Stelvin” is
the proprietary name of one particular brand of Screwcap, manufactured in
France by Pechiney. It consists of two parts: a. The Outer, which is made of
malleable aluminium alloy, and is rolled on to closely fit the bottle
profile, and b. The Liner, which provides the seal between the closure and
This cutaway diagram shows clearly the configuration of the screwcap
components, and their positioning on the bottle. The screwcap outer has four
It presents the liner in the correct position for sealing,
It moulds to the bottle and holds the liner in place against the bottle
mouth with the required pressure,
It provides the thread to help the eventual removal of the cap, and
It provides space for decoration.
The liner is the crucial part of the sealing system. It is a polylaminate of
An 80 micron layer of neutral PVDC film, which is the only part which can
come in direct contact with the wine, and also with the bottle rim.
A 19 micron layer of tin foil, which provides the impermeable gas barrier,
A wad of approximately 2mm depth, made of expanded polyethylene, which
gives the elastic resilience to maintain the compression of the liner onto
the bottle rim.
The capping machine compresses the liner onto the bottle with a force of
120kg, and this is held in place by the aluminium outer being wrapped onto
the bottle. The resultant seal is a high pressure hermetic seal which is
capable of withstanding relatively large pressure and temperature increases.
This is in stark contrast to cylindrical stoppers, which are low pressure
elastic seals that rely on the elasticity of their material (cork or
synthetic) and their length of contact with the glass of the bottle neck.
Such seals are very susceptible to temperature changes, and are often
subject to leakage.
The Screwcap, on the other hand, offers a total gas barrier which means
total exclusion of air (including oxygen), and inclusion of dissolved gases
such as CO2. This is not the case with cylindrical stoppers, which all
exhibit some degree of gas permeability in both directions. One of the
superior features of the new generation Stelvin closures is that of
“redraw”. This is performed by the specially designed capping block which
draws the aluminium outer very closely around the glass rim profile. In so
doing it removes any air gaps that may provide room for indentation that
then could result in seal failure. This is a major improvement in sealing
efficiency and durability, and makes the seal safe from external impact.
Screwcaps, and the technology behind them, are not new. They have been in
commercial use now for well over thirty years and there is a great base of
practical, commercial experience underpinning their use. There have been a
number of key research projects comparing their performance to other
closures, and most of these have been conducted in Australia.
The first such research was published in 1976 by Eric, Leyland & Rankine.
Their main conclusion was that the metal closure with 358 wad, which was the
predecessor of the current Stelvin liner, had performed well and that its
use as a commercial closure for table wines was justified.
In 1980 further results from the same group: Rankine, Leyland & Strain, were
published. This time they confirmed the superiority of the Stelvin closures
over cork, also stating that those wines sealed with Stelvin had matured
better than under cork.
It was not until 2001 that another study was conducted looking at wine
closure types. This time Godden et al commenced a very wide-ranging and
complex trial which tested 14 different closures, including traditional
natural cork, technical corks, synthetic closures, and one type of Screwcap.
The scope of the trial meant that it was practical to use only one wine, and
the wine chosen was a Semillon from the Clare Valley. The first results were
published after 20 months of bottle maturation in controlled conditions.
The Chemical Analysis results gave the following conclusions regarding
Lowest reduction in Free and Total SO2.
Highest retained Free SO2.
Highest retained Ascorbic Acid.
Lowest incidence of Browning, as measured by Optical Density at 420nm
Least variation between bottles for all compositional variables.
The Sensory Evaluation of the samples in the trial also yielded the
following conclusions about screwcaps:
Highest in Overall Fruit.
Lowest in developed and oxidised characters.
Zero TCA (cork taint)
In presenting our Stelvin-sealed wines to consumers we are often met with
surprise, in some cases shock, and many questions. The type of questions
that typically arise include: “But isn’t the cork needed for the wine to
develop and age properly?” or “That’s OK for current consumption wines, but
great wines need corks to age, don’t they?” and “The wine is a living thing,
and it needs to breathe through the cork, doesn’t it?”
Science has shown us otherwise, and in every case the answer is no.
The evidence comes from a great body of research that commenced in France in
the 1930’s, and continues through to the present day.
J. Ribéreau-Gayon et al (1976)“Traité d’Oenologie – Sciences et Techniques
du Vin” Vol.3
“…les quantités d’oxygène qui pénètrent normalement dans les bouteilles sont
infimes sinon nulles. L’oxygène n’est pas l’agent du vieillissement normal
In English: “…the quantities of oxygen that normally penetrate into the
bottles are infinitesimal if not zero. Oxygen is not the agent of normal
This is quite categorical, and stems from research in the 1930’s that showed
that oxygen is not involved in bottle maturation. Therefore, the wine in
bottle does not need to “breathe” in order to age properly.
E. Peynaud (1984) “Knowing and Making Wine”:
“It is the opposite of oxidation, a process of reduction or asphyxia, by
which wine develops in the bottle.”
Peynaud states what the bottle ageing process actually involves:
The reactions involved in bottle maturation are reductive rather than
as opposed to
The reduction of acetaldehyde to ethanol is the last step in alcoholic
fermentation. The oxidation of ethanol to acetaldehyde during barrel
maturation is a key step in the polymerisation of tannins in red wine.
P. Ribéreau-Gayon et al (2000) “Handbook of Enology - Vol.2 The Chemistry of
Wine Stabilization and Treatments”:
“During bottle aging, wines develop in a reducing environment, tending
towards greater organoleptic quality than they initially possessed”
P. Ribéreau-Gayon et al (2000) “Handbook of Enology - Vol.2 The Chemistry of
Wine Stabilization and Treatments”:
“When a wine ages in the bottle, the oxidation – reduction potential
decreases regularly until it reaches a minimum value, depending on how well
the bottle is sealed. Reactions that take place in bottled wine do not
Bottle Maturation by be defined as the aging process which takes place in
the bottle after sealing. There is necessarily a long time frame involved,
and that is typically in the order of years rather than months. It is
responsible for the development of bottle “bouquet”, and it is high quality
wines which will benefit the most. The development of favourable bottle
maturation bouquet arises from slow, reductive reactions, in the absence of
The Advantages of using Screwcaps to seal wine bottles may be summarised as
There is a total absence of taint
Superior retention of wine quality characteristics, both analytical and
The ability to allow further bottle maturation, because the seal is
perfect and still allows for reductive development in the absence of
Screwcaps are very convenient in terms of both removal and re-sealing.
on all the evidence accumulated over 30 years, and on current observations
of our own Kumeu River wines as well as others, we have decided now to
bottle all of our wines with the Stelvin closure.
In the twelve months since our first experimentation with Stelvin, we have
been most encouraged by what we have seen. In the white wines there is an
immediate and consistent improvement in wine quality over the same wines
sealed with cork. The fruit characters are brighter and cleaner, and yet the
wines are still able to mature gracefully in the bottle.
The differences are less clear in the red wines because of the red wine
phenolics (tannins), although with Stelvin we can now perceive the absence
of a dusty, woody character in the wine that we previously attributed to
barrel ageing. We are very confident that over a two or three year
time-frame the differences in the red wine will become even more pronounced,
and that the wines sealed with Stelvin will be markedly superior.