Keeving - What's That?

Keeving  - French and English tradition

'Keeving' is a way of making the ultimate style of naturally sweet sparkling cider. This is traditional both in Western England and the northwest of France, but whereas it has virtually died out as a commercial proposition in the UK, it is still very much alive for the production of ‘cidre bouché’ in France. The underlying principle is to remove nutrients from the juice by complexation with pectin at an early stage, to ensure a long slow fermentation which finishes and can be bottled while still sweet and without any fear of excessive re-fermentation later.


Traditional keeving

The traditional process begins with a blend of late season fully ripened mostly bittersweet fruit, taken from mature orchards which are naturally low in nutrients but fairly high in tannin. Dessert fruit is much less likely to be successful here, due to its generally low tannin and high nutrient levels. The sugar level should be at least 12% (SG 1.055). The fruit is stored until a cold day late in the year when the temperature is about 5 °C and expected to remain so for a week or more. The fruit is washed and milled in the normal way, but the pulp is then packed into barrels to stand for up to 24 hours. This is the procedure of 'maceration' or 'cuvage', terms with no particular English equivalent. During this time, oxidation slowly proceeds which develops some juice colour, but more especially pectin leaches out of the apple cell walls into the juice. The juice is then pressed out, rich in colour and thick in texture, and is run into clean barrels (previously sulphited by burning 10 grams of ‘sulphurated string’  inside them!).  No yeast is added.

Since the temperature is low, no significant yeast fermentation takes place in the first few days, but the natural pectin esterase enzymes in the apple juice slowly change the pectin to pectic acid. This forms a gel as it cross-links with the natural calcium in the juice and a 'brown cap' (the 'chapeau brun') rises slowly to the surface, as it picks up gas bubbles from the early stages of a very slow fermentation conducted by yeasts trapped in the gel. In 17th century English writings this cap was picturesquely known as the ‘flying lees’! Some of the pectin also combines with juice protein and tannin and falls as a sediment to the bottom, leaving a clear juice between the two. In French this process is known as ‘défécation’! If excessive yeast growth begins too soon, the ‘chapeau brun’ is replaced by the normal ‘chapeau blanc’ and turbulent fermentation take over. In this case keeving has failed and the fermentation is allowed to proceed in the normal way. With some varieties of apples, no head forms at all and so they cannot be keeved.

If the keeving has been successful, however, the clear juice between the top cap and the bottom sediment is very carefully pumped or syphoned into another pre-sulphited fermentation vat. It is now allowed to ferment under an air-lock in the normal way (with its own yeast), but this fermentation will be very slow because most of the nutrients in the juice will have been left behind in the cap and in the sediment. Because the pectin is negatively charged, but the nutrients such as thiamine and asparagine (the major amino acid in apple) are positively charged, the nutrients are attracted to the pectin gel and hence removed from the system. Scientific studies have shown that the pectin and the amino nitrogen nutrients are reduced by at least 50% during keeving. The number of yeast cells is also substantially lowered by entrapment.  With the slow resulting fermentation it should be no problem to make a naturally sweet cider, by racking initially at S.G. 1.030 and racking again later to ensure the the fermentation slows down to an almost imperceptible crawl (less than one degree loss per week)Keeve in jar. Eventually the ciders are bottled on a cool high-pressure day to retain the maximum gas and minimum suspended yeast, and they slowly continue to develop ‘condition’ in the bottle.

Keeving brought up to date

The advantage of this process is that it can produce a naturally sweet and well-coloured cider, brilliantly clear due to the removal of pectin during keeving, and full of flavour because of the low nutrient levels during fermentation. The disadvantage is that a lot of it depends on luck - the correct fruit, cold weather, benevolent strains of wild yeast and freedom from bacterial infections! However, it has been made much more controllable in recent years by some key items of technology. On a commercial scale in France, for instance, the vats are refrigerated at 4° C, the chapeau is encouraged to rise by bubbling nitrogen gas through the system, and centrifugation is used to remove excess yeast if the fermentation should develop too quickly. These technologies are not available to the amateur or small producer, but there are two important parts of the system which are - calcium and pectin methyl esterase (PME) enzyme.

Calcium is required to form the pectate gel.  Depending on where the apples are grown, there may or may not be sufficient calcium naturally present in the juice, but it can be added in the form of a calcium salt and this makes the keeve much more reliable and gives a thicker denser cap. In the past a mixture of calcium carbonate (chalk) and common salt (sodium chloride) was used, the chloride helping to moderate yeast growth. Nowadays it is possible to buy food grade calcium chloride which can be added instead and does the same job more conveniently (and without raising sodium levels!). The maximum dosage used is 400 parts per million or 4 grams per 10 litres, which is stirred into the juice immediately after pressing.

The PME is required to de-esterify the pectin so that a gel can form. Apples themselves contain natural but variable amounts of PME and until recent years this has been a very considerable constraint on the process. In the last decade or so, however, specialist PME preparations have appeared on the market, which also have other applications in fruit processing. Brand names that were available in late 2008 are Rapidase CME (DSM), Novoshape (Novo) and Crystalzyme AES (Valley Research Enzymes USA).  Not all these are readily obtainable in small quantities, but for cidermakers in the UK and Europe the  enzyme plus calcium chloride are available as a French kit sold under the name ‘Klercidre’. This is undoubtedly the easiest way to get started with keeving. If you are trying to source the enzyme yourself, it is very important to stress that these PME enzymes are not at all the same as the commonly available pectic enzymes used for clarifying purposes. These PME preparations are very specialist and in fact must not contain any trace of the other polygalacturonase or ‘depolymerising’ activities which are used for juice clarification. The pectin must only be de-esterified; if it is de-polymerised too, then the gel will never form.  It is also important to stress that the PME and calcium chloride must never be added together but only as two separate sequential operations. PME dosage depends on enzyme strength which varies by manufacturer. Retail suppliers currently known to me are given at the bottom of this page.

Availability of these two components have helped to take keeving from the realms of art closer to that of a science (and if you cannot get hold of the commercial enzyme then even the addition of calcium chloride is a considerable help) but it is still not a ‘recipe book’ operation. One important step is the ‘maceration’ of the pulp, that is, leaving it for 24 hours. A major purpose of this is to solubilise as much pectin as possible – it is important to have enough pectin to form a good gel and by leaving the pulp overnight it has the chance to solubilise and leach out of the cell walls. There is some evidence that it is useful to add some of the PME enzyme at this stage too, but pectic enzymes are also inhibited by the pulp tannin so it is best to reserve some for addition to the juice also in the normal way.

In traditional keeving, the PME did not work well in acidic (low pH fruit) and hence bittersweets were the fruit of choice; bittersharps like Kingston Black would not generally keeve. This was obviously something of a drawback since it left the fermentation poorly protected against microbial infection at such a high pH. Fortunately, conventional wisdom has now been turned on its head with the new enzymes which are active at much lower pH, and I and other craft cidermakers have carried out successful keeves with blended juices including bittersharps down to pH 3.6 and even below. This is a considerable advance and allows keeving to be much more widely practiced for well balanced juice blends, which was frankly tricky before.

chapeau brun

It will be noted that a wild yeast fermentation is required and that little sulphite is used (in the traditional scheme, the barrels were pre-sulphured). Cultured yeasts obviously cannot be used here since they are far too vigorous, indeed it is interesting that the French industry is still very keen on slow mixed microflora fermentations. On the other hand, incipient fermentation is required to start within a few days since the cap must be buoyed up by trapped gas bubbles. In practice I have found that by applying sulphite at one- half or one-quarter the recommended level for the pH (see table), sufficient wild yeast growth occurs to raise the cap in a  reasonable time (about a week). But there is also a case to be made for no addition at all before keeving. This is to encourage the apiculate yeasts to get going quickly to raise the cap. You can then add a half dose of sulphite after the juice is removed from the cap.

Racking the clear juice from between the chapeau and the sediment can be something of a challenge, but is made easier by translucent HDPE tanks with bottom taps and a small pump. In traditional designs for French cider factories the keeving tanks were set up above ground level, so that the keeved juice could be gently drawn off by gravity from below the ‘chapeau’.  This operation should not be rushed, and the potentially fragile cap should not be allowed to break up. The cap can vary in firmness and crustiness quite considerably and not all the keeved juice will be easily recovered. Sometimes it is possible to let the cap re-form and take a second crop of juice after a couple of days. But some irretrievable volume loss (maybe around 20%) must be expected during keeving.  It is also critical to watch the process daily and to rack off just before the cap begins to break up and to fall back into the vat. If you are a hobby cidermaker and this point is reached on a Tuesday, it may be too late if you wait until you have more time at the weekend. By then, all may be lost!

Once the fermentation is under way, it needs monitoring and then some sequential racking to slow it right down, with a view to bottling at say SG 1.015 or 1.010. There is some interesting recent French work which shows that a first racking once the SG has dropped by 10 points is the optimum time for initial yeast crop removal.  Traditionally, a keeve might be set up in November and the bottling carried out in April. It will then take several months to develop ‘condition’ in bottle. You must be sure that the fermentation is imperceptibly slow by then (a maximum loss of 2 degrees SG over 3 weeks) and champagne bottles are recommended in case of accidents, because in theory you may be over the limit at which the bottles could burst should all the sugar re-ferment. In France, keeved ciders are bottled in champagne bottles with slightly undersize special ‘cidre bouché’ corks (25 x 38 mm) and stored upright. Then, if excess gas should develop, it can escape slowly past the cork rather than blow the bottle. The bottles should be stored in a cool place (17th century writers refer to them being stored in cool running spring water to avoid them becoming ‘potgun’ cider)

Anyone trying the art of keevng should realise they are experimentalists and are pushing the boat out. It is not a recipe based operation (do this, do that and do the next thing and you will get a perfect result!). There are lots of uncontrolled variables and until you have built up some experience you won't know what works best for you. I say all this because I don't want people to be disappointed or to moan to me because "it didn't do what it says on the tin"! There are many factors to balance for a successful keeved cider and it is not for the faint-hearted. Balancing all these factors relies on judgement and experience which you can build up over several seasons. This was once part of the skill of the traditional cider maker, although he knew absolutely nothing of the biochemistry behind it as we do now.  But there are many enthusiastic amateurs rediscovering the finer points of this technique and, with an increasing focus and understanding of the science and availability of the PME enzyme, it seems to be set for something of a revival in the craft cidermaking community.  


The Klercidre kit which contains both PME and calcium chloride, is sold retail in French agricultural suppliers such as Ruaux and Agrial I understand, if you happen to make a trip to Normandy.  You may be able to negotiate with Standa to buy it direct if you live elsewhere.  However, from autumn 2013, the Klercidre enzyme together with food grade calcium chloride is being sold in the UK by Vigo

In the US and Canada you can buy supplies of PME and calcium chloride from here  If you live in Europe or Australia / New Zealand you can also buy direct from Cider Supply but it may be cheaper on carriage to source the calcium choride locally from a home brew supplier.

You can buy the PME enzyme as 'Novoshape'  in retail quantities (sold  for fruit firming) from the Modernist Pantry in the USA. They will also supply by mail order overseas. You will need to source the calcium chloride locally.  I would suggest an enzyme dosage rate for 'Novoshape' of about 20 ml per 100 litres of juice. (Novoshape seems not as potent as other PME's on the market. It works fine; you just need more of it).

Further Information

Last updated 6th September 2013

Return to Contents Page