This extract from my book may help. You may need more than the two rackings f described. Claude’s book is also worth reading for hints on naturally sweet ciders.
Andrew
—————————
Traditionally, naturally sweet ciders were made from slow fermentations which were poor in nutrients. Such fermentations were not uncommon in the days when most cider fruit came from big old trees with low nutrient status. Those orchards were never fertilized and often had livestock running in them during the summer. As the animals grew they absorbed the nutrients from the grass and took them when they left, leaving none for the trees. Modern fruit from dedicated bush orchards has much higher nutrient status and so is generally unsuitable for making naturally sweet ciders. It is also worth noting that most ‘vintage’ cider apple varieties take up less nitrogen from the soil in any case, and therefore can be inherently slow fermenters if grown in orchards which have very little fertilizer input. So if you have an old orchard of your own, or your orchard is low in nutrients because of the trees you grow and the way you manage it, then you may be able to make naturally sweet ciders in the old way. This is also more likely to be successful if you use wild yeast fermentation rather than adding a vigorous cultured yeast. Juices which are fermenting inherently slowly and which show an SG loss of less than one degree per day are suitable for this treatment.
How to make it
The cider is racked initially into a new clean tank when it reaches SG 1.020, leaving most of the yeast behind. The airlock is re-fitted. The continued fermentation will then become even slower and the sweet cider is racked again (and preferably filtered) at SG 1.015–1.012 for a medium sweet cider. After this racking it is worth waiting several weeks (under an airlock) to ensure that no further fermentation takes place, before sealing the vat or bottling off. Where bottles are used, they should preferably be of the champagne type in case any re-fermentation does take place. If all that sugar were to re-ferment you would be over the safe limit for a ‘bottle bomb’. The factors that generally limit yeast growth are the lack of nutrients and the build-up of CO2 in a closed bottle (since excess CO2, although produced by the yeast, is also toxic to it to a certain extent). It is wise to make a test bottling after the second racking and lay the bottle down in a closed box for three weeks in a warm place (25°C). Then, wearing gloves and goggles, open the bottle and assess the level of carbonation. If it is obviously excessive, leave the bulk for a fortnight longer and re-test before doing the final bottling. If the carbonation is nil to slight, it will be safe to bottle the bulk. Either way, the bottles should be stored cool and in a place where they would cause no damage or injury should the worst happen. A carbonated drink at 30°C contains up to twice the internal pressure of one at 15°C as the gas is driven out of solution by the heat. It is best to choose days on which the temperature is low and the barometric pressure is high for the racking and bottling operations, since this will help to keep suspended yeast to a minimum and will retain the maximum amount of dissolved carbon dioxide in the cider. The success of the whole process depends on reducing both yeast and nutrient levels to a minimum so that re-fermentation of the remaining sugar is unlikely to take place. Sweet ciders of this sort may have a slight ‘prickle’ to them, particularly in bottle, since a very slow fermentation may continue to generate carbon dioxide until the increased CO2 level itself intoxicates the remaining already stressed yeast.
Naturally sweet
The procedure described is ideal for single-variety demonstration ciders or for those which need no further blending – the flavour tends to be ‘fruitier’ since the sweetness is derived from unfermented juice rather than from added sugar. The alcohol level in the cider is of course less than if it had been fermented to dryness because only a part of the sugar has been converted. It is abundantly clear from descriptions in the literature that many (perhaps most) bottled ciders were made in this way for several hundred years until the early twentieth century, and even as late as the 1960s in some cases, and were naturally sweet and sometimes slightly sparkling to some degree. These have almost entirely disappeared from memory in a couple of generations, in part due to the ‘law of unintended consequences’ following changes in apple orcharding practice and hence higher nutrient levels in the juice and cider. In addition, perceived customer demand for a totally consistent retail product led the large cider producers to embrace added sweetening, forced carbonation and pasteurization techniques from the early years of the twentieth century onwards.