Hi. Coming to this a bit late. there has been a lot of good sense written here, however I think it is important to consider the known information and the impact this has on cause analysis:
On this basis, there are the only possibilities I can see:
A
- Since the over carbonation applies to batches not individual bottles, the issue cannot be a sanitation problem in the bottles or inadequately mixed sugar in secondary as that would be variable bottle to bottle.
- The cold crash will increase carbonation, however unless fermenting under pressure and cold crashing from pressure, this will have only a minor impact, and would affect all your batches.
- The over carbonated beers were with the batches with comparatively less sugar added. While microgram scales are necessary for an individual bottle priming approach, when scaled to a whole batch normal kitchen scales will be adequately accurate to 1g scale, so this is unlikely to do with the amount of sugar added.
- in bottle carbonation only occurs with the additional production of CO2, either by bacteria or yeast.
On this basis, there are the only possibilities I can see:
- Bacterial contamination at the time of secondary mixing with the additional sugar leading to digestion of residual sugars in the beer in addition to those added at the time of bottling - this will be evident by a progressive increase in carbonation. Ie if left in a warm situation the bacteria may slowly progressively continue to digest the carbohydrates yeast cannot metabolise - otherwise known as a bottle bomb, some weeks after bottling.
- Sugars available to the yeast which had not been fermented out, which includes all the discussion above about temperature at the end of fermentation to ensure it is complete.
- That you are fermenting under pressure - which I've not seen reference to.
- That you miscalculated the scale up of sugar to volume not allowing for trub, ie added sugar on the volume in the fermenter rather than the volume being bottled - this would have to be way off though, which seems unlikely based on your approach.
A