Belgian Candi Sugar Part III

Yesterday, Brulosophy published a post where they compared "my" candi sugar recipe side-by-side with a commercial candi syrup. The much more rigorous testing conducted by Brulosophy mirrored my own less scientific experience - i.e. namely that in beer the differences between them are hard to detect, but there are differences. The ability to find these differences appears to vary between people, as in the Brulosophy the majority were not able to tell the differences, but the minority that could tell the difference could do so consistently.

Since my previous posts on the topic (1, 2 and "3", #2 is the primary resource) I've refined my method further. In the discussions following the Brulosophy post its become apparent that I should share these changes as a fair number of people are using my old posts as a starting point in their own sugar experiments (hello, Reddit homebrewers).

The changes I have made seem to address the issues others (and I) have noted - namely an occasional acrid/burnt character. An issue was also brought up by one commenter which I think is worth addressing here.

The changes I've made to my method:

1. I've greatly reduced the amount of DME used, as the amount of protein in previous batches was excessive. For 1 kg of sugar (2.2 lbs), I am currently using 5 ml (~1 tsp) of DME. Previously I was using 1 tbs (~15 ml)
2. I avoid mixing the sugar as much as possible - I mix to dissolve the sugar into water, and I mix when adding the lye, but I do not otherwise mix.
3. I am much more careful and slow with my temperature changes. Most of the mixing I did previously was to add cold water to cool the sugar if I overshot the desired temperature.
4. I now usually add corn sugar (fructose) at a rate of 1% volume/mass (i.e. 1 ml corn sugar per 100g sugar). This does not change the flavour of the final candi, but does reduce crystallization. It is easier to then blend the mix into a syrup or cast rocks with the non-crystallized sugar.

These changes have led to a candi which is much closer (to my palate) to commercial candis, one without the unpleasant flavours some of the previous batches had. Others on the thread have mentioned using pressure cookers and other methods with great success. Hopefully, as a collective we will be able to formulate a better method of producing a consistent and flavourful candi sugar for home brewing.

The "Issue":

An issue brought up the commenter 'Chino' in the Brulosophy thread was that the 30 minute inversion time that I recommend is insufficient to completely invert the sugar, with individuals over a Reddit working on ways to get improved inversions. I partially agree with what Chino states - given the rate of the reaction and the fact that it is an equilibrium reaction, a 30 minute inversion period without the addition of something to accelerate the process (e.g. acid) will only invert 8-10% of the sugar. Where I disagree with Chino is that I don't think this matters. Mallard products comprise a pretty small portion of the final sugar - assuming 100% of the protein added via the DME is converted to Mallard products, the Mallard products would comprise about 0.07% of the final candi by weight. Although multi-step reactions, the formation of most Mallard products requires only one sugar molecule per amino acid, meaning that you need "only" 0.07% inverted sugar to be able to (in theory) produce the full array of Mallard products. The 8-10% inverted sugar is a huge excess compared to what is required - this does offer an advantage in terms of reaction rates, but its hard to imagine that increasing inversion to 25% (theoretical maximum using heat alone) or 50-75% (theoretical maximum using acid) would offer further improvements.