Working with these Haiti beans has been a real learning experience. When assessing a bean I usually use a standard roast and make the chocolate at a standard 70% formula, a combination I have been using pretty much since day 1 of commercial chocolate making. But all that had to change when I started using the Haiti.

To do a test roast I use a normal kitchen convection oven, spread the beans out on thin oven trays one bean deep, around 500 grams. I preheat the oven to around 180 and put the beans in with a thermometer to watch the temperature.


That seems to have worked for the last couple of years, and still does. But what made this bean challenging is that the flavours we found when roasting in the small oven weren’t there when we put the beans through the big roaster, and each sample roast/batch came out different. We certainly found some lovely flavours in the samples but trying to replicate it on a larger scale was like herding cats. I had to be very methodical in my note taking and recording of the processing so I knew exactly where the changes were being created but that didn’t really help either. I started thinking this is simply one of those beans that has a mind of its own.

The whole process was taking ages because we found the chocolate was changing day by day after it was made, at the beginning of the week a sample tasted banana, cinnamon. By the end of the week the same sample had changed to coconut. Was it the chocolate or was it our palate that was changing?


Even the timing of the sugar was making a significant difference. If the sugar went in late the chocolate was mellow, if the sugar went in shortly after the nibs the chocolate seemed to retain more intensity. The amount of cocoa butter we used was fiddled with as well. The cocoa solids it the chocolate ended up being 67.5%, how ridiculous is that? We rounded it up on the package.


The main factor was definitely the roasting. Slight changes in the roasting were making significant changes in the chocolate. Simply warming up the roaster vs putting the beans in a cold roaster made all the difference. We got the best results from putting the beans in cold and using a very low flame to bring the temperature up slowly, this flies in the face of convention where beans require strong heat for a short time, mainly to vaporise any moisture in the bean and puff the shell away from the nib, we didn’t have that luxury with the Haiti, we were basically warming them up slowly, causing the water vapour to express gradually. We started calling the chocolate Voodoo Chocolate.


What I love about this is it has got me thinking how these changes to process might affect our other beans, I’ll be tinkering away with it in the coming months, seeing if I can nuance out some interesting flavours from our existing range. We also have Bolivia and Guatemala to test and one other exciting addition I won’t mention now.