Recipe optimization at scale

Recipe Optimization at Scale

The first challenge in moving from lab or artisanal to industrial baking is adjusting recipes for large-scale production. Ingredients behave sometimes differently in larger quantities, and the texture, flavor, and appearance of products may shift.  Relative simple things as environmental and raw material temperatures might be different, mixing and kneading intensities are different and thus the dough might behave differently. Where gluten development is required (such as bread, puff pastry, certain crackers, etc.) this influences hydration levels and dough temperatures. Sometimes also these parameters have to be adjusted to fixed process parameters of a large line: it is challenging to have different fermentation times of products if they move in a continuous way compared to a batch method. This might even lead to small adjustments of ingredient proportions to facilitate these fixed timings (think of more or less yeast or yeast food).

In cakes a parameter might be the specific gravity or density of the batter. In the case that mixing might be done in let’s say 6 minutes in 2^nd gear (of 3), a larger 400 or 600 liter mixing bowl with 2 tools or even a continuous aerator will give completely different timings and intensities to reach to that value. Evaluating viscosity and consistency then firstly leads to optimizing speed and time, but sometimes also requires other adaptations such as the addition or removal of certain ingredients (such as starch or hydrocolloids) or the proportions within the recipe.

The better familiar you are with the conditions on the larger scaled production, the more you can try and solve for at lab scale: the degrees of freedom to adjust and adapt are there huge with 2 or 5 kg batches, compared to 400 kg or more.

It is always wise to compare end product to end product with a small trained team to ensure that product quality attributes and consistencies can be met; despite that in the processes steps had to be taken to deviate from what the expected norm was.

Recent posts