The aim of a scale-up of spice mixtures is to ensure that the homogeneity achieved in the laboratory or pilot plant, the mixing times and the gentle treatment of the ingredients are also maintained with considerably larger quantities of bulk material, larger plants and different process environments.

Spice mixtures are particularly demanding here, because they contain components with widely differing particle sizes and bulk densities. Examples are coarse pepper and fine paprika powder, heavy salt and light herbs. Added to this are differing flow properties and cohesion, micro-components used in very small quantities such as flavourings or vitamins, and constituents sensitive to moisture and fat such as aromatic and essential oils. A purely volumetric 1:1 scale-up here frequently leads to deviating sensory properties and must be replaced by systematic scale-up planning.

Before every scale-up, the bulk material properties of the individual components and of the final mixture are characterised. This includes, for example, the particle size distribution, bulk and tapped density, angle of repose and wall friction, residual moisture and water activity, fat or oil content and the tendency to form deposits. In parallel, the established mixing process is documented in the pilot plant, since it forms the reference for the production scale. Mixer type, fill level, speed, mixing time, energy input and order of addition are documented, among other things.

Three similarity criteria are decisive for transferable mixing behaviour:

• geometric similarity (the same design principle, comparable length-to-diameter ratios, similar tool geometry)
• kinematic similarity (retention of the movement characteristics of the mixing bed, for example via a constant Froude number describing the ratio of centrifugal to gravitational forces)
• dynamic similarity (comparable loading and energy input per particle, in order to avoid breakage, attrition and loss of aroma)

The speed in the production mixer must be adapted to the larger radius, so that the mixing regime is maintained and the specific energy input lies within the same range. The Froude number plays a role here only for mixing systems with a horizontally mounted mixing tool.

In the scale-up, the production mixer should be as geometrically similar as possible to the pilot plant unit, ideally with the same mixing principle in a larger size. The following proven mixer types have become established for spice mixtures:

• vertical helical mixers for three-dimensional mixing at a low circumferential speed,
• conical mixers with a central mixing shaft for a wide fill-level range,
• ploughshare or paddle mixers for high mixing quality with large differences in density,
• conical screw mixers with a planetary orbiting screw for particularly gentle treatment of sensitive herb leaves,
• ribbon blenders for large batches with longer mixing times, and
• twin-shaft mixers for short mixing times at high batch sizes.

Overfilled mixers create dead zones and lengthen the mixing time, while fill levels that are too low reduce the effectiveness of the cutting rotor, which can have a negative effect on homogeneity. At the same time, the dynamic pressure and the surface-to-volume ratio change at large scale, which influences the friction conditions and the formation of deposits, particularly with oil-containing mixtures.

Sensitive components such as herb leaves or freeze-dried pieces of vegetable and fruit react sensitively to high mechanical loading. If the circumferential speed increases too sharply during scale-up, attrition and breakage increase.

Mixing quality should be recorded quantitatively, typically via the coefficient of variation of a lead or tracer component. In the pilot plant, samples are taken at defined points in time in order to obtain a mixing time versus mixing quality curve and to determine the optimum mixing point. At production scale, multi-point sampling from different zones of the mixer is additionally carried out in order to obtain representative data on homogeneity and to avoid overmixing. Mixing times that are too long can in turn favour segregation through demixing or centrifugal forces.

Recipe management, the order of addition and the dosing strategy are equally important. Micro-components are frequently prediluted in premixes with part of a main component, in order to avoid dosing errors and local over-concentrations. The order of addition is optimised so that very heavy or very fine components do not separate at an early stage. Oil and flavouring components are preferably dosed via spray or injection systems, so that droplet size, wetting and lump formation can be controlled. At the same time the shear input is managed deliberately, in order to achieve good distribution while preserving the structure.

Many quality deviations arise only in the downstream process steps. Drop heights should therefore be minimised and conveying paths kept as short as possible. Conveying principles should also be chosen that do not favour segregation. Silo and hopper discharge should ideally be designed for mass flow, since air streams can transport fine particles selectively and thus disturb homogeneity.

At production scale, the specific energy input must also be monitored. An increased energy input leads to warming and can impair aromatic oils and volatile components. Condensation can occur in the packaged final product.

The core principles can be summarised as follows:

• choose a geometrically similar mixer type from pilot plant to production,
• maintain the mixing regime as far as possible via suitable characteristic numbers,
• keep the fill level constant,
• demonstrate the mixing quality quantitatively,
• ensure product protection through a limited circumferential speed and controlled energy input,
• introduce micro-components and liquid additives reliably,
• examine downstream processes for segregation risks,
• secure the scale-up through several validation stages.