Full power from flour

by Emily Wilson
Share This:

In one of the most in-depth studies of its kind, Crop & Food Research, an independent, New Zealand-based, government-owned research company, has concentrated on understanding how milling affects the nature and distribution of flour’s basic components within millstreams — especially carbohydrates and proteins — and the resulting chemical, structural and functional effects in flour.

While much is known about the effects of the gross variations in flour, little work had been done on variations at the molecular level. Results from this work will allow flour millers to more precisely tailor flour to specialized food or industrial products.

The "Millstreams" project will also provide greater knowledge for addressing flour production or quality concerns. Other spin-offs include finding new, higher-value uses for low-value byproducts, and increasing the output of higher-value products.

"Just as wheat grains are not uniform in shape and components are distributed unevenly within the individual wheat grains, the way the grain is broken, crushed and conveyed can have a significant effect on the chemical composition of the millstream," said Millstreams project leader, Lyall Simmons.

For example, in one sample, wheat with a total protein content of 16.4% gave millstream fractions varying between 11.85% to 29.85% in protein content. The levels of specific protein subgroups and other molecular flour constituents varied greatly between millstreams, even ones of broadly similar appearance.

"Storage proteins are strongly segregated by the milling process," Simmons said. "For example, in our research, soluble polymeric proteins increased from the first to fourth break, decreased for the initial reduction flours and slowly increased for the successive reduction flours." (See figure 1.) Similar trends were seen for other types of storage protein, total protein and for polysaccharides.

"Almost every component is segregated by the milling process," he said.

Flour composition and processing

As part of the Millstreams project research, basic flour chemistry measurements were compared to processing characteristics in four small-scale commercial process simulations — MDD baking (a ranking of volume/texture in a baked item), biscuit (or cookie) quality, pastry quality and pasting (viscosity). Simmons and his colleagues, Kevin Sutton, Dale Every, Jafar Al-Hakkak, Marcela Ross and Russell Sara, found strong relationships between the molecular composition of millstreams and flour processing qualities.

As expected, total protein was not a good indicator of MDD bake score, although specific protein sub-groups, such as the insoluble gluten proteins, are a better indicator as they are proportional to MDD bake score. An increase in the amount of insoluble pentosans tended to reduce bake score.

The quantity of albumin and globulin proteins related strongly to the flow of biscuit dough during baking. Pastry quality could not be linked to a specific molecular component, although it clearly varied between individual millstreams. (See figure 2.)

Milling of the samples was carried out at the pilot mill at BRI (Bread Research Institue) Australia in Sydney, using two tonnes of each of four different wheat varieties — two New Zealand and two Australian wheats — milled as pure cultivar lines.

Two cultivars, one Australian and one New Zealand, were low-protein, low-work input lines. The other two cultivars were high-protein, high-work input lines. (See table 1.)

Laboratory tests included moisture, color grade, falling number, farinograph, mixograph, extensograph, total protein, damaged starch, ash, enzyme activity, HPLC protein profile, and carbohydrate components.

The mixographs were found to give the best estimates of mixing behavior and could be correlated with soluble polymeric protein content.

Commercial applications

Simmons said millstreams research can have huge benefits for industry, especially firms with specialty products.

"In milling, there is still a lot of handed-down knowledge, but this work was looking for the molecular basis to this knowledge," he said. "We think we can now help millers achieve the best balance of components for a particular purpose.

"This approach allows you to look at the profile of the streams and to change the blends using a scientific base. You can optimize the material available. If there is variability occurring, you can look into why this is happening."

Analyzing a facility’s individual millstreams may also be an approach of interest to firms producing specialty components from flour or flour products.

"Companies can identify component-rich fractions to augment normal production," Simmons explained. "Using different streams or combinations may also allow the production of specialized products, such as specialized glutens with different characteristics."

Simmons said this type of dry fractionation could be a cheaper alternative to wet fractionation. High-cost wheat in the grist could be replaced by a strong stream from a utility wheat.

"Of course, all this depends on the product and how its milled, but all types of production situations could potentially benefit from this approach — and of course, implementation doesn’t require expenditure on new equipment."

A millstreams analysis by Crop & Food Research involves an initial consultation, with samples taken while the mill is operational, he said.


Proteins and carbohydrates are the main molecular components of flour. The flour proteins comprise 9% to 13% of flour weight and include albumins, globulins, prolamins, glutelins, gliadins and glutenins.
The carbohydrate fraction isn’t just starch; it includes fiber, pentosans, dextrins, sugars and non-starch polysaccharides.
The millstreams study investigated the composition and relative proportions of each individual class of protein and carbohydrate.

Table 1: Wheats of the millstream project

Work Input, Wh/Kg

Protein, %
















Georgina Hall is a journalist with Crop & Food Research, Lincoln, New Zealand.

For more information, E-mail Lyall Simmons at simmonsl@crop.cri.nz or Tim Lindley at lindleyt@crop.cri.nz.