Technical Profile: Starch Damage and British Baking Practices

by Teresa Acklin
Share This:

   Contributed by suppliers, technical profiles feature new technology, products, specific applications or proprietary concepts. This material was presented by R.N. Butler, Satake U.K. Ltd., Stockport, U.K., at a 1996 conference in France.

   It is perfectly reasonable that a conference section concerning starch damage should emphasize British practices. After all, it is our particular baking process, which is almost unique to the United Kingdom, that determines our millers' requirement for starch damage.

   It needs to be made clear what we mean by starch damage. In the U.K., millers and bakers nearly always measure starch damage by the increased susceptibility of starch granules to degradation by amylolytic enzymes.

   The Farrand method is normally employed for this measurement.1 Caution must be exercised in converting Farrand units to other units of measurement based on the increased extractability of amylase from damaged starch granules.

British Baking Practice

   Most large bakers use a no-time dough process of the Chorleywood Bread Process (C.B.P.) type, in which intensive mechanical work input to the dough during mixing replaces much of the action of yeast in the traditional bulk fermentation baking process.

   Because of the reduced time for diastatic activity, less water is released, allowing additional water to be added to the recipe to achieve the correct consistency for dough handling. This can amount to as much as 4% to 5% extra water. The intense action of a C.B.P. mixer produces near optimum formation of the gluten matrix and in fact permits the use of somewhat lower protein flours than would be the case with bulk fermentation.

   In general, the lower protein wheats used to produce this flour will be softer and hence, all other things being equal, will produce a flour with lower levels of damaged starch and hence lower water absorption. In order to restore the balance, we grind harder on the head reductions.

   The enhanced levels of starch damage produced can be tolerated because of the short time available for enzymicaction, as already explained. The production of excess sugars can be controlled by adjusting the levels of fungal amylase used. Of course, this is all possible in the U.K. because bread is sold there by net weight and not by dry solids content.

   Eric Farrand recognized early in his work that to make a stable dough and a loaf of good volume and texture, the gluten must spread far enough to contain the total surface of the starch. There is a limit to how far the gluten will spread. The potential to damage starch and increase water absorption will therefore be limited by protein content.

   Farrand's formula2 for optimal balance was

Starch Damage=Protein2

   The factor 6 is nowadays considered rather high and a factor of 5.2 is more usual. Thus, if we were using a hard wheat grist of 12.5% protein, optimal starch damage would be:


   In this same study, Farrand derived a relationship between optimum water absorption, protein, starch damage and moisture content:

WA = 1.4P + 0.38SD - {1.6M + 0.004SD (M + P)} + 12 {6SD - 1} + 57.3
P{P2 }

   In this context, water absorption is defined as that measured by the Brabender Farinograph at the 600 BU line. This is not necessarily the actual percentage of water used by the baker.

   Today, several formulas exist in the U.K. industry relating water absorption to starch damage, protein, moisture content and in some cases amylase activity. Some are more sophisticated than others, but they follow a similar basic pattern

% WA=F1P + F2SD + F3M + F4
where F1, F2, F3 and F4 are factors determined by the bakers from sustained research and development. If we take possible factors of
then for the 12.5% grist
WA=(1.4 x 12.5) + (0.25 x 30) + 37
=17.5 + 7.5 + 37

Why Damaged Starch

    Are we looking for starch damage and water absorption just to get maximum water into a bag of flour, or are we looking for a well balanced dough and a loaf of excellent volume and texture? If we start looking for a similar water absorption from European wheats of say, 10.5% protein, then we run the risk, which I regret to say we see too often in practice, of excessive starch damage.

IfWA=1.4P + 0.25SD + 37
and62=14.7 + 0.25SD + 37

but SD should equal10.52=21.2

   Therefore, we have excess starch damage, total lack of balance between gluten and starch surface area, a bad loaf and short shelf life.

   Is this British practice? Not always, of course, but it does exist, and the risk is heightened in countries with lower proteins for the same wheat varieties.

   If one has a wheat protein of 9%, starch damage should not exceed 16, but I suspect many may be attaining 20 (water absorption would then be about 55%). They are probably grinding too hard. But with exceptionally low roll surfaces, they probably have to, in order to achieve extraction.

Changing Current Practice

   At the beginning of this paper, I commented on the current basic practice of grinding harder on lower protein soft wheats to enhance starch damage and water absorption, and I have just mentioned the inherent risks.

   It is essential that we establish milling technology and practice that give optimum control of the factors that can affect starch damage. It is our belief at Satake that our unique PeriTec system, a fundamental change to the milling process in which the bran is removed from the endosperm before milling, responds with many other advantages to the control criteria required in the context of starch damage.

   First, there is the question of controlling alpha amylase. We have already seen that in C.B.P. we can use greater amounts of fungal alpha amylase. This can lead to larger loaf volumes without the side effects of stickiness associated with high levels of cereal amylase.

   We have been fortunate in recent years with few parts of Europe experiencing adverse weather conditions at harvest. However, this will not always be so, and those of us who have been privileged to work in this industry for many years will well remember when wet harvests and low Hagberg Falling Numbers seemed to be the norm. Problems only could be avoided by strict checks of all deliveries at intake or by using gravity tables or other density separators to remove affected kernels.

   In conventional milling, some aleurone and sub-aleurone material is present in any straight run flour. Since most enzyme activity is initiated in the aleurone, this is where the greatest concentration of alpha amylase is to be found. Inevitably then, conventionally milled flour will be contaminated.

   In PeriTec, by contrast, it is possible to control the material removed from the grain. When necessary, this can be used to significantly increase the Falling Number of flour from low Hagberg wheats by as much as 60 seconds.

   Then there is the well-known grinding parameter of roll speed differential. Increasing the differential increases the starch damage.

   Unfortunately, this has a tendency to increase ash. We must therefore develop milling systems that provide the cleanest possible semolinas and middlings from the break system to provide low ash feeds to the head of the reduction system.

   By removing the bran before milling, a very high release of very clean large semolina is produced from the break system, providing remarkably clean feeds to the reduction system and a consequently very flat ash curve at high extraction.

   There also is the question of particle size. This is perhaps a contentious subject with diverse and opposing expert opinions. But it is perfectly reasonable to propose that we should be measuring constantly the total surface of the starch and relating this to the covering power of the gluten.

   Debranned wheat first breaks down into very clean large semolina. This is then far more easily and controllably reduced to a regular fine particle spectrum without excessive grinding and consequent excess starch damage and reduced baking quality. Correct fluting for sizing is essential, and the use of fluted head reductions can improve control, even with high loadings.

   Removing the bran before milling reduces the core milling machine content and space requirement and minimizes initial investment, and the simplicity of the flow reduces operating costs. These are important considerations for those who may suffer from the constraints and compromise created with excessively short roll surfaces on the head reduction passages, the key area in which grinding practice controls starch damage.

   By introducing a PeriTec debranning system in front of the existing mill, tail end break and reduction rolls can be liberated to supplement the head reduction passages and so provide the essential grinding surface and control.

   After debranning, the PeriTec hydration system permits the addition of up to 3% moisture to the debranned wheat with a very short absorption time of approximately one hour. This feature allows accurate control of finished flour moisture because variations can be identified and corrected in a very short time.

   The debranning process makes possible the controlled inclusion of aleurone material — uncontaminated by bran powder — that can remain in the flour to enhance baking quality and nutritional value. The removal of bran before the milling process significantly reduces contamination by bacteria, insect fragments and residual chemicals.

   These latter factors are directed at flour quality improvement. No less important in this challenge is the ability to obtain accurate control of starch damage, which should be directed at improved baking quality. This is why milling practice is changing, moving into the 21st century to give an adequate, economic and proper response to the quality needs of the baker to satisfy the preferences of the consumer.

   1Flour properties in relation to the modern bread processes in the U.K. with special reference to alpha amylase and starch damage, Farrand, E.A., Cereal Chem., 41, pp 98-111.

   2Starch damage and alpha amylase as bases for mathematical models relating to flour water absorption, Farrand, E.A., Cereal Chem., 46, pp 103-116.

   3Flour Characteristics and Fungal Alpha Amylase in C.B.P., Report 121, March 1985.