The latest generation of near-infrared systems for online measurements in grain, flour and semolina open up new possibilities regarding gluten, water absorption and starch damage. These allow millers to optimize flour production directly and individually.
Thomas ZIOLKO
Bühler AG

The grain processing industry has been using the near-infrared system (NIR) to continuously monitor the contents of raw, intermediate and end products for years. More and more companies are now using the advantages NIR technology offers for optimizing processes in real time. NIR analysis makes it possible to assure consistent product quality, and it makes a substantial contribution to the profitability of a mill.
NEW GENERATION OF NIR
The first generation of NIR equipment focused on analyzing protein content, moisture and ash. These parameters were and are the 'classics' of most online systems. In addition, however, there are other quality parameters that require compliance by the mills. For example, in wheat flour, the amount of gluten, the flour’s ability to absorb water or the extent to which the starch has been damaged are important factors. Older generations of online NIR systems were only able to determine these parameters with insufficient accuracy and reproducibility. By employing photodiode arrays (diode array, DA), as in Bühler's NIR Multi Online Analyzer MYRG, the latest generation of NIR systems offers new possibilities for such additional parameters and thus new potential for millers. These additional calibrations eliminate long periods of waiting for laboratory results and correctional measures can be performed without delay.
POSSIBILITIES FOR NIR
Near-infrared spectroscopy is very well suited for determining the dominant contents in organic materials such as food and feed products. But to do so, an NIR system needs not only good hardware in order to function reliably, but a good mathematical model (calibration) to determine the properties of an unknown product.
The most important and NIR-active biochemical components, such as water, starch, protein and fat, can be modeled (or calibrated) without difficulty. However, other physical or biochemical product properties can also be correlated with the NIR spectra as long as they are dependent on the combination of the dominant contents and other properties such as particle size distribution. An example of this is the ability of flour to absorb and retain water during the making of dough. The ability to absorb water depends on the volume and quality of the protein, the condition of the starch grains and the particle size distribution, among other things.
LIMITS OF NIR
Only those characteristics which actually leave information in the NIR spectra and correlate with them can be calibrated. An example of this is the so-called "sample" which is used to evaluate the activity of certain enzymes in the flour. NIR technology is only conditionally suited for lower ranges of concentration. It is technically impossible to calibrate enzymes for NIR because it takes only few ppm enzymes in the flour to become active.
For generating NIR calibrations, the breadth of data is decisive for accuracy and the sturdiness of the model. Models can only predict products if their characteristics have already been included in the model itself. On the one hand, it is essential to cover the entire range of features that are to be measured because the models are not allowed to extrapolate. On the other hand, any disturbance factors, such as specific product characteristics (particle size, temperatures, source, chemical composition), the instrumentation used and the surroundings, must also be taken into consideration along with the characteristics which are supposed to be measured. For the best calibration, several hundred samples can easily become necessary under these conditions.
PROCESS
In order to compare spectra and samples, data must be prepared and calculated using certain algorithms (chemometry). The preparation of the spectra data using various mathematical functions depends on the product itself as well as the hardware being used and is necessary for better separation of the interesting information (such as the protein contents) from the uninteresting information (such as particle size distribution). Various possibilities exist which fall under the skill set of chemometricians. The quantitative calibration models are usually calculated with the PLS (partial least squares) algorithm which searches for the largest differences in spectra and links these with the characteristics to be calibrated.
NIR SYSTEMS AT WORK
The accuracy of an online NIR measurement system is usually indicated by SEP (standard error of prediction). SEP is a random standard error which is found between the reference laboratory and the online measurement during at least 20 validation measurements. The random error in NIR (SEP) cannot be smaller than the random error of the lab (SEL) since the calibration is based on the data from the reference laboratory. For inhomogeneous samples, where taking a representative sampling is already a large problem, an online NIR measurement can be significantly more accurate simply because of the size of the sample volume.
NIR devices require constant adjusting. First, the hardware (i.e., the light source, measuring window) must be frequently checked, and secondly, the NIR calibrations themselves also need regular monitoring and expanding since the product can undergo a natural change in an unknown direction after a certain amount of time.
QUALITY PARAMETERS
GLUTEN
The protein content in wheat flour consists of 90 percent gluten (gluten protein). The important proteins in gluten are gliadin and glutenin in equal portions. Gluten is a more or less flexible-elastic substance which results when wheat flour dough is allowed to rise.
In other words, it is essentially soaked gliadin and glutenin. Since a higher protein content does not always mean a higher content of gluten, an NIR calibration of gluten can offer a high added value. The gluten contents of wheat flour and its texture are a decisive determinant of the dough's behavior during kneading and baking. In general: The higher the gluten content, the greater the water absorption, the gas-retention ability and the expected volume of baked product. Good gluten values: 30–34 %.
The reproducibility of lab measurements (Method ICC137/1) is 0.4. With the NIR Multi Online Analyzer MYRG, 0.7 is achieved.
STARCH DAMAGE
"Starch damage," from a scientific point of view, refers to mechanically deformed starch. Compared to intact starch, mechanically deformed starch can absorb five times more water. That makes it the most important factor in water absorption for flour and dough yield besides the protein content. Starch damage occurs during the various passages in the milling process. If the technologist knows the desired degree of starch damage he can adjust the grinding process according to expectations, for example, by dimensioning the roll lengths (the longer, the higher the starch damage) or the grinding pressure.
The reproducibility of lab measurements (Method AACC 76-33) is 0.7. With the NIR Multi Online Analyzer MYRG, 0.8 is achieved.
WATER ABSORPTION
Water absorption [%] is the amount of water which must be added to a flour in order to achieve a fixed dough consistency of 500 farinograph units (FU). For determining the water absorption capacity, the farinograph from the Brabender company is frequently used in the laboratory.
Industrial bakeries need raw materials of consistent quality so that the process doesn't need constant adjustment. This includes the water absorption capacity of the flour. The water absorption capacity is important for proper dough preparation and controlling the dough during the rising and baking process, among other things. So naturally, the amount of water which is added in the baking process will depend on the water absorption capacity. There are possible corrective measures that a miller can undertake: Adjusting the grinding process, performing various types of conditioning, or adding attrition flour when the water absorption is too low.
The reproducibility of lab measurements (Method ICC115/1) is 0.8. With the NIR Multi Online Analyzer MYRG, 1.0 is achieved.
ACCURACY
The following table summarizes the accuracy of the various calibration systems which are available with the Bühler NIR Multi Online Analyzer MYRG. A comparison of the SEP values with the accuracies of the lab methods shows that the newest generation of NIR spectrometers can determine additional parameters in continuous production with amazing accuracy.
Wheat Flour
Parameters |
Reference Method |
Range |
Target SEP |
Moisture |
TGA 701 @130°C |
7 – 16 % |
0.20 % |
Protein |
Dumas / Kjeldahl |
8 – 23 % dm |
0.25 % |
Ash |
TGA 701 @900°C |
0.3 – 0.9 % dm |
0.03 % |
0.9 – 2.5 % dm |
0.05 % |
Wet gluten |
ICC 137 |
18 – 47 % mb 14 % |
0.7 % |
Water absorption |
Brabender Farinogram |
47 – 85 % mb 14 % |
1.0 % |
Starch damage |
Chopin SDmatic
AACC 76-33 |
5 – 31 UCD |
1.0 UCD |
Sandstedt & Mattern
AACC 76-30 |
3.5 – 17 % |
0.8 % |
|