Wheat milling at the Grain Research Laboratory

Milling for quality evaluation of new wheat lines

Marta Izydorczyk, PhD, Jerry Kletke, and Dave Turnock

The Grain Research Laboratory uses the Allis-Chalmers Laboratory Mill for milling new wheat lines tested in the Western Canadian wheat breeding programs. The Allis-Chalmers mill is a batch mill with stock being conveyed manually (photograph 1).

Allis-Chalmers Laboratory Mill. Described below.
Photograph 1. The Allis-Chalmers Laboratory Mill

The milling flow consists of three corrugated roll break (B) passages, three sizing (S) passages, four smooth roll reduction (M) passages, and 2 bran finisher passages. The sifters used in conjunction with the Allis-Chalmers mill were designed at the GRL (photograph 2) to minimize sample loss during sifting and to reduce differences in results between operators.

Sifter units. Described below.
Photograph 2. The sifter units

The sifters operate in a side-to-side mode that carries the stock in a ziz-zag pattern down the surface of the sloped sieves. The sifters are equipped with light weight chains and air nozzles that assist in stirring the stock while cleaning the surface. Compared to the Buhler Laboratory Flour Mill, which has six grinding passages in total (three fluted break passages and three smooth roll reduction passages), the Allis-Chalmers Mill allows for more extensive stock separation and produces more flour streams. One of the advantages of the Allis-Chalmers Laboratory Mill is a high sample throughput as several samples can be milled simultaneously. The Buhler mill allows only one sample to be milled at a time, and while the sample throughput is fast, cleaning is time consuming. Another advantage of the Allis system is that milling flows can be easily adjusted (e.g., adding a break passage or reduction passage), which cannot be easily done with the Buhler system.

For the evaluation of milling and functional qualities of late generation lines in the Western Canadian wheat breeding programs, the extraction rate (flour yield) is kept constant for all lines within a particular wheat trial. The exact extraction rate depends on the wheat class and on the specific crop year, and it is determined after all lines are milled once. For hard wheat, the extraction rate is usually between 73-75%, whereas for soft wheat 72-74%. The constant extraction rate is used so that samples can be compared on an equivalent basis, in particular for ash content. A sample with lower ash content than the check varieties, prepared at the same extraction rate, indicates that a miller could extract more flour compared with the current varieties while meeting customer specifications for ash. In contrast, if a sample has higher ash content than the check varieties at equivalent extraction rate then it would have to be milled to a lower extraction rate in order to produce flour with equivalent ash content.

Handling and tempering wheat before milling

  1. Wheat samples are cleaned using a Carter dockage tester to remove impurities. After cleaning, a test weight is performed, and the grain is cleaned again using a Forster Cyclone Grain scourer to further remove any dust and damaged kernels.
  2. Wheat moisture is determined with a Model 919 dielectric moisture meter prior to tempering.
  3. The samples are tempered to different moisture levels depending on the wheat class; e.g., Canada Western Red Spring wheat is tempered to 16.3% moisture, Canada Western Amber Durum wheat to 16.0% moisture, Canada Western Red Winter to 15.5% moisture, Canada Western Soft White wheat to 14.5% moisture.
  4. Depending on the moisture content of wheat samples, an appropriate amount of water is added to the sample or the temper is adjusted by drying with forced air if the moisture content is too high. Reducing wheat moisture by forced air is a rare event with breeding material which is usually very dry (after drying with forced air it is important to equilibrate the wheat in an airtight container to make sure the bran is not so dry as to fragment during milling). If the moisture content of grain is below 10%, wheat is tempered in two stages: initially to ~13% and then to the desired level. Wheat samples are placed in 6 L plastic containers with airtight lids. The samples are shaken by hand after the initial addition of water and then inverted intermittently by a mechanical mixer for 6 hours. Generally wheat samples are tempered overnight for at least 18 hours. The milling performance does not change if the total tempering time ranges from 12 to 48 hours, therefore, the samples can be milled over several hours without fear that the milling results will change from the first to the last sample.
  5. After tempering, the moisture content of wheat samples is determined again to confirm the desired level.


  1. Milling is performed using the Allis-Chalmers Laboratory Mill in a temperature and humidity-controlled room: 21°C and 60% humidity to assure repeatable results, regardless of the ambient conditions.
  2. The routine Grain Research Laboratory common wheat mill flow is illustrated in Figure 1.
  3. The B3/S1, S2/S3 and M1 to M4 rolls are warmed up to 37.5-39°C (overnight using light bulbs) to avoid any mill warm up time and to ensure constant temperature of the rolls during milling. Variations in grinding roll temperature will influence milling results, since rolls expand as they warm up, decreasing space between rolls. Decrease in roll spacing increases grinding action and may affect the bran and flour yield and the level of starch damage. Each individual roll stand has a fixed roll gap that remains constant for milling of all samples.
  4. After verifying the room temperature and humidity, roll temperature, roll gaps, and the vibratory feed settings, the mill is ready for operation. The tempered wheat is weighed to the appropriate amount on a 14% moisture basis (e.g., 1.028 kg at 16.3% moisture equals 1 kg at 14.0% moisture).
  5. A 2 kg check sample is milled on a daily basis at various times of the day. The following parameters for the mill check are determined and documented to ensure proper mill performance and reproducibility of results:
    • Flour yield
    • Flour moisture
    • Flour colour
    • Farinograph absorption
    • Flour ash
  6. It takes approx. 1.5 hours to mill a 2 kg sample, but several samples can be in the mill simultaneously at various stages of milling, thereby increasing the sample throughput without having to calibrate several mills against each other.
  7. Wheat lines tested in the 1st year breeding trials are milled once (3 kg portion), whereas samples in the 2nd and 3rd year trials are milled in duplicate (2 x 2 kg portion). Duplicate samples are milled on different days and at a different time of the day.
  8. Milling generates several flour streams. The B1-B3, S1-S3, M1, M2 flour streams and the bran flour (BF) are combined immediately and kept separate from the M3 and M4 flour streams. The moisture content of flour streams is determined using the Brabender moisture oven.
  9. The exact extraction rate (flour yield) depends on the wheat class and the specific crop year, and it is determined after all lines are milled once. The extraction rate is kept constant for all lines within a particular wheat trial. Taking into consideration the moisture content of the flour streams (B1-M2 and bran flour, M3 flour, and M4 flour), the desired extraction rate is reached by blending the streams in appropriate proportions. Flour streams are combined in a tumbler and mixed for 10 minutes (photograph 3).
  10. Flour yield (extraction rate) is expressed as % of clean wheat on a constant moisture basis (14%).
Flour tumbler. Described in text.
Photograph 3. Flour tumbler for mixing flour streams
Illustration of the durum wheat mill flow in the GRL.
Figure 1. Illustration of the Allis-Chalmers laboratory mill flow for durum wheat semolina incorporating modified GRL sifter unit
  • B = break rolls
  • M = middling (reduction) rolls
  • S = sizing rolls
  • BF = bran finisher, corrugations/centimeter
  • B1 - 6.3 / B2 - 7.9 / B3 - 9.5
  • All seives use nitex bolting cloth.

Reliability of laboratory milling of common wheat

To maintain a high degree of accuracy and precision of milling and test results, reliable laboratory milling procedures and tolerances for control samples have been established at the Grain Research Laboratory to produce milled products that are comparable and repeatable not only day-to-day but year-to-year and longer. Standard control samples ('mill check'), included with the test samples every day, provide constant verification of the mill performance and related test procedures. The 'mill check' samples are prepared in large quantities by blending No. 1 CWRS wheat from two crop years; the ‘mill checks’ are stored < 5°C.

Target values and acceptable tolerances for CWRS 'mill checks'
  • flour moisture: 15.0% ± 0.3%
  • flour yield: 75.9% ± 1.0%
  • Minolta colour: L* 94.0 ± 0.2
  • farinogragh absorption: 66.3% ± 0.5%
  • flour ash: 0.50% ± 0.03%

Experimental wheat samples are thoroughly cleaned, blended, and sub-sampled. Tempering moisture can strongly influence wheat milling properties, and improper wheat conditioning moisture will affect the flour yield, flour moisture, ash content and flour colour. If the tempered wheat is too dry, more of the bran will be ground finer, causing flour yield and ash to increase and colour to decrease. If the tempered wheat is too wet then more of the endosperm will adhere to the bran, causing a flour yield decrease. Proper tempering procedures, including determination of moisture in wheat before and after tempering, determination of moisture content in flour, and temper time and temperature, must be ensured. Variations in ambient temperature and relative humidity in the mill room can have marked effects on experimental milling results; for instance, flour quality as measured by ash and colour will deteriorate as the mill room relative humidity decreases. Milling at the Grain Research Laboratory is performed in a temperature and humidity-controlled mill room.

Standard procedures involved in preparation of the mill include pre-warming mill rolls overnight, checking the mill roll temperature, checking the proper roll gap setting and the vibratory feed settings. These practices guarantee the correct mill performance and achieving flour yields and flour farinograph absorption within the set tolerances. Proper maintenance of sifters to prevent leakage, tears, clogging and thorough cleaning of roll stands and sifters during the milling process will ensure correct flour yields and recovery, and prevent loss and/or contamination of samples.

Durum wheat milling

The experimental durum wheat lines tested in the Western Canadian wheat breeding programs are milled using the Allis-Chalmers Laboratory Mill (photograph 4) in conjunction with a laboratory purifier (photograph 5).

Allis-Chalmers Laboratory Mill for durum wheat.
Photograph 4. General view of the Allis-Chalmers Laboratory Mill for durum wheat
Laboratory purifier for durum semolina.
Photograph 5. General view of the purifier for durum semolina

The procedures for handling and tempering of durum wheat are similar to those used for common wheat. Durum wheat is tempered to 16% moisture before milling. Samples are milled according to the flow illustrated in Figure 2. The flow consists of four corrugated break roll passages, five corrugated sizing roll passages and ten purification steps. Roll gaps on the first three breaks are kept relatively wide to produce semolina with coarse granulation. The purification steps in durum milling ensure better particle size separation and bran removal. Semolina streams are collected after each purification step, whereas the coarser material retained on a 630 µm screen is passed through the sizing rolls and sifted on a box sifter. After sizing passages, streams with fine particles passing through the 180 µm sieve are collected as flour, whereas the coarser streams are passed through the purifier. Streams with particles retained on the 183 µm screens but passing through the 571 µm screens are collected as semolina. The distribution of particle size in semolina produced with our mill is shown in Figure 3.

Durum wheat lines being tested in the Western Canadian wheat breeding programs are milled in duplicate (2 x 2 kg portions) to a constant extraction rate of 65% semolina and 75% total yield (semolina and flour).

Illustration of the durum wheat mill flow in the GRL (65% extraction rate).
Figure 2. Illustration of the Allis-Chalmers laboratory mill flow for durum wheat semolina incorporating modified Grain Research Laboratory sifter unit (65% extraction rate)
  • B = break rolls
  • S = sizing rolls
  • P = purifier
  • SEMO = semolina
  • REB = stock to rebolted

Sieving times of break and sizing stocks for rebolting on the box sifter.

Graph of distribution of particle size in durum semolina
Figure 3. Distribution of particle size in durum semolina

Reliability of laboratory milling of durum wheat

Major sources of error in experimental milling of durum are similar to those involved in milling of common wheat and include: improper sampling, incomplete removal of foreign material, inappropriate tempering, variations in ambient conditions, poor conditions of milling equipment, and improper mill settings. Durum wheat semolina milling is particularly sensitive to fluctuation in relative humidity of the mill room because stocks are exposed to air during purification passages. Milling of durum wheat occurs in a humidity and temperature-controlled room (21°C and 60% humidity). The mill performance is verified by milling the standard ‘mill check’ sample and ensuring that the yield and the results for moisture, colour, and ash fall within the defined tolerances.

Target values and acceptable tolerances for durum wheat 'mill checks'
  • tempered wheat moisture: 16% ± 0.2%
  • semolina moisture: 15% ± 0.3%
  • semolina ash: 0.66% ± 0.02%
  • semolina yield: 66.7% ± 1.0%
  • total yield: 75.0% ± 1.0%
  • Minolta color: L* 86.5 ± 0.2


AACC International Methods 2012. Approved Methods. Experimental milling: introduction, equipment, sample preparation, and tempering. Method No. 26-10-02.

Black, H.C. 1966. Laboratory Purifier for durum semolina. Cereal Science Today 11:533.

Dexter, J.E. and Tipples, K.H. 1987. Wheat milling at the Grain Research Laboratory. Part 1. Objectives and the reliability of test results. Milling 180:22.

Dexter, J.E. and Tipples, K.H. 1987. Wheat milling at the Grain Research Laboratory. Part 2. Equipment and procedures. Milling 180:16.

Further information

Marta Izydorczyk, Ph.D.
Email: marta.izydorczyk@grainscanada.gc.ca

Jerry Kletke
Email: jerry.kletke@grainscanada.gc.ca

Dave Turnock
Email: dave.turnock@grainscanada.gc.ca