I developed a web application to perform blend optimization calculations directly in your browser — no installation or server required. All computations run in-browser using a built-in LP and MIP solver.

The application covers two modules used in industrial flour milling:

• Silo Blend — find the minimum-cost or maximum-quantity blend of wheat or flour silos that satisfies protein, ash, moisture and other quality constraints.
• Flour Stream — allocate mill passage streams to multiple flour product types to maximise revenue while meeting quality and demand requirements per grade.

▶ Open Application

Wheat blending is one of the most critical operations in flour milling. Selecting and combining wheat varieties to achieve target flour quality parameters — at minimum cost — is a complex multi-variable optimization problem. This page summarizes research and practical advances in wheat blending optimization methods relevant to industrial flour milling.

Related publication: Alfin, F., 2019, Optimization of Wheat Blending, Miller Magazine, 13(111):70.  Article Download

Optimal wheat blend formulations must simultaneously satisfy multiple flour quality targets:

Linear Programming (LP)

The classical approach to wheat blending uses linear programming to minimize blend cost subject to quality constraints. Variables represent the proportion of each wheat variety in the blend. LP is efficient and well-suited to large-scale commercial mill operations with well-defined, linear quality relationships.

Nonlinear Programming

When blend properties exhibit nonlinear interactions — such as gluten quality responses or rheological parameters — nonlinear optimization models offer improved accuracy. These methods require more computational resources but better capture real-world blending behavior.

Artificial Neural Networks (ANN)

ANN-based models can learn complex, nonlinear relationships between blend composition and resulting flour quality from experimental data. They are particularly useful when analytical relationships are unknown or difficult to define, and have been applied to predict dough rheology from blend composition.

Response Surface Methodology (RSM)

RSM provides a structured approach to exploring the relationship between multiple blending variables and output quality parameters. It is widely used to identify optimal blend ratios through a reduced number of designed experiments.

Multi-objective Optimization

Modern blending problems often require simultaneously optimizing several competing objectives — minimizing cost while maximizing quality, or balancing multiple flour quality parameters. Pareto-based multi-objective methods produce a set of non-dominated solutions allowing millers to make informed trade-off decisions.

Successful wheat blending optimization in practice requires reliable analytical data for each incoming wheat lot, a well-maintained laboratory capable of measuring key quality parameters, and software tools that can handle the optimization calculations. Many modern mill management systems include built-in blending modules based on linear programming principles.

Regular calibration of blend models against actual production data is essential, as wheat quality can vary significantly between harvest seasons, growing regions, and storage conditions.

For further resources on this topic, see the Teaching page for lecture materials on modeling in food engineering, or the Publications page for related journal articles.

1. Annon, 2016, Buğday Paçallama ve Tavlama İšlemleri, Miller Magazine, Yıl 10, Sayı 82, sayfa 45.
2. Dündar, A. ve Zerenler, M., 2011, Bir Un Fabrikasında Hedef Programlama Uygulaması, Sosyal Ekonomik Araştırmalar Dergisi, 11(21), 73–94.
3. Elgün, A. ve Ertugay, Z., 2002, Tahıl İšleme Teknolojisi, Atatürk Üniversitesi Yayınları No. 718. Ziraat Fakültesi No. 297. Ders Kitapları Serisi No. 52, Erzurum.
4. Fowler, M., 2009, Blending for Value, World Grain, October, 62.
5. Fowler, M., 2012, Wheat Blending, World Grain, November, 94.
6. Haas, N. C., 2011, Optimizing Wheat Blends For Customer Value Creation: A Special Case Of Solvent Retention Capacity, Master of Agribusiness, Department of Agricultural Economics, Kansas State University, Manhattan, Kansas.
7. Hayta, M., and Çakmaklı, Ü., 2001, Optimization of Wheat Blending to Produce Bread Baking Flour, Journal of Food Process Engineering, 179–192.
8. Özkaya, H. ve Özkaya, B., 2005, Öğütme Teknolojisi, Gıda Teknolojisi Derneği Yayınları No. 30. Ankara.
9. Posner, E. S., and Hibbs, A. N., 2005, Wheat Flour Milling, 2nd Ed., American Association of Cereal Chemists, St. Paul, Minnesota.
10. Posner, E. S., 2009, Wheat Flour Milling, in Khan, K. and Shewry, P. R., Wheat Chemistry and Technology, 4th ed., American Association of Cereal Chemists, St. Paul, Minnesota.
11. Sugden, D., 1996, Wheat blending and mixing, World Grain, January 1, 37–38.
12. Tanase, T., 2014, Tahılı Öğütme İšlemine Hazırlama: Paçallama ve Tavlama, Miller Magazine, Yıl 8, Sayı 50, Şubat, Sayfa 44.
13. Türker, S. ve Elgün, A., 1997, Farklı iki Protein Düzeyine Sahip Bezostaya-1 ve Gerek-79 Buğdayları ile Optimum Ekmeklik Paçal Hazırlanması Üzerine Bir Araştırma, Gıda 22(1), 25–33.
14. Ünver, E., 2005, Buğday Politikamız ve Hububat Teknolojisi Üzerine Düşünceler 5. Depolama, Paçal ve İnce Temizleme İšlemleri, Un Mamuller Teknolojisi, Yıl 14, Sayı 70.