Example: OR in the Steel Industry
The steel industry provides a wide field of application concerning instruments for Operations Research. In cooperation with the AG der Dillinger Hüttenwerke (Dillinger Hütte), Europe's leading heavy plate producer, there are a lot of research projects in progress aiming to improve the production process by applying OR methods. In many production areas, such methods are already integrated as an inherent part of the production planning.
Production process and material flow
The following figure shows the production process and the material flow of the integrated iron and steel works of the AG der Dillinger Hüttenwerke along with the ROGESA Roheisengesellschaft mbH (ROGESA) and the Zentralkokerei Saar GmbH (ZKS) as sub companies (each to 50 %):
Production process and material flow of the integrated iron and steel works
of the AG der Dillinger Hüttenwerke, ROGESA and ZKS inclusively
(Source: AG der Dillinger Hüttenwerke)
In the coking plant (ZKS), low-sulphur coal gets heated under exclusion of air. The resulting coke as well as ore and other alloys reach the blast furnaces of the ROGESA as bulk material. A blast furnace works on the counter flow principle. This means that the input materials filled from top get heated from the bottom up. Hot metal is liquid at temperatures over 1500°C and it is tapped from the blast furnace at regular intervals. In the steel plant of the Dillinger Hütte, the desired steel properties are set by adding alloying elements as well as through metallurgical treatment. Then, the liquid steel is casted into long blocks and these slabs are transported to the rolling mill. There, they are rolled into plates. After diverse finishing treatments, the plates are ready for the delivery to the customers. The transport of the different intermediate goods results in high demands concerning the logistical planning. There are band-conveyors as well as cranes, flatbed trailers and freight cars to get handled.
Application areas for Operations Research
Application areas for OR techniques are manifold. Previous research projects dealt for example with the following problems:
Warehouse logistics and stacking problems
Particularly in the rolling mill, the temporary storage of the intermediate and final goods (slabs and plates) presents problems with potential of optimization in various areas. Due to lack of space, the slabs as well as the plates have to be stored in stacks arranged by different cranes. The following figure shows a view over a slab yard of the of the Dillinger Hütte:
View over a slab yard of the AG der Dillinger Hüttenwerke
(Source: AG der Dillinger Hüttenwerke)
Commonly, the problem of such storage areas is that the next slab (plate) needed for the next production step is often not the one on the top of the respective stack. Therefore, a resorting has to be performed in order to access the desired slab (plate). Since this resorting is time (and energy) intensive and the resorting should not destroy any preplanned structures, optimization algorithms may have a great impact. This applies specific algorithms acting on the one hand on the stack formation and on the other hand on the resorting.
There are also a lot of scheduling problems arising in the rolling mill. Examples are the sequencing of plates on diverse machines, in particular on furnaces or the roll stands. The following figure shows a drawing of a roller furnace used for the heat treatment of plates:
Drawing of a roller heat furnace
Due to many different technical, physical and logistical restrictions, the design of the processing sequence of the plates holds significant potential of optimization. Currently, a specifically developed genetic algorithm is applied to determine the sequence.
Simulation and Optimization
Besides pure optimization problems, the steel industry is also frequently confronted with different problems which are not able to analyze without help of simulations for reasons of dimension and complexity. The planning for a new production line in the slab adjustment department is an example for such problems.
The slab adjustment department is located between the steel plant and the rolling mill. The new production line should contain flame cutting machines and deburrers as well as among others a marking machine and a machine for the measuring of the slabs. The problem is to determine the required number and the positioning of the machines in order to achieve the desired throughput. For this purpose, a specifically developed simulation tool can be used as decision support. An interesting research area arises in particular by integrating Simulation and optimization techniques.
Currently, the following ongoing projects in cooperation with the AG der Dillinger Hüttenwerke are:
- Optimization methods in the heat treatment of steel
- Simulation of a new production line in the slab adjustment department
- Optimizing the assignment of steel plates to bunkers and warm stacks
- Sequencing the sample processing in the steel plant laboratory
- Optimizing the freight car loading of slabs for a subsidiary rolling mill