Cement production is energy-intensive and therefore expensive. The main process steps in cement production are the preparation of the raw material in the raw mill, the subsequent burning process in the rotary kiln, and finally the production of the desired cement type in the cement mill.
The subprocesses described have in common that their reaction times to a control intervention are very long (large dead times). Therefore, classical controllers are only suitable for this task to a limited extent. This is a major reason for the still rather rudimentary state of automation in the cement industry.
In order to control these processes in an automated way despite their demanding characteristics mentioned above, modern methods using artificial intelligence are used. For example, the use of forecasts makes it possible to anticipate the behavior of the process and in this way to make the necessary control interventions at an early stage. This methodology results in the process having significantly lower fluctuations (process stability), which translates into better product quality, increased production if necessary, and ultimately a reduction in specific energy consumption.
Our solutions for the cement industry are designed to automate these processes and at the same time to optimize them. In this way, it is possible to reduce the specific energy consumption of each sub-process. In the grinding processes, energy consumption is dominated by electricity consumption, in the kiln process by fuel consumption.
Therefore, expect an improvement in the energy balance of your plant and, as a consequence, a reduction in energy costs.
For the optimization of grinding processes, no additional sensors are installed; instead, the available process data are used to train special process models. The process model created in this way provides, among other things, important process variables that cannot be obtained otherwise. One example is the amount of material circulating in the mill.
The additional information gained in this way enables early action to be taken so that the product quality remains within specification, the process runs more stably and thus consumes less energy.
When optimizing the thermal sub-process, consisting of preheater, rotary kiln and clinker cooler, our thermographic camera is used as an additional sensor. The image of the main flame helps to control the process so that the free lime content remains within the permissible limits. In conjunction with the free lime forecast, a laboratory value available at best every two hours thus becomes a quasi-continuously available process value.
Fields of application for our solution in the hot part of the process are furnace optimization and optimization of flue gas denitrification by our highly efficient heSNCR process.
In the course of the furnace optimization, our system determines the setpoints for the essential actuators. For example, these include the speed of the induced draft fans, all fuel quantities at the main burner and the calciner, the material feed, the combustion air quantities, the cooling air fans in the cooler, and the grate speeds.
Our heSNCR process is capable of keeping NOx emissions below 200 mg/Nm3 as a daily average even under difficult conditions. In a direct comparison with an existing plant, the use of our heSNCR plant led in many cases to a reduction in consumption of ammonia water by double-digit percentages.
By combining classic control technology with artificial intelligence, we were able to show in many projects that increased process stability results in a notable reduction in energy consumption.