Challenges for the Material Flow Analysis at an enterprise level | Herausforderungen für die Stoffflussanalyse auf betrieblicher Ebene

The method of Material Flow Analysis (MFA) allows tracing flows and stocks of gods and substances within a system, defined in time and space. The analysis hence can be conducted at regional scale, at national scale, at global scale, but also on the level of an enterprise. In the present paper the application of MFA to a blast furnace process is discussed. In particular heavy metal flows through the blast furnace plant are modelled using MFA. Thereto, in a first step transfer coefficients (TC) of heavy metals have been determined for one particular year (2008) and checked for plausibility. In the second step, these TC have been applied to simulate heavy metal flows in the outputs of the plant for the years 2009 and 2011. These flows have subsequently been compared with measured data. The results show that the application of TC for Ni delivers plausible results. Also for Cd the application of the TC seems to be reasonable, although a detailed investigation of the Cd balance indicates significant differences between the total input and output of Cd. As amounts of Cd detected in the output are larger than in the input, this difference originates most likely from non-detectable Cd contents (< LOD limit of determination) in different input flows (lump ores, pellets). Moreover, it has been shown that constant TC for Zn may not be applicable, as some output flows of Zn determined by TC are significantly different to observed data (e.g. differences in the Zn content of the hot metal of up to 100 %). Additionally, the results show that differences in total input and output flow of Zn may vary over time, as the differences can be positive (Input > Output) as well as negative.Based on the results of the investigations the following challenges for applying MFA on an enterprise level could be identified: The applicability of TC is depending on the respective element. Thus, it is also depending on the behavior of the element in the process. However, the system modelled should be investigated in detail even if the application of TC delivers possible results. In the case of Cd, the application of TC was reasonable, although big differences in the balance (input vs. output) are observable. Such differences might occur because the concentrations of the input or the output flows are underestimated. Another reason might be that samples taken and analyzed are not representative for the flow of interest. In case that significant differences between simulated results (based on TC) and measured data are observed (like for Zn), the thermochemical behavior of the respective element should be considered within the investigations and subsequently also with MFA.