Since 2008 the refractories company RHI runs a magnesia plant in Dashiqiao City, Liaoning province, Peoples Republic of China. For production, most of the raw magnesite is supplied by the joint-venture partner JDMG, which operates the Huaziyu open pit mine. To produce high quality magnesia products at Aoding plant the knowledge about the various procedures in the process line is elementary. Therefore, this study was initiated by RHI to analyse the following tasks. The first aim of this thesis is to analyse the ore supply process from mine to mill, in order to detect potentials to improve the quality control of the crude ore supply. The evaluation of the current raw material supply process showed that the quality and the processability of the mined crude ore cannot be correlated with its geographic origin in the deposit. Therefore, the second aim of this study is to find methods to survey the origin of delivered magnesite within the Huaziyu deposit and to achieve traceability of delivered crude ore. The third aim of this work is to establish an unbiased sampling station at the primary crushing line because currently the sampling of the crushed ore in the plant is biased. Various surveying methods for the localisation of mining areas in the deposit were evaluated in order to decide about their applicability in the mine. Due to permission by JDMG to measure with GPS it is possible now to trace the source area of delivered raw magnesite in the Huaziyu deposit. For the development of an alternative sampling station at the primary crushing line, extraordinary sampling and chemical analysis of raw ore samples were carried out to do the design of the sampling station according to specific material flow parameters. In the scope of this thesis an alternative sampling station was designed in form of an automated sampler to overcome the deficiencies of the current manual sampling. Additionally, a new sample processing procedure was developed for the laboratory. This allows representative sampling without bias and creates results of defined accuracy within limits of confidence.