7TH EUROPEAN COKE AND IRONMAKING CONGRESS - ECIC 2016

The amount and type of different mineral species in coke oven feed coal, the size of these mineral grains and their association with fusible and infusible macerals can have a significant impact on coke quality. The Coal Grain Analysis (CGA) system is an optical reflected light microscope imaging system that collects mosaics of contiguous high resolution colour images to provide detailed information on particles up to 8mm in size. The systems capability has recently been extended by developing optical mineral markers which provide details on the minerals present in samples. To develop the “Mineral Markers” images collected from a scanning electron microscope (SEM) and an optical microscope were registered. The distribution intensities for each individual component were compared to ensure correct identification of each mineral species. Supervised learning algorithms combined with other parameters such as variation in colours and textural information were used to develop an optical mineral marker database which is able to identify two different clay minerals, quartz, siderite, calcite, pyrite and apatite. Incorporation of these mineral markers into the CGA system will provide a cost effective mechanism to obtain mineral abundance, size and mineral/maceral association information on a particle by particle basis in a single analysis.