In the context of new power system construction and the normalization of deep peak shaving, increasing flexibility requirements are imposed on thermal power units. Significant operating cost pressures were imposed on coal-fired power plants. Real-time control throughout the coal process from plant intake to furnace entry was considered essential for thermal power units under economically flexible operation. The study was focused on the complex coal-handling system, which includes unloading, stacking, reclaiming, and conveying. A Petri net was adopted as the theoretical framework. Color sets were introduced to characterize the quantity, quality, and cost attributes of coal flow. Time functions were introduced to describe the time delays triggered by operational actions during transfer. A dynamic arc-weight adjustment mechanism based on reclaiming quantity was proposed to overcome the limitation of traditional Petri nets. Coal-flow attribute update rules were established to describe changes in coal-flow properties at different stages and to achieve accurate coal-flow tracing. A real-time estimation method for furnace coal quality was developed based on the traced quantity, quality, and cost information. Application research was conducted using operational data from a large coal-fired power plant in China. The results show that a coal-flow tracing accuracy of 97.7% is achieved and that the average error in furnace coal calorific value calculation is maintained within 3.50%, satisfying engineering accuracy requirements. The study has important practical significance for enhancing unit flexibility, reducing costs, and improving efficiency in coal-fired power plants.