News Information

For a long time, the coking industry has relied on the traditional production model based on "the experience of veteran workers", facing development bottlenecks such as data silos and difficult-to-understand cost black boxes. At the recently held 8th 6th Standing Committee (Expanding) Meeting of the China Coking Association and the Professional Committee Exchange Annual Conference, several experts attending the meeting pointed out the development path: To promote the coking industry to shift from "experience-based coking" to "scientific coking", it is necessary to rely on the "industry big data + industry mechanism + industry large model + general large model" path, to lay a solid data foundation and build a full life cycle digital management system covering procurement, production, and management. 

The pain points of industry intelligent transformation remain to be addressed. 

As an industry with traditional characteristics, the coking industry faces multiple challenges in its intelligent transformation. Wang Yan, the director of the Coking Department of the Coal Science and Technology Research Institute Co., Ltd., pointed out that raw coal is the fundamental factor for the "cost, quality and greenness" of coking enterprises, but there are obvious shortcomings in the management of coking coal. 

In terms of cost control, "high prices" and "black boxes" coexist, resulting in inaccurate cost control. Wang Yan analyzed that, against the backdrop of a general increase in raw coal prices, coking enterprises lack real-time and transparent data for the entire process, often falling into the predicament of "blind purchasing", leading to a sharp increase in hidden costs. At the same time, the data in the procurement, production, and inventory stages are disconnected, and the "island" phenomenon intensifies, resulting in the lack of overall cost coordination and optimization in the supply chain. The quality losses caused by internal inefficiencies ultimately fall on the coking plants, further increasing their cost burden. 

The quality inspection process also has significant problems. Wang Yan explained that the traditional "sampling - sample preparation - testing" model is lagging and cannot provide real-time guidance for production adjustments, resulting in the production process being "out of control" to some extent and affecting the accuracy of coking coal quality inspection. In the error composition of coking coal quality inspection, sampling error accounts for 70% of the total error, sample preparation error accounts for 20%, and testing error accounts for 10%. 

These issues ultimately led to the excessive reliance on "experience" in the coking industry. "The fragmentation of data has become the primary obstacle to the intelligent transformation of the industry," emphasized Wang Yan. Due to the inability to integrate data from various departments, coking enterprises have difficulty systematically exploring the key cost reduction potential point of "coal blending". From the lag in incoming material inspection, inaccurate inventory盘点， to the rough model in the coal blending process， and the disconnection between theoretical calculations and actual operations， it ultimately affects the stability of coke quality and the accuracy of production cost control. 

Build a complete chain of intelligent control management 

Facing the challenges of industry intelligent transformation, Shi Yanfeng, the president of the China Coke Association, proposed that we should accelerate the construction and improvement of a centralized control platform for production and operation integration that integrates distributed control systems (DCS), manufacturing execution systems (MES), and enterprise resource planning (ERP), to achieve real-time and accurate collection, analysis, and sharing of all elements' data. At the same time, by combining real-time production data and applying and developing relevant process expert systems, we can realize intelligent optimization of local process units and gradually move towards the systematic optimization of the entire coke production process. Additionally, we need to fully implement the energy management system (EMS), conducting real-time measurement, online analysis, and dynamic balance of key energy media such as gas, steam, electricity, and water resources, and deeply exploring the "leaks, spills, drips, and leaks" and energy waste points within the system. 

Shi Yanfeng divides the digitalization and control promotion of the coking industry into three stages. The first stage is the infrastructure construction and data integration stage, where a unified IoT platform and data center are established to realize the onlineization of core business processes such as incoming materials, testing, and storage yards, as well as automatic data collection. Visual dashboards are developed to achieve the preliminary display of key data. The second stage is the deepening of intelligent applications, where digital twins are introduced to achieve 3D visualization of the storage yards, and an intelligent coal blending optimization module is developed and launched, establishing a preliminary cost accounting and early warning system. The final stage is the advancement of intelligent decision-making, where AI applications are deepened, such as demand forecasting and predictive maintenance of equipment, and the decision support center is improved to provide advanced data mining and simulation analysis functions, and an industrial collaboration ecosystem is constructed. 

Wang Yan pointed out that the "Intelligent Digital Management System for the Whole Life Cycle of Metallurgical Raw Coal" developed by this institute covers aspects such as intelligent procurement and supplier management, intelligent coal blending and cost optimization, full life cycle cost accounting, refined on-site and storage yard management, quality management and traceability of coke, as well as production process monitoring and feedback. It can establish intelligent management of storage and stacking, digital coal yards, and intelligent sampling and processing equipment, etc., addressing issues such as difficulty in inventory counting and the black box of logistics, and achieving "no disruption and no loss of authenticity" during storage and transportation processes, thereby reducing the cost per ton of coke. 

The implementation of intelligent applications has achieved remarkable results. 

How can coking enterprises achieve intelligence? Zheng Helalei, the deputy chief engineer of Huatai Yongchuang (Beijing) Technology Co., Ltd., stated that the intelligence of coking enterprises cannot be achieved without industry mechanisms and production equipment data. It is necessary to rely on the "industry big data + industry mechanism + industry large model + general large model" path to create platform-based, iterative, and low-barrier software products. 

Taking the dry quenching process as an example, Zheng Helalei introduced the practical application of AI in solving boiler tube burst and desulfurization problems. For new projects, the problem of boiler tube burst can be fundamentally solved by using technologies such as cyclone dust removal; for existing projects, AI prediction and in-furnace desulfurization technology can effectively address the issue. The company's HTZW-01 dry quenching furnace internal desulfurization intelligent device automatically, uniformly and stably distributes the desulfurization agent by real-time detection of the SO2 content in the flue gas discharged from the dry quenching furnace. This device can be installed online without production suspension, achieving an annual cost reduction of approximately 3 million yuan (calculated based on an annual production capacity of 1 million tons of coke), reducing the SO2 content in the circulating gas inside the furnace by 35% to 55%, and reducing the amount of desulfurization agent used outside the dry quenching furnace by 25% to 45%. 

Xu Xiuli, the director of the Coal Resources Department of China Steel Group Anshan Thermal Energy Research Institute Co., Ltd., believes that the key for coking enterprises to reduce costs lies in "expanding the sources of coking coal, deepening the understanding of coal quality, and optimizing the coal blending structure". She pointed out: "The prerequisite for achieving intelligent coal blending is data governance. Sometimes, it takes us three months or even longer to conduct data governance for the enterprises, eliminating invalid and inaccurate data, and laying a reliable foundation for the model." 

Xu Xiuli introduced that the company's developed intelligent coal blending software has established a "resource identity card" system for coal and coke, enabling full-process data traceability. Through mechanism research and a new understanding of coal quality, this software can optimize coal blending predictions. For instance, using lean coal and rich coal to replace coking coal and thin coal, and using coking coal to replace expensive rich coal and thin coal, the cost of coal blending can be reduced by approximately 40 yuan per ton. 

Coal blending should not be limited to the final proportion adjustment; it should cover the entire process from coal source selection, procurement, transportation to storage and coal blending. Wang Yan believes that by building an intelligent coal blending system with the lowest cost and the best quality, enterprises can address challenges such as large fluctuations in coal quality, high blending costs, inaccurate prediction of coke quality, and unutilized data value. This will truly enable the transformation from "empirical coking" to "scientific coking".