Thermal behavior and oxidation kinetics of coal gangue under varying moisture conditions: An experimental and predictive modeling study
This study elucidates the dual regulatory mechanism of moisture on coal gangue oxidation—where moderate levels (8.61%) promote self-heating while excessive moisture (18%) inhibits it—and develops a high-accuracy HPSO-GRU predictive model, supported by molecular dynamics simulations and a graphical interface, to enable early warning of spontaneous combustion risks.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the shadow of coal mines across China, massive piles of discarded rock and dirt, known as coal gangue, sit exposed to the elements. These are not inert mounds; they are chemically active. Over time, the oxygen in the air reacts with the minerals and remaining carbon inside the rock, a slow process that generates heat. If this heat cannot escape, it builds up until the pile ignites on its own, a phenomenon called spontaneous combustion. Such fires release toxic gases, pollute the air, and can destabilize the ground, threatening nearby communities. The key to stopping these fires lies in understanding the internal temperature of the pile and predicting when it will rise dangerously high. However, the relationship between the gases released during this slow burn and the rising temperature is complex and changes depending on how wet the rock is.
A team of researchers from the Inner Mongolia University of Science and Technology set out to decode this relationship, focusing specifically on how moisture content influences the heating process. They began by gathering fresh coal gangue from a local mine, crushing it, and carefully drying it to remove all natural water. They then re-introduced precise amounts of water to create samples with five different moisture levels, ranging from a nearly dry 2% to a damp 18%. These samples were placed in a specialized furnace and heated slowly, mimicking the natural warming of a real pile. As the temperature rose from a cool 30°C to a scorching 550°C, the researchers captured the gases escaping from the rock, measuring the levels of oxygen, carbon dioxide, carbon monoxide, and various hydrocarbons.
The experiments revealed a surprising dual role for water. Contrary to the simple idea that water always cools things down, the researchers found that a moderate amount of moisture actually speeds up the heating process. When the moisture content was around 8.61%, the rock consumed oxygen and produced heat at its fastest rate. In this state, the water molecules seemed to help oxygen gather near the reactive spots on the rock's surface, acting like a bridge that made it easier for the fire to start. However, once the moisture level climbed higher, to 18%, the effect flipped. The excess water began to block the tiny pores in the rock and cover the active sites, effectively smothering the reaction and slowing down the heat release. This means that keeping a coal gangue pile very wet could be a strategy to prevent fires, while a slightly damp pile might be more dangerous than a dry one.
To understand exactly how water molecules move and interact with the rock at a microscopic level, the team ran computer simulations. They built a digital model of the rock's primary mineral, kaolinite, and introduced water molecules to see how they behaved at different temperatures. The simulations confirmed the experimental findings: at that critical 8.61% moisture level, water molecules moved in a way that helped oxygen reach the rock's surface more easily. But at 18%, the water formed a thick layer that trapped the oxygen away from the rock, preventing the reaction from gaining momentum. This microscopic view provided a clear physical explanation for the macroscopic behavior observed in the furnace.
With this understanding of how moisture changes the chemistry, the researchers turned to the challenge of prediction. They needed a way to forecast the internal temperature of a gangue pile based on the gases being released, a task complicated by the changing moisture levels. They developed a new computer model that combines a type of artificial intelligence designed to learn from sequences of data with a smart optimization algorithm that fine-tunes its own settings. This model was trained on the experimental data, learning to recognize patterns in the gas levels that corresponded to specific temperatures. The system was tested against several other existing models and proved to be the most accurate, with its predictions staying very close to the actual measured temperatures.
The study identified that the most important clues for predicting temperature were the amount of oxygen remaining and a specific ratio involving carbon monoxide and the drop in oxygen. These two factors, which the model learned to weigh heavily, provided the clearest signal of how intense the heating reaction was. To make this technology useful for engineers and safety officers, the team built a simple computer interface. With this tool, a user can simply input the current moisture level and the readings from gas sensors, and the system will instantly calculate the internal temperature and identify what stage of heating the pile is in. This approach offers a practical, data-driven method to monitor coal gangue piles, allowing for early warnings before a spontaneous fire can take hold, ultimately helping to protect both the environment and public safety.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.