TL;DR
This paper introduces a new 3D dynamic prediction model for mining subsidence, called the "strip unit mining model," which accounts for the time-dependent nature of subsidence. By integrating time factors into the probability integral method and validating it with field data from the Weiqiang coalfield, the model accurately predicts surface deformation, offering a more realistic approach to understanding and mitigating mining impacts.
What Problem Does This Paper Address?
Underground coal mining inevitably leads to surface subsidence, posing significant risks to the environment and infrastructure. While various prediction models exist, most simplify this complex, four-dimensional (space and time) phenomenon into 2D or 3D static problems. This simplification fails to capture the crucial time-dependent evolution of subsidence, especially the initial, active, and declining stages. Accurately predicting this dynamic behavior is essential for effective land management, infrastructure protection, and environmental sustainability in mining areas.
Methods
The authors propose a novel three-dimensional dynamic prediction model called the strip unit mining model. This model builds upon the established probability integral method, which treats rock mass as a random medium. The core innovation lies in:
- Strip Unit Concept: Longwall mining is conceptualized as the superposition of numerous "strip units" of mining. This approach better reflects the actual advancement of the working face.
- Dual Time Factor Integration: The model explicitly incorporates two time-related factors:
- Mining Time (T): The time elapsed since a specific strip unit was mined.
- Advancement Time (t): The time related to the working face's forward movement.
- Knothe Time Function: The classic Knothe time function is adapted to describe the dynamic subsidence process, accounting for the initial, active, and decay stages of subsidence.
- Parameter Determination: Crucial prediction parameters (subsidence coefficient, tangent of main influence angle, etc.) are determined by analyzing field-measured data from the Weiqiang coalfield.
Datasets
The study utilizes field data collected from the Weiqiang coalfield in northern Shaanxi, China. This includes:
- Surface Movement Observations: Two systematic observations and 27 daily observations were conducted at the first mining working face (1301).
- Measured Data: A total of 1508 sets of reliable observation data were obtained, including plane coordinates and elevations of measuring points.
- Geological and Mining Conditions: Information on coal seam thickness, dip angle, buried depth, and mining technique (longwall mining with full caving) was also utilized.
No new benchmarks were introduced; the study focuses on validating the proposed model against existing field data.
Key Insights and Ideas
- Time is Crucial: The study strongly emphasizes that mining subsidence is not instantaneous but a dynamic process that evolves over time. Ignoring this temporal aspect leads to incomplete and potentially inaccurate predictions.
- Strip Units Mimic Reality: Conceptualizing longwall mining as a series of "strip units" moving forward provides a more granular and realistic representation of the mining process compared to treating the entire working face as a single entity.
- Dual Time Factors Enhance Accuracy: Incorporating both the time since a strip was mined and the overall advancement time significantly improves the model's ability to capture the complex spatio-temporal evolution of subsidence.
- Field Data is King: The successful fitting of the model to measured data highlights the importance of accurate field observations for calibrating and validating prediction models. The parameters derived from the Weiqiang coalfield data proved effective.
Key Results
- The proposed strip unit mining model was developed and formulated, incorporating dual time factors into the probability integral method.
- The model was validated using field data from the Weiqiang coalfield.
- The predicted subsidence curves showed good agreement with the measured subsidence curves, demonstrating the effectiveness of the new model.
- The study confirmed that surface movement typically progresses through initiation, active, and decline stages, consistent with general subsidence laws.
- The maximum predicted surface subsidence was found to be approximately 1.4 m, which was consistent with measured values.
Limitations and Future Work
- Geological Simplicity: The study notes that the geological structure in the Weiqiang coalfield is relatively simple, with horizontal mining conditions. The assumption of a constant subsidence coefficient might need re-evaluation in more complex geological settings.
- Time Step Simplification: While the study uses "day" as the time unit for practical calculation, it acknowledges that subsidence is a continuous time function. Further refinement could explore finer time steps, though this would increase computational complexity.
- Future Work: The paper suggests that the dynamic predicted subsidence velocity can be obtained by partial differentiation of the final three-dimensional subsidence model. Further research could explore the application of this model to different mining conditions and geological environments to assess its broader applicability.
Summary updated