基于贝叶斯网络的煤矿全生命周期生态风险评估

Ecological Risk Assessment of Coal Mines Across the Whole Life Cycle Based on Bayesian Network

  • 摘要: 为了保证煤矿的正常有序生产,有效降低煤矿生态环境事故风险,将贝叶斯网络应用于煤矿开发全生命周期生态环境风险评价。通过对煤矿全生命周期中各个阶段风险点的识别,构建出包含80个风险因素及其相互关系的贝叶斯网络模型,可用于煤矿全生命周期的风险评估、风险预测和风险诊断。综合考虑煤矿从建设到运营再到关闭过程中可能遇到的各种环境风险因素及其相互关系,通过模糊理论、层次分析法以及逻辑分析法三种方法确定拓扑结构中的参数,并利用概率论的方法来评估环境风险。为了进一步验证模型的适用性,以山西某煤矿为例,分析了该矿区存在的生态环境风险,综合评估环境风险值由模型初始基底风险从20%提高到28%,主要的风险为固废风险和环境管理风险,其风险值分别上涨38%和16%,此外,地下水风险、地表水风险、大气风险以及生态风险也有不同程度的上升。这与环保督察对该煤矿判定的削山取土导致植被覆盖度降低、保水能力减弱、水土流失加剧以及排矸场周围扬尘污染、煤矸石淋溶产生的氟化物浓度超标定性趋势一致,说明了贝叶斯网络可用于煤矿的生态环境风险评价。其对于有效防范煤矿前期—建设期—生产运营—闭坑的完整生命周期中的生态环境污染风险隐患提供了一种科学的方法。

     

    Abstract: To ensure the orderly and sustainable production of coal mines and effectively reduce the risks of ecological and environmental accidents, a Bayesian network is adopted to evaluate ecological environmental risks throughout the full life cycle of coal mine exploitation. By identifying risk points in each stage of the coal mine life cycle, a Bayesian network model covering 80 risk factors and their intercorrelations is established, which supports risk assessment, risk prediction and risk diagnosis for the whole life cycle of coal mines. Various environmental risk factors and their interactions occurring across coal mine construction, operation and closure phases are comprehensively considered. The parameters of the network topology are quantified via three integrated approaches: fuzzy theory, analytic hierarchy process and logical analysis, and environmental risks are further evaluated based on probability theory. To verify the applicability of the constructed model, a coal mine in Shanxi Province is selected as the research case to assess its ecological environmental risks. The results reveal that the overall comprehensive environmental risk probability rises from the baseline value of 20% to 28%. Solid waste risk and environmental management risk represent the dominant hazards, with their risk levels increasing by 38% and 16%, respectively. Meanwhile, groundwater risk, surface water risk, atmospheric risk and ecological risk all rise to varying degrees. This variation trend is consistent with the qualitative conclusions of environmental supervision on this coal mine, including reduced vegetation coverage, weakened water retention capacity and aggravated soil erosion caused by mountain excavation, as well as excessive fluoride concentrations generated by gangue leaching and dust pollution around gangue stockpiles. Such findings demonstrate that the Bayesian network approach is feasible for ecological environmental risk assessment of coal mines. This study provides a scientific technical framework to prevent potential ecological pollution hazards throughout the complete life cycle of coal mines, covering the pre−construction, construction, production and operation, and mine closure stages.

     

/

返回文章
返回