ISSN 1000-3665 CN 11-2202/P

    高放废物深地质处置地下水流数值模拟方法研究进展

    Advances in groundwater numerical simulation in deep geological disposal of high-level radioactive waste

    • 摘要: 地下水流数值模型不仅是认识深部水动力场形成演化机制的有效工具,也是建立核素迁移数值模型的基础,因而是高放废物处置场选址和安全评价中重要的技术手段。高放废物深地质处置地下水流数值模拟方法较多,如何选择适当的方法也是值得关注的问题。针对高放废物深地质处置地下水流数值模拟技术展开研究,通过阅读大量国内外文献,文章系统阐述了目前常用的4 类地下水流数值模拟方法的研究进展、适用条件和实例应用;综述了深地质处置中常用的模型不确定性分析方法及研究成果,列表给出了适用于放射性废物地质处置的地下水流数值模拟软件及其在废物处置选择和安全评价中的应用。研究结果表明:等效连续介质模型适用于大区域、长序列、裂隙发育程度较高或较均匀的地区,该类模型方法成熟、所需的数据和参数易于获得,但是不能精确刻画裂隙介质中地下水的流动特征。离散裂隙网络模型适合解决处置场地、储罐尺度等需要精细刻画的地下水流问题,但由于需要大量裂隙及其连通性数据、相关参数等,该方法存在着工作量大、耗时多的缺点。双重介质模型主要用于解决区域尺度裂隙水流问题,但并不能表现出裂隙介质的各向异性、不连续性等特征,因而适用范围存在一定的限制。等效-离散耦合模型可以通过区域分解法对裂隙密度大的区域采用等效连续介质模型,对于裂隙密度较小的地区采用离散裂隙网络模型,从而更符合一般地质条件下裂隙渗流的特征,但也存在交换量难以确定、模型耦合技术问题。通过灵敏度分析,将不同敏感因子对模型敏感指标的影响程度进行排序,提高模型精度、减少参数不确定性分析的工作量。蒙特卡罗法是目前常用的一种模型不确定性方法,原理简单、易于实现。文章展望了数值模型在仿真性、不确定性分析、预测和多介质耦合等方面的研究前景。

       

      Abstract: Groundwater numerical model is not only an effective tool for understanding the formation and evolution mechanism of deep groundwater dynamic field, but also the basis for establishing numerical model of nuclide migration. Therefore, it is an important technical means in the site selection and safety assessment of high-level radioactive waste (HLW) disposal repository. There are many numerical simulation methods for groundwater flow in deep geological disposal of HLW, and how to choose the suitable method is also a problem worth paying attention to. This article focuses on the research of groundwater numerical simulation technology for deep geological disposal of HLW. Through reviewing a lot of relevant papers, systematically expounds the research progress, applicable conditions and practical applications of four kinds of commonly used groundwater numerical simulation methods. In addition, summarizes the model uncertainty analysis methods and research results commonly used in deep geological disposal, and lists the numerical simulation software of groundwater flow suitable for geological disposal of HLW and its application in waste disposal selection and safety assessment. The results show that the equivalent continuum model is suitable for large, long sequence and high fracture development or uniform areas, with the advantages of mature method and easy to obtain the required data and parameters, but the flow characteristics of groundwater in fractured media cannot be described accurately. The discrete fracture network model is suitable for solving the groundwater flow problems that need to be finely described, such as disposal site and repository canister. However, due to the need for a large number of fracture characteristics, connectivity and related parameters data, this method has the disadvantages of heavy workload and time-consuming. The dual medium model is mainly used to solve the problem of regional-scale fractured groundwater flow, but it cannot show the characteristics of anisotropy and discontinuity of fractured media, so the scope of application has certain limitations. The equivalent discrete coupling model can adopt the equivalent continuum model for the area with high fracture density and the discrete fracture network model for the area with low fracture density through the domain decomposition method, which is more in line with the characteristics of fracture seepage under general geological conditions, but there is also the problems that the exchange capacity is difficult to determine and the coupling technology of two models. Sensitivity analysis sorts the influence degree of different sensitive factors on the model sensitive indexes, so as to improve the model accuracy and reduce the workload of parameter uncertainty analysis. Monte Carlo method is a commonly used method for model uncertainty analysis, which is simple in principle and beneficial to implementation. Finally, the author points out that numerical model simulation ability, uncertainty analysis, prediction simulation and multi-medium coupling model research should be strengthen in the future.

       

    /

    返回文章
    返回