Abstract Within the framework of safe disposal of high and intermediate level radioactive waste, it is well known that for both the Boom Clay and overlying Eigenbilzen Sands (early Oligocene; Rupelian), the composition varies with depth. Though, a detailed understanding of how these intrinsic variations may affect the basic transport properties is still missing. Therefore, a small but comprehensive and unique data set, containing transport parameters (hydraulic conductivity and diffusivity) and related properties (mineral composition and texture, grain size distribution, specific surface area and pore size distribution) of different samples of the Boom Clay and Eigenbilzen Sands has been gathered during recent years. First, the variability in transport parameters and petrophysical properties is discussed and mutual relationships are investigated. Second, the performance of predictive modelling is explored. The multivariate analysis clearly illustrates that the data are mainly explained by two main axes of variation. One corresponds to the logarithmic hydraulic conductivity (logK) and all petrophysical properties that are very strongly correlated to it. The second axis is independent of logK, and relates to the size of the diffusing molecules. Finally, predictive modelling seems promising at this point, but the limited quality of the currently available test cases prohibits being conclusive on the model performance and its generalisability.

中文翻译：

---

Boom Clay 和 Eigenbilzen Sands 的岩石物理特性和输运特性之间的关系

摘要 在高、中放废物安全处置的框架内，众所周知，对于繁荣粘土和上覆的 Eigenbilzen Sands（渐新世早期；Rupelian），其成分随深度而变化。尽管如此，仍然缺乏对这些内在变化如何影响基本传输特性的详细了解。因此，一个小而全面且独特的数据集，包含 Boom Clay 和 Eigenbilzen 不同样品的传输参数（水力传导率和扩散率）和相关特性（矿物成分和质地、粒度分布、比表面积和孔径分布）沙子是近年来聚集的。首先，讨论了输运参数和岩石物理特性的可变性，并研究了相互关系。其次，探索了预测建模的性能。多元分析清楚地表明，数据主要由两个主要的变异轴解释。一个对应于对数水力传导率 (logK) 和所有与其密切相关的岩石物理特性。第二个轴与 logK 无关，与扩散分子的大小有关。最后，预测建模在这一点上似乎很有希望，但当前可用测试用例的质量有限，无法对模型性能及其普遍性做出定论。一个对应于对数水力传导率 (logK) 以及与其密切相关的所有岩石物理特性。第二个轴与 logK 无关，与扩散分子的大小有关。最后，预测建模在这一点上似乎很有希望，但当前可用测试用例的质量有限，无法对模型性能及其普遍性做出定论。一个对应于对数水力传导率 (logK) 和所有与其密切相关的岩石物理特性。第二个轴与 logK 无关，与扩散分子的大小有关。最后，预测建模在这一点上似乎很有希望，但当前可用测试用例的质量有限，无法对模型性能及其普遍性做出定论。