Clay liners are widely used as porous membrane barriers to control solute transport and to prevent the leakage of leachate both in horizontal and vertical flow scenarios, such as the isolated base and ramps of sanitary landfills. Despite the primary importance of saturated hydraulic conductivity in a reliable simulation of fluid flow through clay barriers, there is no model to predict hydraulic conductivity of clayey soils permeated with saline aqueous solutions because most of the current models were developed for pure water. Therefore, the main motivation behind this study is to derive semi-empirical models for simulating the hydraulic conductivity of clayey soils in the presence of arbitrary solute concentrations in addition to deionized water. In order to achieve this goal, a relatively comprehensive dataset of 842 measured hydraulic conductivities was retrieved from the experimental literature, where almost 44% of them are related to certain solute concentrations. Afterwards, two modelling approaches were introduced; the first one is a modified form of Mbonimpa et al.'s (2002) model, in which the constants are adjusted to take into consideration the variations in liquid limit due to a change in solute concentration. A modification term was added to the model for the sake of accuracy. In the second approach, a new form of solute concentration-dependent hydraulic conductivity function was proposed, where special attention was given to void ratio and adaptive liquid limit as effective parameters. The results revealed that hydraulic conductivity predictions could be erroneous if the influence of solute concentrations in permeating fluid is ignored. An error analysis was conducted to examine the models' applicability and deviations. A blind independent set of data, including 132 data points, was also used to verify models. On the other hand, both newly proposed models could predict the hydraulic conductivity for a variety of soils, salt species, and concentrations well. Therefore, the proposed modelling approaches are somehow unique by considering the salinity of the pore fluid in addition to deionized water. More importantly, both models are comprised of easy-to-measure parameters with clear physics-based implications.

粘土衬里被广泛用作多孔膜屏障，以控制溶质运输并防止渗滤液在水平和垂直流动情况下的泄漏，例如卫生填埋场的隔离底部和坡道。尽管饱和至关重要在通过粘土屏障的流体流动的可靠模拟中的导水率，没有模型来预测渗透盐水溶液的粘土的导水率，因为大多数当前模型都是为纯水开发的。因此，本研究背后的主要动机是推导半经验模型，用于模拟除去离子水外存在任意溶质浓度的粘土土壤的水力传导率。为了实现这一目标，从实验文献中检索了 842 个测量水力传导率的相对全面的数据集，其中几乎 44% 与某些溶质浓度有关。随后，介绍了两种建模方法；第一个是 Mbonimpa 等人 (2002) 模型的修改形式，其中常数被调整以考虑由于溶质浓度变化引起的液限变化。为了准确起见，在模型中添加了修改项。在第二种方法中，提出了一种新形式的溶质浓度依赖性水力传导率函数，其中特别注意空隙率和自适应液限作为有效参数。结果表明，如果忽略渗透液中溶质浓度的影响，水力传导率预测可能是错误的。进行了误差分析以检查模型的适用性和偏差。一组独立的盲数据，包括 132 个数据点，也用于验证模型。另一方面，两个新提出的模型都可以很好地预测各种土壤、盐种和浓度的导水率。因此，通过考虑除去离子水之外的孔隙流体的盐度，所提出的建模方法在某种程度上是独特的。更重要的是，这两个模型都包含易于测量的参数，具有明确的基于物理的含义。