Identification of Preferential Flow Paths at Sites of Groundwater Contamination

小行星9526888 大量的研究表明，渗透系数的空间分布是地下污染物运动的重要控制因素。 许多理论已经发展到量化的空间变化的水力传导性对污染物的传输使用随机过程或分形表示模型的传导性变化的影响。然而，越来越明显的是，使用例如从岩心数据估计的平稳随机过程的前两个时刻来对场地处的电导率变化进行建模在由复杂的岩性混合物组成的单元中可能具有限制。 显然，现场的特定功能的水力传导率分布的一个更大的规模需要量化，以可靠地预测污染物在这样的系统中的运动。 特别是，知识的存在，横向连续的区域的高水力传导性，作为优先的流动路径，往往是至关重要的污染物传输的网站的成功建模。然而，在现场对这些小虫进行实地鉴定已被证明是一项相当困难的任务。然而，实地确定这些区域已证明是一项相当困难的任务。常规的现场技术仅提供高度平均性质的信息或仅限于测试井附近的信息。 这里提出的研究的最终目标是开发一个全新的领域的方法来估计的空间变化的水力传导率。 尽管该方法是为了估计饱和地层中水力传导系数变化的一般任务而开发的，但该方法对于识别优先流动路径特别有效。 这种新方法包括三个主要部分：（1）最近提出的水力层析成像方法，（2）通常用于大规模示踪剂试验以获得垂直隔离水样的多级取样威尔斯，和（3）最近开发的用于生物医学应用的微型光纤压力传感器。 水力层析成像的现场应用的主要限制之一是需要关于泵引起的水头变化的垂直变化的详细信息。 然而，在多级取样器的小管中使用微型光纤压力传感器应能克服这一限制，并实现液压层析成像的巨大潜力。 对方法进行为期一年的初步评估是本报告所建议工作的主要目的。 该评估将包括水力层析成像方法的理论扩展/分析，以及在一个装有大量仪器的现场对该方法进行的初步现场评价，在该现场，可以从以前的大规模示踪剂试验中获得有关水力传导率井间变化的一些信息。 将多层采样器中的微型光纤传感器提供的详细头部数据与层析成像反演方法相结合，有可能提供 关于控制污染物输送的特定场地水力特征的非常有用的信息。 将这些特征纳入场地模型，应能大大提高由此产生的模型预测的质量，从而使风险评估更加可靠，并能更有效地为场地特征描述和补救活动分配资源。 此外，这种方法应产生更详细的描述水力传导率的变化比以前可能的。 这些更详细的描述应该有助于完善现有的电导率变化建模方法，并希望有助于更好地将电导率变化与其地质基础联系起来。 ??

9526888 Butler A considerable body of research has identified the spatial distribution of hydraulic conductivity as a significant control on the movement of contaminants in the subsurface. A number of theories have been developed to quantify the influence of spatial variation of hydraulic conductivity on contaminant transport using stochastic processes or fractal representations to model the conductivity variations. It is increasingly apparent , however, that the modeling of the conductivity variations at a site using, for example, the first two moments of a stationary stochastic process estimated from core data may have limitations in unites composed of a complex mixture of lithologies. Clearly, site-specific features of the hydraulic conductivity distribution of a larger scale need to be quantified in order to reliably predict contaminant movement in such systems. In particular, knowledge of the existence of laterally contiguous zones of high hydraulic conductivity, which serve as preferential flow paths, is often critical for the successful modeling of contaminant transport a site. The field identification of such somes, however, has proven to be a rather difficult task at site. The field identification of such zones, however, has proven to be a rather difficult task. Conventional field techniques only provide information of a highly averaged nature or information restricted to the immediate vicinity of the test well. The ultimate objective of the research proposed here is to develop a completely new field method for the estimation of spatial variations in hydraulic conductivity. Although developed for the general task of estimation of hydraulic conductivity variations in saturated formations, this method would be especially effective for the identification of preferential flow paths. This new methodology involves three primary elements: (1) a recently proposed methods for hydraulic tomography, (2) multilevel sampling wells commonly employed in large-scale tracer tests for obtaini ng vertically isolated water samples, and (3) miniature fiber-optic pressure sensors recentlydeveloped for biomedical applications. One of the primary constraints on the field application of hydraulic tomography has been the need for detailed information about vertical variations in pumping-induced head changes. The use of miniature fiber-optic pressure sensors in the small tubing of the multilevel samplers, however, should enable that constraint to be overcome and the considerable potential of hydraulic tomography to be realized. An initial one-year assessment of the methodology is the primary purpose of the work proposed here. This assessment will include a theoretical extension/analysis of the hydraulic tomography approach and a preliminary field evaluation of the methodology at a very heavily instrumented field site where some information about the interwell variations in hydraulic conductivity is available from a previous large-scale tracer test. The coupling of the detailed head data provided by miniature fiber-optic sensors in multilevel samplers with the tomographic inversion method has the potential of providing extremely useful information about the site-specific hydraulic features controlling contaminant transport. The incorporation of such features into a site model should dramatically improve the quality of the resulting model predictions, thus leading to more reliable risk assessments and a more efficient allocation of resources for site characterization and remediation activities. In addition, this methodology should produce much more detailed description of hydraulic conductivity variations than has previously been possible. These more detailed descriptions should help refine existing approaches for the modeling of conductivity variations and, hopefully, help better tie the conductivity variations to their geologic basis. ??