Department of Environmental Engineering Sciences, University of Florida, Gainesville, FL 32611, USA.
J Contam Hydrol. 2013 Sep;152:147-58. doi: 10.1016/j.jconhyd.2013.07.001. Epub 2013 Jul 11.
The equilibrium streamtube model (EST) has demonstrated the ability to accurately predict dense nonaqueous phase liquid (DNAPL) dissolution in laboratory experiments and numerical simulations. Here the model is applied to predict DNAPL dissolution at a tetrachloroethylene (PCE)-contaminated dry cleaner site, located in Jacksonville, Florida. The EST model is an analytical solution with field-measurable input parameters. Measured data from a field-scale partitioning tracer test were used to parameterize the EST model and the predicted PCE dissolution was compared to measured data from an in-situ ethanol flood. In addition, a simulated partitioning tracer test from a calibrated, three-dimensional, spatially explicit multiphase flow model (UTCHEM) was also used to parameterize the EST analytical solution. The EST ethanol prediction based on both the field partitioning tracer test and the simulation closely matched the total recovery well field ethanol data with Nash-Sutcliffe efficiency E=0.96 and 0.90, respectively. The EST PCE predictions showed a peak shift to earlier arrival times for models based on either field-measured or simulated partitioning tracer tests, resulting in poorer matches to the field PCE data in both cases. The peak shifts were concluded to be caused by well screen interval differences between the field tracer test and ethanol flood. Both the EST model and UTCHEM were also used to predict PCE aqueous dissolution under natural gradient conditions, which has a much less complex flow pattern than the forced-gradient double five spot used for the ethanol flood. The natural gradient EST predictions based on parameters determined from tracer tests conducted with a complex flow pattern underestimated the UTCHEM-simulated natural gradient total mass removal by 12% after 170 pore volumes of water flushing indicating that some mass was not detected by the tracers likely due to stagnation zones in the flow field. These findings highlight the important influence of well configuration and the associated flow patterns on dissolution.
平衡流管模型（EST）已证明能够准确预测实验室实验和数值模拟中的稠密非水相液体（DNAPL）溶解。在这里，该模型被应用于预测佛罗里达州杰克逊维尔的一处四氯乙烯（PCE）污染干洗店的 DNAPL 溶解。EST 模型是一种具有现场可测量输入参数的分析解决方案。现场尺度分配示踪剂试验的测量数据用于参数化 EST 模型，并将预测的 PCE 溶解与现场乙醇驱替的测量数据进行比较。此外，还使用校准的、三维的、空间显式多相流模型（UTCHEM）的模拟分配示踪剂试验来参数化 EST 分析解。基于现场分配示踪剂试验和模拟的 EST 乙醇预测与总回收井场乙醇数据非常吻合，纳什-苏特克里夫效率 E 分别为 0.96 和 0.90。EST PCE 预测结果显示，基于现场测量或模拟分配示踪剂试验的模型的峰值出现了更早到达时间的偏移，导致在这两种情况下与现场 PCE 数据的拟合都较差。结论认为，这种峰值偏移是由于现场示踪剂试验和乙醇驱替之间的井筛间隔差异引起的。EST 模型和 UTCHEM 都被用于预测自然梯度条件下的 PCE 水相溶解，其流动模式比用于乙醇驱替的强制梯度双五点要简单得多。基于在具有复杂流动模式的示踪剂试验中确定的参数进行的自然梯度 EST 预测，在经过 170 个孔隙体积的水冲洗后，低估了 UTCHEM 模拟的自然梯度总质量去除量 12%，这表明一些质量没有被示踪剂检测到，可能是由于流动场中的停滞区。这些发现强调了井配置和相关流动模式对溶解的重要影响。
