非均质性对DNAPL污染源区结构特征影响的实验研究

重非水相液体(DNPALs)在地下环境中的运移分布与残余捕获受多种因素控制。选择四氯乙烯(PCE)作为DNAPLs特征污染物,通过二维砂箱实验探究介质非均质性对PCE运移及污染源区结构特征的影响。采用透射光法监测PCE在砂箱内的运移过程,定量测量PCE的饱和度。采用空间矩分析PCE污染体的平均运移行为随时间的变化,并采用离散状与池状PCE体积比(GTP)定量表征污染源区结构特征。结果表明,由于无法克服毛细压力,PCE在细砂透镜体上方聚积并侧向扩散,水平扩散范围显著增大。随机非均质介质中,PCE绕开细砂透镜体,沿着粗砂透镜体构成的优势通道运移,运移路径的不规则性及饱和度分布的空间变异性增强。PCE的捕获主要为毛细屏障和拉断效应两种形式。运移路径延长,路径上残留的离散状PCE增多,低饱和度的比例与GTP增大。非均质性显著影响PCE的运移路径、捕获及空间分布。     

盐度对多孔介质中DNAPL运移和分布的影响

选用四氯乙烯（PCE）作为典型DNAPL污染物,以NaCl作为地下水中溶解盐代表,研究盐度对DNAPL在饱和多孔介质中运移和分布的影响。通过批次实验测定NaCl水溶液/石英砂/PCE三相体系下的接触角和界面张力,结果表明,PCE在石英砂表面的接触角随着水中NaCl浓度的增大而减小,而PCE和NaCl水溶液的界面张力随着NaCl浓度的增大而增大,尤其当氯化钠浓度较高时（>0.1 mol/L）,影响程度更为显著。在此基础上,采用透射光法监测不同介质情景下DNAPL在二维砂箱中的运移和分布,定量测定DNAPL在介质中的饱和度。实验结果表明,地下水盐度的增加将促进DNAPL的垂向入渗,减少被截留在运移路径上的DNAPL量,使得DNAPL运移路径及累积形成的池状DNAPL（pool）向水流方向偏移。在均质多孔介质和含有透镜体的非均质多孔介质中,随着盐度的增加,DNAPL在横向和垂向上的展布均呈现出增加趋势,导致污染源区变大,且介质中以离散状存在的DNAPL量明显增加。     

利用Tween80溶液冲洗修复萘污染地下水模拟实验

采用Tween80溶液冲洗修复萘污染地下水,分析了Tween80溶液质量浓度和冲洗流速对修复效果的影响,并考察了Tween80溶液对萘的增溶效果及对萘在介质上吸附特性的影响。结果表明：萘和Tween80在中砂介质上的吸附分别符合线性吸附和F型吸附;Tween80增溶萘效果较好,10.0 g/L Tween80溶液中,萘的表观溶解度达到500 mg/L,约为水相饱和溶解度的17倍;确定Tween80洗脱萘的临界解吸质量浓度为2.0 g/L;不同质量浓度、不同冲洗流速修复萘污染地下水实验结果表明,10.0 g/L Tween80溶液、3.0 mL/min流速为最佳冲洗修复参数,萘的去除率达到90%以上。     

表面活性剂强化的DNAPLs污染含水层修复过程的数值模拟

根据含水层中水、表面活性剂和DNAPLs的运移规律和相互作用机理, 建立三维多相流数值模拟模型, 用以模拟表面活性剂强化的DNAPLs污染含水层的修复过程.将所建立的模型应用于一个被PCE污染的非均质含水层中, 并分别对污染物的污染过程以及修复过程进行模拟.研究结果表明: 数值模拟模型给出了表面活性剂强化含水层修复过程中非水相流体迁移转化的数学描述, 能够在短时间内、参数有限的条件下真实地刻画DNAPLs在含水层中的运移规律, 并能有效地模拟表面活性剂的修复过程.此外, 模拟结果显示, 由于表面活性剂对PCE的增溶增流作用, 有效地提高了PCE在水中的溶解性和迁移性, 其修复40 d的去除率达到63.5%, 与抽出处理法(去除率为31.8%)相比修复效果明显增强.      

裂隙介质中流速及倾斜度对四氯乙烯的运移影响研究

文章选用四氯乙烯（PCE）为代表性重非水相液体（DNAPL）,采用透射光法监测PCE在单裂隙介质中的运移过程和形态分布,探究流速及裂隙面倾斜度对PCE运移分布行为的影响。结果表明,单裂隙介质中PCE呈团块状脱离注样针,并在运移过程中呈近椭圆状。地下水流速越大,初始面积Fs越大,近椭圆的PCE离心率e越小;裂隙面倾斜度越大,初始面积Fs越小,近椭圆的PCE离心率e越小。不同流速或倾斜度条件下,入渗稳定后的PCE团块面积与离心率均明显下降。地下水流速及裂隙面倾斜度的增大对于PCE纵向运移具有促进作用,使PCE污染物前端锋面运移速率增大,在相同时间内更快运移到裂隙底部。PCE入渗停止后,流速增大促使污染源区PCE饱和度降低,使饱和度峰值随水流速度增大而减小;倾斜度增大抑制污染源区PCE饱和度降低,使饱和度峰值随倾斜度增大而增大。     

非均质介质中重非水相污染物运移受泄漏速率影响数值分析

重非水相污染物(DNAPL)在地下介质中运移和分布受多种因素控制,包括DNAPL本身的物理化学性质,土的性质,泄漏条件等等。由于介质的非均质性,使得多相流运移行为更为复杂。基于地下水随机理论构建渗透率随机场,采用蒙特卡罗方法探讨泄漏速率对非均质饱和介质中DNAPL运移的影响。数值结果表明,在泄漏总量一定的情况下,泄漏速率越低,介质非均质性对DNAPL运移的影响程度越高。反之,DNAPL的渗漏速率越高,小尺度地层的非均质性影响越低。由于DNAPL运移过程中在垂直方向受重力的影响,污染羽在空间上的质心位置(一阶矩)以及展布范围(二阶矩)在垂直方向上的变异程度要高于水平方向。     

黄原胶在含水层的迁移特性及对KMnO4迁移的影响

向地下水注入化学药剂进行修复时,药剂迁移主要集中在渗透性相对较高的区域,致使低渗透区内的污染物无法有效去除。通过注入聚合物（黄原胶）对地下水进行黏度控制,可以有效提高修复药剂在低渗透区的迁移能力,从而提高修复效果。黏性流体在地层中的迁移特性是该技术应用的理论基础,因此本研究运用一维模拟柱实验分析了含水层介质对黄原胶流体的阻滞作用,黄原胶注入前后介质的压力及渗透系数的变化以及黄原胶与修复药剂KMnO4迁移同步性。实验结果表明:当黄原胶溶液注入到介质后,介质对黄原胶的阻滞导致其有效孔隙度减小,因此会在一定程度上加速后续注入溶液溶质的运移,且介质渗透系数越小,对黄原胶阻滞作用越明显;黄原胶注入导致含水层渗透性降低,流体运移阻力增加,特别是在细砂和粉砂介质中,渗透系数都降低了一个数量级;虽然黄原胶和KMnO4在迁移锋面存在一定差异,但经过2 h后迁移速率基本相同,具有较好的同步性。     

裂隙宽度空间变异性和泄漏条件对网络裂隙中DNAPLs运移影响研究

DNAPLs本身的化学性质、裂隙的几何性质以及泄漏条件等影响重非水相流体（DNAPLs）在裂隙介质中的运移和分布。针对DNAPLs的运移规律研究多集中在孔隙介质和单裂隙中,在随机网络裂隙中的研究较少。本研究生成随机网络裂隙是基于蒙特卡罗方法,运用像素扫描识别并输出裂隙的坐标和宽度,然后采用PetraSim模拟四氯乙烯(PCE)在随机网络裂隙中的运移,探讨裂隙宽度空间变异性和泄漏条件（包括泄漏速率和泄漏位置）对DNAPLs运移的影响。数值模拟结果表明:DNAPLs的空间展布和运移路径受裂隙宽度空间变异性影响,随着网络裂隙中裂隙宽度空间变异性增大,出现优势通道,DNAPLs运移的速率加快,DNAPLs的质心位置和饱和度空间分布发生明显变化;DNAPLs的运移速率和空间展布受泄漏速率影响,泄漏速率越大,DNAPLs运移速率越快,模型底部蓄积的DNAPLs的饱和度越大,DNAPLs的空间展布也越大;同一网络裂隙中,泄漏位置不同,导致DNAPLs的运移路径及分布范围不同,不同的泄漏位置重力方向裂隙空间变异性不同,导致DNAPLs运移路径和空间展布各不相同。研究结果可以丰富裂隙介质中DNAPLs运移机理研究,为裂隙介质中DNAPLs污染修复提供模型参考。     

基于改进稀疏网格替代模拟的地下水DNAPLs运移不确定性分析

由于实际的水文地质条件具有复杂性和变异性,地下水DNAPLs运移数值模拟的不确定性不可避免。为解决不确定性分析中因多次调用DNAPLs运移模型导致的计算耗时问题,本次研究在传统稀疏网格（SG）替代模型的基础上,提出了一种将局部自适应（LA）和维数自适应（DA）耦合的改进替代模型DA-LA-SG。通过两个解析案例和一个PCE运移室内实验,验证了DA-LA-SG的替代效率和精度,并将其用于室内砂箱PCE运移模拟不确定性分析。研究结果表明:在替代模型构建初期,LA-SG的替代效率优于或接近于DA-SG和DA-LA-SG,随着插值节点数量的增多,DA-SG和DA-LA-SG的替代效率逐渐优于LA-SG,DA-LA-SG具有最高的替代效率。DA-LA-SG能够高效、高精度地建立PCE运移模型的似然函数替代模型,且被用于PCE运移模拟参数不确定性分析。结果表明:背景介质渗透率k1、拟合系数n1和透镜体渗透率k2的可识别性较强,背景介质孔隙度μ1、透镜体孔隙度μ2和拟合系数n2的识别效果较差,这些参数对饱和度观测数据不敏感。     

非均质裂隙介质中重非水相流体运移

由于裂隙介质具有强烈非均质性,使得重非水相流体(DNAPLs)在裂隙介质中的运移行为较孔隙介质更为复杂.基于指示模拟算法模拟裂隙介质的二元变量(粗糙面接触和裂隙开口),综合模拟退火算法构建裂隙渗透率随机场.采用T2VOC模拟DNAPLs在裂隙介质的运移,探讨粗糙面接触的相关长度、各向异性比、倾角以及非均质性程度对DNAPLs运移分布的影响.数值分析结果表明,粗糙面接触的相关性越差,DNAPLs污染范围越大;粗糙面接触的各向异性比和倾角增大,将导致更多DNAPLs残留;而相关长度和各向异性比增大以及倾角减小,都会引起DNAPLs锋面运移速率增大;随着渗透率非均质性增强,会导致局部粗糙面接触上蓄积的DNAPLs饱和度增大,运移速率减小.     

Use of geostatistical models in DNAPL source zone architecture and dissolution profiles assessment in spatially variable aquifer

Following the accidental subsurface release of dense nonaqueous phase liquids (DNAPLs), spatial variability of physical and chemical soil/contaminant properties can exert a controlling influence on infiltration pathways and organic entrapment. DNAPL spreading, fingering, and pooling typically result in source zones characterized by irregular contaminated regions with complex boundaries. Spatial variability in aquifer properties also influences subsequent DNAPL dissolution and aqueous transport dynamics. An increasing number of studies have investigated the effects of subsurface heterogeneity on the fate of DNAPL; however, previous work was limited to the examination of the behavior of single-component DNAPL in systems with simple and well-defined aqueous and solid surface chemistry. From a DNAPL remediation point of view, such an idealized assumption will bring a large discrepancy between the designs based on the model simulation and the reality. The research undertaken in this study seeks to stochastically explore the influence of spatially variable porous media on DNAPL entrapment and dissolution profiles in the saturated groundwater aquifer. A 3D, multicomponent, multiphase, compositional model, UTCHEM, was used to simulate natural gradient water flooding processes in spatially variable soils. Porosity was assumed to be uniform or simulated using sequential Gaussian simulation (SGS) and sequential indicator simulation (SIS). Soil permeability was treated as a spatially random variable and modeled independently of porosity, and a geostatistical method was used to generate random distributions of soil permeability using SGS and SIS (derived from measured grain size distribution curves). Equally possible 3D ensembles of aquifer realizations with spatially variable permeability accounting of physical heterogeneity could be generated. Tetrachloroethene (PCE) was selected as a DNAPL representative as it was frequently discovered at many contaminated groundwater sites worldwide, including Thailand. The randomly generated permeability fields were incorporated into UTCHEM to simulate DNAPL source zone architecture under 96-L hypothetical PCE spill in heterogeneous media and stochastic analysis was conducted based on the simulated results. Simulations revealed considerable variations in the predicted PCE source zone architecture with a similar degree of heterogeneity, and complex initial PCE source zone distribution profoundly affected PCE recovery time in heterogeneous media when subject to natural gradient water flush. The necessary time to lower PCE concentrations below Thai groundwater quality standard ranged from 39 years to more than 55 years, suggesting that spatial variability of subsurface formation significantly affected the dissolution behavior of entrapped PCE. The temporal distributions of PCE saturation were significantly altered owing to natural gradient water flush. Therefore, soil heterogeneity is a critical factor to design strategies for characterization and remediation of DNAPL contaminated sites. The systematic and comprehensive design algorithm developed and described herein perhaps serves as a template for application at other DNAPL sites in Thailand.     

Electrokinetic remediation of perchloroethylene-contaminated soil

One large group of persistent and toxic contaminants is the hydrophobic organic contaminants. Among them, perchloroethylene (PCE) has been recognized as a representative group of these pollutants with low solubility. This study reports on the effects of electrokinetic remediation with non-ionic surfactant on PCE-contaminated soil. The performance of electrokinetic process was investigated in the treatment of clay soil that artificially contaminated with two levels: 10,000 and 30,000 mg/kg PCE and 0.33 g/kg Triton X-100. A DC power supply with electric voltage (1 V/cm) was used for 8–16 days. A negatively charged soil surface resulted in a more negative zeta potential and greater electroosmotic flow toward the cathode. The PCE was measured after extraction using n-hexane and analyzed by Fourier transform infrared spectroscopy instrument. The water content of soil was kept 25 % (w/w). Results were shown that PCE removal efficiency achieved was 74 and 89 % for 10,000 and 30,000 mg/kg PCE, respectively, for 16 days. Therefore, in this study, the integration of electrokinetic with non-ionic surfactant as a hybrid method was most effective for the remediation of PCE-contaminated soils.     

EK-Fenton process for removal of phenanthrene in a two-dimensional soil system

Feasibility of electrokinetic (EK) process combined with Fenton-like reaction was investigated for the removal of phenanthrene in a two-dimensional cell. Sandy soil and bentonite were selected as a model soil and a filling material to inhibit the leak of electrolyte solution within the electrode reservoirs into the soil by hydraulic pressure difference, respectively. The effects of parameters including current, electroosmotic flow (EOF), electrolyte pH, and moisture content on the removal efficiency were examined under constant voltage.

At the end of operation for 21 days, the concentration of phenanthrene near the anode was lower than the other positions of soil specimen and increased gradually towards the cathode because hydrogen peroxide solution was supplied from anode to cathode region following the direction of EOF. The concentration of phenanthrene at the bottom soil was lower than that at the top soil. Because capillary attraction in the sandy soil with high porosity was too low to maintain appropriate moisture at the top of the cell, EOF moved through the bottom soil with higher moisture content. Overall removal efficiency at 140 V was 81.6%, which was higher than 68.9% at 100 V because total EOF increased by a factor of 1.5 upon increase of the voltage from 100 to 140 V. In addition, power consumptions at 100 and 140 V were 7.2 and 19.4 kWh, respectively.