Natural micro-scale heterogeneity induced solute and nanoparticle retardation in fractured crystalline rock

详细信息	查看全文 | 推荐本文 |

• 作者：F. Huber^a ; ^b ; ^{florian.huber@kit.edu} ; F. Enzmann^c ; A. Wenka^d ; M. Bouby^a ; M. Dentz^e ; T. Schä ; fer^a ; ^b
• 关键词：Colloids ; Computational fluid dynamics ; Residence time ; Non-Fickian transport
• 刊名：Journal of Contaminant Hydrology
• 出版年：2012
• 期刊代码：43_01697722
• 类别：env
• 出版时间：15 May, 2012
• 卷：133
• 期：Complete
• 页码：40-52
• 文件大小：1119 K

摘要

We studied tracer (Tritiated Water (HTO); Tritium replaces one of the stable hydrogen atoms in the Hb>2b>O molecule) and nanoparticle (quantum dots (QD)) transport by means of column migration experiments and comparison to 3D CFD modeling. Concerning the modeling approach, a natural single fracture was scanned using micro computed tomography (渭CT) serving as direct input for the model generation. The 3D simulation does not incorporate any chemical processes besides the molecular diffusion coefficient solely reflecting the impact of fracture heterogeneity on mass (solute and nanoparticles) transport. Complex fluid velocity distributions (flow channeling and flowpath heterogeneity) evolve as direct function of fracture geometry. Both experimental and simulated solute and colloidal breakthrough curves show heavy tailing (non-Fickian transport behavior), respectively. Regarding the type of quantum dots and geochemical conditions prevailing (Grimsel ground water chemistry, QD and diorite surface charge, respectively and porosity of the 脛sp枚 diorite drill core) experimental breakthrough of the quantum dots always arrives faster than the solute tracer in line with the modeling results. Besides retardation processes like sorption, filtration, straining or matrix diffusion, the results show that natural 3D fracture heterogeneity represents an important additional retardation mechanism for solutes and colloidal phases. This is clearly verified by the numerical simulations, where the 3D real natural fracture geometry and the resulting complex flow velocity distribution is the only possible process causing solute/nanoparticle retardation. Differences between the experimental results and the simulations are discussed with respect to uncertainties in the 渭CT measurements and experimental and simulation boundary conditions, respectively.