Use this URL to cite or link to this record in EThOS:
Title: Modelling of transport in highly heterogeneous porous media, with application to the flushing of waste
Author: Woodman, Nicholas Daniel
ISNI:       0000 0004 2745 0554
Awarding Body: University of London
Current Institution: University College London (University of London)
Date of Award: 2008
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
This thesis is concerned with predicting the fate of contaminants using tracer and contaminant flushing datasets in otherwise poorly characterised heterogeneous media. A deductive approach towards constraining prediction uncertainty is adopted, by narrowing a 'pool' of plausible process models and associated parameter sets. Mathematical comparison of simple transport models is used to allow enhanced diagnosis. Transport models are put into a framework. Then they are compared using commonly used measures of breakthrough curves (least squares fitting, method of moments and late-time concentration gradients). Experimental data and models for transport in waste are then reviewed using this framework. Two new tracer datasets, at laboratory and lysimeter scales, are analysed and modelled. Strong evidence is shown of 'dual-porosity' diffusive exchange effects being an important component in transport in municipal solid waste. However, it is not possible to discriminate between contributions to dispersion through advective and diffusive effects. Approaches to enhance the diagnostic capability of future waste experiments are developed. One possible 'entrant' to the 'pool' of models is examined: a stochastic channel network model. Histograms of nodes and average heads are evaluated at different distances from a range of boundary conditions for large ensembles revealing key insights into network flow behaviour near different types of boundary. Key properties of the network are related to typical continuum analyses, including dimensionality of flow and genesis of apparent 'skins' near boundaries. 'Classical' percolation networks (with zero spatial correlation) are also analysed, revealing new insights into their network properties within finite boundaries. In summary, progress towards predicting the fate of contaminants in highly heterogeneous systems is made through a thorough delineation of the behaviour of simple models in tandem with adoption of a fertile approach to adding new conceptual models. The increased rigorousness of using multiple diagnostic criteria in conjunction with adopting a philosophy of working with multiple working hypotheses is commended as a methodology that provides a sound basis from which to employ multiple-model uncertainty analysis for predictive purposes in the future.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available