Lectura de Tesis Doctoral
Mechanisms and stochastic dynamics of transport in Darcy-scale heterogeneous porous media
Thesis advisors: Dr. Marco Dentz / Dr. Daniel Fernández García (UPC)
The defense will take place:
thursday, june 14th 2018, 11:00
UPC, Campus Nord
Building C1. Classroom: 002
C/Jordi Girona, 1-3
a cargo de: Arash Massoudieh (Catholic University of America)
Fecha: Martes 12 de Junio a las 12:15 h Lugar: Departamento de Ingeniería Civil y Ambienta, Modulo D2-Aula CIHS, Planta Baja
In this presentation GIFMOD, a new flexible and user-friendly modeling tool for forward and inverse modeling of environmental processes in the surface, subsurface, and the vadose zone will be introduced. GIFMOD can be used to model flow, particle transport and reactive transport in systems composed of streams, ponds, soil, groundwater, catchments, and plants. Transport of multiple classes of particles undergoing settling, resuspension, filtration, and remobilization can also be modeled. The model also allows prediction of transport and reaction of user defined water quality constituents in the system based on user-provided reaction networks and biokinetics rate expressions. Plants can be considered as individual blocks uptaking water and chemicals through their root system. The modeling framework allows users to represent the system they intend to model with the desired level of complexity and only include the processes they deem essential in the processes. GIFMOD also has a built-in parameter estimation capability both for doing point estimates using a hybrid genetic algorithm and probabilistic parameter estimation using Markov chain Monte Carlo algorithm. The flexibility of the tool allows it to be applied to a wide range of systems ranging from stream networks, stream-catchment systems, surface water-groundwater interaction, batch and column experiments, best management practices, and groundwater flow and reactive transport among others. I will demonstrate the utility of GIFMOD to simulate flow and transport in a stream, bioretention, infiltration basin and permeable pavement GI systems.
a cargo de: Tomás Aquino (IDAEA-CSIC)
Fecha: Jueves 31 de Mayo a las 12:15 h
Lugar: Departamento de Ingeniería Civil y Ambiental, Modulo D2-Aula CIHS, Planta Baja
Accurately simulating reactive transport through heterogeneous media requires resolving the spatial scales at which mixing takes place. This typically requires a fine spatial discretization (Eulerian methods) or large numbers of particles (Lagrangian methods), leading to prohibitively expensive simulations for large-scale transport. We explore a different approach and consider the question: In heterogeneous chemically reactive systems, is it possible to describe the evolution of macroscopic reactant concentrations without explicitly resolving the spatial transport? Traditional Kinetic Monte Carlo methods, such as the Gillespie algorithm , model chemical reactions as random walks in particle number space, without the introduction of spatial coordinates. The inter-reaction times are exponentially distributed under the assumption that the system is well mixed. In real systems, transport limitations lead to incomplete mixing and decreased reaction efficiency. We introduce an arbitrary inter-reaction time distribution, which may account for the impact of incomplete mixing. The resulting process defines an inhomogeneous continuous time random walk in particle number space, from which we derive a generalized chemical master equation and formulate a generalized Gillespie algorithm . We then determine the modified chemical rate laws for different inter-reaction time distributions, which describe the macroscopic reaction kinetics. We trace Michaelis–Menten-type kinetics back to finite-mean delay times, and predict time-nonlocal macroscopic reaction kinetics as a consequence of broadly distributed delays. Non-Markovian kinetics exhibit weak ergodicity breaking and show key features of reactions under local non-equilibrium.
a cargo de: Arnau Canelles García
Día: Jueves 17 de Mayo a las 12:15 h
Lurgar: Departamento de Ingeniería Civil y Ambienta, Modulo D2-Aula CIHS, Planta Baja
Managed Aquifer Recharge (MAR) can be affected by a series of events that can be considered an issue for its operation. Those events represent different parts of the operation of a MAR facility like recharge, water quality and quantity, engineering works failure, etc. These events can be classified into different issue groups: technical (quality, quantity, specific targets and structural damages) and non-technical (social, economic, legislation and governance). The combination of those events can be used to quantify the risk of failure of the MAR facility. In this project, risk is quantified by means of a Fault Tree within a Probabilistic Risk Assessment (PRA) framework. The Fault tree developed consist of 65 events applicable to the operation phase of a MAR facility. That combination of events can lead to a failure (partial in the case of this work) of the facility which probability represents the risk. This methodology has been applied to six different MAR sites located in five countries in the Mediterranean Basin. The probabilities of occurrence for each event have been defined by “expert criteria”, provided by each manager of the studied facilities. We can conclude that for all sites, the perception from experts about the non-technical aspects is similar or even higher than those of the technical ones. Regarding risk values, in three out of six sites the probability of failure evaluated has exceeded 90%, while the other three sites, presented lower risks (75%, 29% and 18%) those differences in values are related mainly to each site’s characteristics.