arXiv subjects

Alberto Guadagnini

Publications and source records attributed to Alberto Guadagnini.

8 recordsLinked to original sources

arXiv2026

Probabilistic Inverse Modeling of Contaminant Transport via a Conditioned-on-Design Bayesian Physics Informed Neural Network

We address the inverse problem of reactive transport in heterogeneous porous media, where unknown model parameters must be inferred from sparse experimental observations. The problem is complicated by strong nonlinearities, spatial heterogeneity, and limited data availability. We propose a Conditioned-on-Design Bayesian Physics-Informed Neural Network (CoDe-BPINN), which combines a domain-decomposed PINN solver with a Bayesian inference network that learns the conditional distribution of model parameters given experimental design variables. The framework is trained by maximizing a physics-informed Evidence Lower Bound (ELBO), enabling simultaneous reconstruction of spatiotemporal concentration fields, probabilistic parameter estimation, and uncertainty quantification. We demonstrate the approach using laboratory experiments on contaminant transport through a multilayer porous column with an iodinated contrast medium. The model accurately reproduces breakthrough dynamics while revealing systematic parameter dependence on flow conditions. In particular, it identifies a nonlinear decrease in effective sorption capacity with increasing flow rate and porosity, consistent with kinetic limitations and reduced adsorbent mass. The Bayesian formulation also uncovers a strong negative correlation between sorption affinity and sorption capacity, quantifying the intrinsic non-identifiability of the inverse problem. CoDe-BPINN provides a robust framework for parameter inference and uncertainty quantification in data-scarce reactive transport problems.

physics.comp-ph↗

arXiv2026

Pore-shape and its spatial organization control intrinsic permeability of porous media

The structure of a porous material, and in particular its spatial variability, is known to control the intrinsic permeability of the system. We investigate how dead-end pores influence the intrinsic permeability of a porous medium beyond their contribution to total pore volume. Dead-end pores are ubiquitous in porous media, yet they are often treated as hydraulically inactive regions whose influence is assumed to be negligible or absorbed into effective-porosity descriptions. We perform pore-scale flow simulations across different dead-end pore structures, including heterogeneous arrangements, controlled granular assemblies, and a minimal single-channel model to study their impact on the system macroscopic permeability. This strategy allows us to isolate the effects of dead-end pore density, depth, and orientation while preserving the transmitting network. We find that dead-end pores can influence intrinsic permeability: increasing the density of dead-end pores along percolating flow paths enhances permeability, whereas pore depth and junction orientation have negligible effects. The observed permeability enhancement originates from localized hydrodynamic interactions at junctions between transmitting and dead-end pores. Based on these results, we propose an effective formulation that relates the density and spatial organization of dead-end pores relative to the transmitting network to macroscopic permeability. Our findings show that dead-end pore architecture provides an additional geometric control on intrinsic permeability beyond porosity and pore-size statistics.

cond-mat.soft↗

arXiv2026

Physics Informed Differentiable Solvers for Learning Parametric Solution Manifolds in Heterogeneous Physical Systems

Learning the full family of solutions to parameterized partial differential equations (PDEs) is a central challenge to our ability to model the behavior of heterogeneous systems, with a variety of fundamental and application-oriented implications in fields such as hydrogeology where system properties exhibit significant (and often uncertain) spatial heterogeneity. We address this by reformulating a Physics-Informed Neural Network (PINN) as a differentiable solver that learns the continuous solution manifold for steady-state Darcy flow. Our framework requires only a single training run, circumventing the need for costly re-training for each new parameter instance. Its versatility is demonstrated through two representations of spatially heterogeneous hydraulic conductivity fields: a direct analytical form and a novel data-driven formulation resting on an autoencoder to create a low-dimensional latent encoding. A key innovation is the integration of the differentiable decoder into the physics-informed loss function, enabling on-the-fly reconstruction of complex conductivity fields via automatic differentiation. The approach yields accurate, mass-conserving flow solutions and supports efficient uncertainty quantification, providing a general methodology for physics-constrained data-driven modeling of heterogeneous systems.

physics.comp-ph↗

arXiv2025

Spatial organization of biomass controls intrinsic permeability of porous systems

Biofilms in porous media critically influence hydraulic properties in environmental and engineered systems. However, a mechanistic understanding of how microbial life controls permeability remains elusive. By combining microfluidics, controlled pressure gradient and time-lapse microscopy, we quantify how motile and non-motile bacteria colonize a porous landscape and alter its resistance to flow. We find that while both strains achieve nearly identical total biomass, they cause drastically different permeability reductions - 78% for motile cells versus 94% for non-motile cells. This divergence stems from motility, which limits biomass spatial accumulation, whereas non-motile cells clog the entire system. We develop a mechanistic model that accurately predicts permeability dynamics from the pore-scale biomass distribution. We conclude that the spatial organization of biomass, not its total amount, is the primary factor controlling permeability.

physics.bio-ph↗

arXiv2024

Modelling parametric uncertainty in PDEs models via Physics-Informed Neural Networks

We provide an approach enabling one to employ physics-informed neural networks (PINNs) for uncertainty quantification. Our approach is applicable to systems where observations are scarce (or even lacking), these being typical situations associated with subsurface water bodies. Our novel physics-informed neural network under uncertainty (PINN-UU) integrates the space-time domain across which processes take place and uncertain parameter spaces within a unique computational domain. PINN-UU is then trained to satisfy the relevant physical principles (e.g., mass conservation) in the defined input domain. We employ a stage training approach via transfer learning to accommodate high-dimensional solution spaces. We demonstrate the effectiveness of PINN-UU in a scenario associated with reactive transport in porous media, showcasing its reliability, efficiency, and applicability to sensitivity analysis. PINN-UU emerges as a promising tool for robust uncertainty quantification, with broad applicability to groundwater systems. As such, it can be considered as a valuable alternative to traditional methods such as multi-realization Monte Carlo simulations based on direct solvers or black-box surrogate models.

physics.data-an↗

arXiv2024

Intrinsic permeability of heterogeneous porous media

Providing a sound appraisal of the nature of the relationship between flow $(Q)$ and pressure drop $(ΔP)$ for porous media is a long-standing fundamental research challenge. A wide variety of environmental, societal and industrial issues, ranging, e.g., from water-soil system remediation to subsurface energy optimization, is affected by this critical issue. While such dependence is well represented by the Kozeny-Carman formulation for homogeneous media, the fundamental nature of such a relationship ($Q$ vs $ΔP$) within heterogeneous porous systems characterized by a broad range of pore sizes is still not fully understood. We design a set of controlled and complex porous structures and quantify their intrinsic permeability through detailed high quality microfluidics experiments. We synthesize the results upon deriving an original analytical formulation relating the overall intrinsic permeability of the porous structure and their key features. Our formulation explicitly embeds the spatial variability of pore sizes into the medium permeability through a conceptualization of the system as a collection of smaller scale porous media arranged in series. The resulting analytical formulation yields permeability values matching their experimentally-based counterparts without the need of additional tunable parameters. Our study then documents and supports the strong role played by the micro-structure on the overall medium permeability.

physics.flu-dyn↗

arXiv2022

Quantification of CO2 generation in sedimentary basins through Carbonate Clays Reactions with uncertain thermodynamic parameters

We develop a methodological framework and mathematical formulation which yields estimates of the uncertainty associated with the amounts of CO2 generated by carbonate-clays reactions (CCR) in large-scale subsurface systems to assist characterization of the main features of this geochemical process. Our approach couples a one-dimensional compaction model, providing the dynamics of the evolution of porosity, temperature and pressure along the vertical direction, with a chemical model able to quantify the partial pressure of CO2 resulting from minerals and pore water interaction. The modeling framework we propose allows (i) estimating the depth at which the source of gases is located and (ii) quantifying the amount of CO2 generated, based on the mineralogy of the sediments involved in the basin formation process. A distinctive objective of the study is the quantification of the way the uncertainty affecting chemical equilibrium constants propagates to model outputs, i.e., the flux of CO2. These parameters are considered as key sources of uncertainty in our modeling approach because temperature and pressure distributions associated with deep burial depths typically fall outside the range of validity of commonly employed geochemical databases and typically used geochemical software. We also analyze the impact of the relative abundancy of primary phases in the sediments on the activation of CCR processes. As a test bed, we consider a computational study where pressure and temperature conditions are representative of those observed in real sedimentary formation. Our results are conducive to the probabilistic assessment of (i) the characteristic pressure and temperature at which CCR leads to generation of CO2 in sedimentary systems, (ii) the order of magnitude of the CO2 generation rate that can be associated with CCR processes.

physics.geo-ph↗

arXiv2013

Origins of Anomalous Transport in Disordered Media: Structural and Dynamic Controls

We quantitatively identify the origin of anomalous transport in a representative model of a heterogeneous system---tracer migration in the complex flow patterns of a lognormally distributed hydraulic conductivity ($K$) field. The transport, determined by a particle tracking technique, is characterized by breakthrough curves; the ensemble averaged curves document anomalous transport in this system, which is entirely accounted for by a truncated power-law distribution of local transition times $ψ(t)$ within the framework of a continuous time random walk. Unique to this study is the linking of $ψ(t)$ directly to the system heterogeneity. We assess the statistics of the dominant preferred pathways by forming a particle-visitation weighted histogram $\{wK\}$. Converting the ln($K$) dependence of $\{wK\}$ into time yields the equivalence of $\{wK\}$ and $ψ(t)$, and shows the part of $\{wK\}$ that forms the power-law of $ψ(t)$, which is the origin of anomalous transport. We also derive an expression defining the power law exponent in terms of the $\{wK\}$ parameters. This equivalence is a remarkable result, particularly given the correlated $K$-field, the complexity of the flow field and the statistics of the particle transitions.

physics.flu-dyn↗