This site uses cookies and third-party cookies to provide you with a better service. This site keeps information concerning your visit and users' preferences. If you continue browsing the site we will understand that you have accepted their use.

A MUltiple Space and Time scale approach for the quAntification of deep saliNe formations for CO2 storaGe

Work package 7 Numerical Model Development and Modeling

WP leader
UNOTT
Participants
UU EWRE CSIC IIT UCAM

Objectives
  
The overall objective of this WP is to provide a comprehensive modeling approach and associated numerical tools for the simulation of flow, transport, reactive transport, thermal and chemical processes occurring to the fluids and to the rock matrix during the injection and storage of CO2. In particular,
  • Adopt the best available THMC models for the simulation of the CO2 flow and transport processes in deep saline formations and update them according to project findings
  • Apply these models to the validation experiment in Heletz and to the test sites, firstly, to validate the models and second, to get an understanding of the site performance, to be used as input in terms of ICE and risk assessment;
  • Update existing codes by incorporating innovative numerical schemes able to quantify the impacts on seal integrity resulting from the injection of CO2, heterogeneity effects

Recent Model developments, some examples (April 2012)


Figure. Example - Adding chemistry to Code-Bright. Saaltnik et al, 2012



Figure. Example: two-phase flow and transport with time dependent tracer reaction/partitioning between the phases (UU, Tong et al. 2012)


Work progress (March 2011)
  
Numerical modeling capabilities related to the existing code have been enhanced. Majority of the assessment simulations are carried out with the established codes TOUGH2 and CODE_BRIGHT.
  • The CODE_BRIGHT code has been extended to supercritical CO2 density and viscosity. In addition, hydro-mechanical coupling has been included. To improve the computational performance, an Object Oriented Numerical Model Framework for Multiphase Flow and Reactive Transport Processes, PROOST++ (Process Object Optimization and Simulation Tool) is in progress. See more outcomes at Hydromechanical Characterization Test for CO2 Sequestration in Deep Saline Aquifers
  • The TOUGH2 model has been implemented to model the heterogeneity effects in a stochastic Monte Carlo framework. Development to use Gaussian emulators to replace the full Monte Carlo approach is underway. Modeling of heterogeneity effects for CO2 spreading is computationally extensive and emulators are a promising approach to replace the full MC analysis. (see ‘Gaussian Process Emulators for Quantifying Uncertainty in CO2’)
  • Model development related to the existing codes includes also incorporation of the novel (at the interface reactive) tracers to simulators of multiphase flow and transport is underway. (see ‘Modelling of novel tracers in a two-phase flow system’)
In terms of simulation of the MUSTANG test sites Generic model analyses have considered
  • Hydromechanical modeling to study possible reactivation/creation of fractures in the caprock, due to overpressure during CO2 injection. Initial stress field proves to have a great effect on the failure mechanisms, and possible opening of preferential paths for CO2. (see progress at Hydromechanical Characterization Test for CO2 Sequestration in Deep Saline Aquifers)
  • Effect of dispersion on the onset of density driven fingering during CO2 sequestration. Dissolution of CO2 causes an unstable high-density diffusive front which will affect the dissolution. Accounting for this, causes a significant reduction on the onset time, implying that CO2 dissolution can be accelerated by activating dispersion, for example by various (fluctuating) injection schemes.(see ‘Mass Transfer Limitations and Non-Locality in Large Scale Reactive Transport’)
  • Numerical modeling for moving phase interface, by combining a multiphase flow model with an analytical moving front tracking algorithm, and application to supercritical CO2 forcing its way into a heterogeneous caprock
  • Effect of outer boundary conditions of the storage aquifer has been addressed (‘Analysis of boundary conditions in numerical simulations’)
Specific model development is carried out for
  • Modeling crack evolution of multiple cracks. RBF meshless methods based on collocation can describe non-homogenous poroelasticity equation with high accuracy. However, they cannot directly handle singularities within the solution domain. Since the evolution of crack depends strongly on behaviour at the crack-tip (a singularity), techniques are needed to include the crack tip. A new mixed approach, fully meshless, is under development, to address this by combining RBF collocation scheme at the local level with integral representational formula at the global level. (‘Numerical analysis of the Coupling effect between dissolution at the interface and fingering evolution’)
  • LH-RBF numerical method has been developed to solve nonlinear heat conduction problems, which is relevant in order to consider changes of phase of the injected sc CO2. During these phase changes, dramatic variations occur in the density, heat capacity and thermal conductivity (ρ, Cp and κ), making the problem strongly nonlinear.

©Mustang - 2009