Marie-Christine Cacas Stentz

Numerical simulation of sedimentary basins (Schneider et al, 2000) has been developed to assist the geoscientist in building consistent integrated models of the deep subsurface over geologic time and in predicting petroleum systems. Beyond these applications, it seems interesting to ask whether it can also help to address the challenges of hydrogen systems. In this communication, we present some recent developments in numerical basin modeling that tentatively address some of these challenges. A demonstration case is presented and several scenarios about natural hydrogen migration and alteration are considered. To be able to model natural H2 systems, we had to extend and develop new functionalities to simulate migration processes that are usually neglected when considering hydrocarbons: 1. Although already designed for two-phase flow simulation, we needed to implement the exchange of hydrogen between the vapor phase and the aqueous phases. To do this, we rely on an analytical formulation of H2 solubility as a function of salinity, pressure, and temperature, and which was calibrated against the equation of state of Sørreide and Whitson (1992). Using this functionality, H2 exsolution or dissolution along flow paths as thermodynamic conditions change is well represented. 2. Dissolved hydrogen diffusion is also simulated with a diffusion coefficient that depends on porosity, temperature, and facies tortuosity. 3. To model possible influx of H2 by hydrothermal flow, we are now able to impose the influx rate at deep fault roots as well as its dissolved hydrogen concentration. Classical gas production by an order-one degradation process of a source material using an Arrhenius degradation rate is also available in predefined lithologies. 4. Since the degradation kinetics of H2 is also expected to be orders of magnitude faster than that of hydrocarbons, a basic degradation model based on an order-1 process has been implemented according to the model proposed by Rosso (1993), where the degradation occurs only in a temperature range. 5. Finally, our detailed modeling of flow in fault zones which simulates the flow in damage zone cells and across the fault gouge using different discretization cells is particularly well suited to simulating H2 migration which is believed to be largely controlled by flow along faults. Using this updated version of our basin simulator, we developed a demonstration case slightly modified from a 2D section of the Perth Basin in Australia (Frery et al, 2017). This demonstration case shows that predicting hydrogen migration is far from intuitive as it is controlled by many factors, including regional-scale hydrodynamics and thermodynamic conditions, the latter controlling the depth and location of exsolution. This makes basin modeling an essential tool to address natural H2 systems.Co-auteur : Marie-Christine Cacas-Stentz*, Benjamin Braconnier*, Francesco Patacchini*, Clémentine Meiller*, Arnaud Pujol*, Adriana Traby*, Françoise Willien*, Sylvie Wolf*