THOMAS Hadrien
Abstract: France, like many countries, is facing the need to reduce the carbon dioxide production related to the energy consumption, particularly for heating. The Paris Basin is densely populated in the Île-de-France region and has interesting potential reservoirs for low-enthalpy geothermal energy exploiting geothermal waters between 300 m and 2500 m depth, for district heating. The carbonate formations of the Middle Jurassic located between 1200 m and 1800 m are a main target. However, the high density of wells exploiting this reservoir and the recent failure of a well near Evry show that there is a non-negligible and unquantified risk of low water flow or interference between geothermal doublets in urban areas with a high density of infrastructures. This geological risk constitutes an obstacle to the future development of geothermal energy in Île-de-France. Optimizing the use of geothermal energy requires an increased knowledge of the heterogeneity of the reservoir in terms of porosity/permeability or reservoir connectivity. The objective of the thesis is to build a detailed 3D geological model of sedimentological, diagenetic and petrophysical heterogeneities integrating all available well data and allowing to reduce geological risk. A database of logging and petrophysical data from 168 wells was compiled. This database was supplemented with 129 m of core samples and 191 thin sections. The result is a high-resolution 3D model (12.2 million cells of 100 m x 100 m sides and 3.7 m height, proposing the spatial distribution of facies, porosity and permeability. Thin section studies show that diagenesis plays an important role in the heterogeneity of porosity, in particular the development or not of blocking calcite, or dolomitization and dissolution phenomena. An outcrop analogue provided a better understanding of the distribution of facies in a well constrained sequence stratigraphy framework (3rd order sequences). The use of a drone to photograph a quarry in the Yonne (sub-kilometer scale) coupled with photogrammetric modelling enabled work on a digital outcrop, analogous to the geothermal reservoir. From this digital outcrop, a methodology was developed to work on a geomodeller, and thus to better understand the distribution of facies and petrophysical properties at a finer scale. The 3D modelling of the reservoir in a part of Île-de-France and of an outcrop analogue allows to better predict and understand the distribution of the reservoir properties and to de-risk future operations (such as Grigny near Evry).