Baptiste Lepillier

In 2020, as the world Energy demand keeps on rising (International Energy Agency (IEA), 2019), and with the global climate warming a reality (The Organisation for Economic Co-operation and Development (OECD), 2020), reducing our societal impact on Earth is of utmost importance. Energy and Climate have always been intrinsically related. Therefore, solving the Energy-Climate problem is a challenge where not one but several solutions should come together. Part of this global solution is the potential of geothermal resources. Geothermal energy is a renewable energy resource which has large potential to reduce the dependency on fossil fuels.

Within the several uses of geothermal resources, a promising technique is titled Enhanced Geothermal Systems. More than renewable, this method has the potential to be sustainable. EGS consists of an originally low permeability reservoir rock that is artificially enhanced. The enhancement can be achieved by different stimulation techniques, such as mechanical, chemical, thermal or a combination of all.

This thesis focuses on the mechanical EGS stimulation, where opening of existing fractures and creation of new ones is achieved by injecting a pressurized fluid in the reservoir rock formation. Such a process results in propagating a hydraulic fracture. The complexity of the EGS technique stands in predicting the hydraulic fracture propagation phenomena.

EGS research and development is part of the GEMex goals. The GEMex project is a collaboration between Mexican institutions and the European Commission, dedicated to the development of non-conventional geothermal techniques. The Acoculco geothermal field, located in Puebla, is foreseen as a potential EGS. Because this field has been explored with two geothermal wells, and because an analogue exhumed system is available nearby, in the Las Minas area, this system constitutes a great research site for developing knowledge on EGS.

In this thesis, I propose an integrated workflow for evaluating the EGS potential of a reservoir rock formation. The workflow starts with Chapter 2, introducing a novel method used to predict the fracture network in the subsurface, by computing Discrete Fracture Networks (DFN) from outcrop data. More specifically, a compilation of scanline survey datasets is acquired at the field. This dataset is analyzed using the newly written Python TM script SkaPy. SkaPy helps compiling Training Images which can then be extrapolated into the field scale using the Multiple Points Statistic method.

In Chapter 3, I describe the rock physics laboratory work realized to characterize the rock properties. These measurements span from mineral analyses to original hydraulic fracture experimental setup. All these measurements are applied to rock samples collected in Acoculco and Las Minas analogue outcrops. The characterization is used to predict the behavior of the potential reservoir rocks during the hydraulic fracture propagation process.

Continuing on fracture characterization, Chapter 4 presents a model built to evaluate the extent to which fractures in the subsurface can influence the reservoir thermal and fluid flow. As the Acoculco geothermal field enhancement could be achieved in different reservoir formations, the model handles a multi-reservoir analysis. And because the fracture distribution within a reservoir formation is not homogeneous, the model compares multi-production scenarios.

Moving on the prediction of the hydraulic fracture propagation, Chapter 5 approaches the question by modeling a hydraulic fracturing process using a perfectly planar 3 dimensional (3D) fracture. In this chapter I analyze the influence of different parameters involved in a hydraulic fracturing process to identify the most influencing ones. Then, using the Acoculco specific geothermal case study, I calculate the relation between the total injected volume and the hydraulic fracture dimensions.

Because the propagating hydraulic fracture can be influenced by reservoir heterogeneities, such as natural pre-existing fractures, Chapter 6 introduces a novel approach. Here, I propose to build on the classical hydraulic fracture method, by adding the reservoir heterogeneities related to pre-existing natural fractures. The model uses the variational phase-field, a smeared representation of the damage, thus the hydraulic fracturing process and its propagation over time are represented using a smooth transition function instead of a sharp interface implemented. Discretization uses a non-conforming mesh, which enables complex geometries. Hence, I can solve the hydraulic fracture propagation process accounting for its interaction with pre-existing natural fractures. The model is applied to the three potential reservoirs studied at the Acoculco geothermal field.

In Chapter 7, I apply the entire workflow to the Marble reservoir formation of the Acoculco Geothermal system. This chapter offers a review of the workflow and tests its application to a real case, by observing the modeled enhancement of the Marble formation.

Finally, in Chapter 8 I discuss the extent to which this workflow could be used in deploying EGS over the world.