Reservoir simulations of H2S production during a SAGD process on a meandering fluvial reservoir taking into account both the aquathermolysis effects and the thermal conductivity.
Abstract
Nowadays, EOR methods such as thermal techniques are widely used to recover theviscous hydrocarbons from heavy oils and bitumen reservoirs. One of the thermal methods isthe Steam-Assisted Gravity Drainage (also called SAGD) consists in injecting steam throughthe injector well into the reservoir to melt the viscous oil and allow its mobility. The meltedoil falls by gravity and is produced by the production well. The hot steam injected, in contactwith heavy oils/bitumen, creates chemical reactions called aquathermolysis reactions. Thesereactions generate gases such as the hydrogen sulfide (H2S) or carbon dioxide (CO2). The H2Sis known to be a highly toxic and corrosive, therefore, it needs to be given a particularattention when it is produced at the surface. Reservoir models have been built to simulatethermal effects during a SAGD but only few publications in the literature deal with theaquathermolysis reactions occurring in the reservoir where steam is injected.
This report will focus on building a reservoir simulation model in order to forecast the H2Sproduction. This model is built based on a PVT description of the heavy oil/bitumen and on anew sulfur-based kinetic model. A literature study was performed to build up the generalunderstanding of the physical and chemical mechanisms occurring during a steam injectionprocess. A description of the heaviest components found in heavy oils/bitumen is madethrough a SARA analysis. A new sulfur-based kinetic model recently developed waspresented and used in the reservoir model to reproduce the aquathermolysis reactions.
A large part of this thesis is dedicated to the topic of building a PVT model for heavy oils.Throughout this chapter, the PVT (Pressure-Volume-Temperature) model is built based oncorrelations specific for heavy oils. This PVT describes the behavior of the fluid (oil) usedduring reservoir simulations.
Then, a chapter presents the reservoir models used for the simulations. Reservoir simulationswere run using both the 2D and 3D models in order to compare the simulation results. A 3Dreservoir model is presented to get to know the main characteristics of this heterogeneousreservoir. This model reproduces the Hangingstone heavy oil field in Canada.
Finally, these simulations results are then presented, such as the oil and gas productions, thesteam injection, the Steam-Oil Ratio as well as the H2S production. A sensitivity analysis wasalso performed in order to see how some parameters, such as the pressure/temperature (at thesaturation conditions) of the steam injected or the rock conductivity, would affect the H2Sproduction or the oil production.
The different simulations presented in this report show consistent results in terms of H2Sproduction at surface for different steam injection pressures. This shows that the fluiddescription and the kinetic model used in the study are relevant for acid gases prediction inthe context of steam injection reservoir simulations.