3D Hydraulic and Natural Fracture Characterization and Simulation

Abstract

ABSTRACT Numerical reservoir models have some uncertainties inherent to them. This is partly because it is impossible to incorporate all complexities the real reservoirs have in nature. Furthermore, it is not uncommon that some of the required data is missing, in which case approximations or guesses are employed. It has been observed that numerical reservoir models with equally decent match with historical production data return significantly different numerical simulation conclusions which begs the obvious question, which conclusions should be taken into account before proceeding to the next step? In this study, a series of numerical reservoir models with different hydraulic fractures geometry are created based on a gas-producing well in an unconventional reservoir. Embedded discrete fracture model (EDFM) is used to embed hydraulic and natural fractures into the structured grid. Results show that the model with longer hydraulic fracture geometry has a significantly higher estimated ultimate recovery (EUR), likely due to it has a bigger total hydraulic fractures area. This suggests that a mistake in determining the most representative numerical reservoir model could lead to overestimation or underestimation of EUR which potentially translates to financial losses. It is recommended to use neighboring wells to determine which hydraulic fractures geometry does a well in question have, so that the potential problem mentioned above can be avoided. The effects of natural fractures presence are also investigated in an attempt to provide insights into how such type of fractures affect the gas production from a low permeability shale reservoir. INTRODUCTION The presence of fractures, in this case those that are manmade, has allowed low permeability shale reservoirs to produce hydrocarbons. Nonetheless, modeling such fractures representative enough to produce "correct" forecast is nontrivial. There are ampleuncertainties related to fractures and its properties in the reservoirs. This study uses a real dataset of a gas-producing well in a deep shale gas reservoir in Sichuan Basin. Three-dimensional discrete fracture network model is added to the model in order to simulate the effect of natural fractures to the well production. History-matched numerical models incorporated with embedded discrete fracture model (EDFM) are then created for simulation purposes. The concept of EDFM is illustrated in Fig. 1. EDFM itself has been argued to be able to accurately model any complex fractures in the reservoir without sacrificing the efficiency of structured grids (Xu et al., 2017, 2019; Sepehrnoori et al., 2020). It has been widely used in modeling complex fractures in unconventional reservoirs that also covers the existence of natural fractures, well spacing optimization with complex fracture hits, automatic history matching, and gas huff-n-puff for enhanced shale oil recovery (Yu and Sepehrnoori, 2018; Xu et al., 2018; Xu and Sepehrnoori, 2019; Ganjdanesh et al., 2019a, 2019b; Tripoppoom et al., 2020a, 2020b). A model has hydraulic fractures geometry greater in height but smaller in length than the other one. Another model has no natural fractures to simulate the effects of natural fractures on the gas production.