Abstract

This study is part of our investigation to translate results from the laboratory scale to match the field pilot results. As the unconventional shale development matures, the industry has been actively seeking new ways to unlock incremental value beyond primary depletion. In particular, the miscible gas injection via the huff-and-puff technique has shown promising results from the pilot tests. However, the pilot tests typically show lower recoveries than originally predicted by laboratory work and simulation studies.

This paper extends the previous work and utilizes the Lorenz/Cauchy distribution to investigate the impact of fracture spacing on gas injection scenarios with and without soaking times. The mathematical model previously developed represents the infiltration of the injected gas into the shale matrix by modeling mass diffusion of a limited volume of well-stirred fluid into a solid body (remaining injected gas in the fracture network at the end of the injection phase as compressed gas) into a porous medium (matrix). To better represent field conditions, the matrix is characterized as an ensemble of rock pillars separated by fracture discontinuities randomly distributed, and the recovery was explored as a function of the micro-fracture distribution.

Although diffusion is the main transport and recovery mechanism, this study found that the fracture geometry created near-wellbore, i.e., fracture spacing & distribution, has a first-order effect on the efficiency of the huff-and-puff process in the field. It was also observed that by varying the soaking times of each cycle, the issue of penetration length could be resolved (as it increases as a function of √time). Additionally, understanding the near-wellbore fracture geometry would help operators optimize their gas injection schemes.

The updated upscaling equation will help understand the huff-and-puff process better and predict the expected recoveries in the field more accurately. Additionally, it would help operators adjust and optimize soaking times for the process using a mechanistic approach.

This content is only available via PDF.
You can access this article if you purchase or spend a download.