Abstract

We present benchmark simulations and a real-field application of a new coupled fracture growth / fluid flow simulator.

The coupled reservoir simulations compare well for various synthetic test scenarios. The application to a real five-spot injection trial demonstrates the importance of the new modelling tool for optimal reservoir management.

Introduction

In many fields secondary recovery is playing an increasingly important role to maintain oil production. The recovery of the remaining oil is a major task that, due to the complexity of many reservoir formations, requires an extensive knowledge of the subsurface. Probably the cheapest and most often applied technique to sweep a reservoir is a waterflooding project.

Injection of water in the reservoir formation is often accompanied by, unnoticed, induced fracture propagation. Dynamically growing or shrinking fractures have a huge impact on the reservoir management strategy and, if not handled properly, may risk the success of the whole water injection process. Hence, dynamic fracture propagation should be an integral part of a standard reservoir simulation study. Then, an optimal waterflood implementation may be achieved which guarantees an overall increase of recoverable oil.

Next to waterflooding are numerous other applications where induced fracture mechanics play a significant role in defining the reservoir behavior. Typical examples are water injections for reservoir pressure maintenance of e.g., green or brown fields, and safe disposal of produced water by re-injection.

In the past, varying approaches have been presented to study fluid flow and fracture growth during reservoir simulations. Implicit and explicit coupling techniques have been developed and tested for varying theoretical and practical scenarios. The main differences are to which extent the interaction and uncertainties of a coupled fluid flow-rock mechanics model are incorporated in the simulation tool. To our knowledge, these tools are not widely used in the industry on a standard basis, since the models need to be purposed-built and, therefore, require an extensive effort to be put into place.

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