After having discussed how geoscientists and geomodelers can efficiently collaborate, this book is now reviewing the place engineers can take in geomodeling projects.
This chapter focuses on the collaboration between reservoir engineers and geomodelers around the topic of flow simulation. The next two chapters will look at geomodeling for reserve estimates and for production engineering respectively.
Many geomodels are built at the request of reservoir engineers. They are in need of a 3D grid, capturing the characteristics of the rocks (porosity, SW, permeability) to feed to their flow simulation software. In fact, this is largely what motivated the development of geomodeling in the first place and several decades ago. Geomodeling can be seen as a bridge between geoscientists and engineers (Figure 1). This places geomodeling at the intersection of two worlds which have always had difficulties communicating one with the other. The first part of this paper describes the common communication issues around geomodels. When not handled properly, these issues can be the cause of failure of our projects.
A good 3D geological grid captures all the geological, petrophysical and geophysical information gathered by geoscientists about the reservoir. The geomodeler then transfers the information stores in the geological grid into a new 3D grid fit for the purpose of flow simulation. This new 3D grid is called a flow simulation grid hereafter. The reason why a specific grid is needed is also covered in the first part of this paper. The second part focuses on the technical processes involved (creation of the flow simulation grid as well as upscaling and downscaling of properties).
Transferring the uncertainties from the geomodel study into the flow simulation study is a major challenge. Which geomodels shall we use in flow simulation when potentially hundreds of 3D petrophysical distributions have been created by geostatistical algorithms? Historically, only one was used. Nowadays, several 3D geomodels are sent to flow simulation. In either case, it is up to the geomodeler to help his team to choose which 3D distributions should be used. It requires ranking the geomodels based on criteria relevant to flow simulation. This topic is covered in the last part of this paper.