An Integrated workflow for gas injection EOR and a successful application for a heterogeneous sandstone reservoir in the Southern North Sea{bullet operator}

Nishikiori, Nobuo (Norske Shell A/S) | Sugai, Keiichiro (Arabian Oil Co. Ltd.) | Normann, Clas (Talisman Energy Norge AS) | Onstein, Arne (Talisman Energy Norge AS) | Melberg, Oddbjoern (DONG Norway) | Eilertsen, Terje (DONG Norway)


This study describes an improved engineering workflow to perform technical evaluation and screening of gas injection EOR. A successful case study demonstrates how field data, engineering analysis and simulation are integrated to precisely model gas injection EOR. This workflow can be adaptable for any type of reservoir and can be utilized as a fast-track screening workflow for gas injection EOR.

The target for this study was the Gyda reservoir located in the southern part of Norwegian North Sea in the Norwegian Continental Shelf. The reservoir is of heavily faulted heterogeneous shallow marine sandstone. As the measure of heterogeneity, a Dykstra-Parson's coefficient1 (VDP) of more than 0.8 has been measured from core plug data.

For the purpose of building a tool that can be utilized for gas injection EOR study, a five-step workflow has been implemented:

  1. High resolution vertical geologic modeling to capture heterogeneity
  2. Flow-based rock typing to model fluid displacement in detail,
  3. Application of a commercial optimizer for adjustment of fault transmissibility to assist and accelerate the process of history match
  4. Inclusion of gas miscible modeling with a simplified 1D Todd and Longstaff representation of phase mixing
  5. A combined engineering and economic approach to establish gas injection strategy.

The results of this case study confirmed the capability of the described workflow to model gas injection EOR for the heterogeneous sandstone reservoir. Potential gas channeling in high permeability streaks and an improved displacement by gas was precisely modeled by the workflow. Injection strategies, such as WAG, SWAG and gas injection have been screened by the model, leading to a conclusion in relatively short period of time.