Field Development Using Compostional Reservoir Simulation and Uncertainty Analysis in the Delaware Basin

Hefner, William (University of Oklahoma) | Davudov, Davud (University of Oklahoma)

OnePetro 

Abstract

This study presents a novel, integrated workflow to maximize recovery using PVT compositional modeling, history matching, and numerical reservoir simulation in a tight oil sand formation, the Second Bone Spring. Advancements in unconventional resource development have enabled the Delaware Basin to become highly significant. However, optimizing the development of each formation is still lacking in understanding. This study is one of multiple future studies over tight reservoirs in the Delaware Basin and exhibits a comprehensive approach. Properties that will be optimized are well spacing, reservoir parameters, and EOR feasibility.

To determine the behavior and optimize the development of each of these reservoirs, data from multiple sources was necessary. The data compiled consisted of reports from PVT analysis, completions design, petrophysical analysis, daily production and pressure, deviation surveys, structure and isopach maps, and well design. This data was then implemented into a 3D numerical reservoir simulator (CMG-GEM), first to confirm PVT output in a compositional simulation (CMG-WINPROP), then to simulate up to 20 years of production, and finally to use uncertainty analysis (CMG-CMOST) to optimize reservoir input parameters. Once a base case scenario was established, we then furthered our investigation of well spacing and EOR feasibility by setting up multiple different scenarios for each and running them for 20 years. EOR scenarios included 1-3 month huff-and-puff CO2, as well as low salinity water injection. Results are normalized per foot of completely lateral length and lab data is implemented in EOR simulations.

Our results confirm that reservoir parameters, once established after uncertainty analysis, have a large impact on both optimizing well spacing and EOR feasibility in the Second Bone Spring formation. With each well having very similar cluster spacing, proppant amount and type, and fracturing fluid and type, up to 250 feet of inter-well spacing is unaccounted for. Optimized models show that closer spacing of at least 150 feet can increase EUR estimates an average of 11.25%. An increase of 5-17% recovery is observed once a smaller spacing is implemented. EOR models showed that CO2 and low salinity water injection are viable candidates for the formation (7.25-9% increase for CO2, 6.25% for LSWI).

This integrated study refines our reservoir parameter estimates and helps identify potential to maximize recovery. It suggests that a tighter spacing is necessary to cover a larger portion of the reservoir, as well as showing that EOR is feasible. An improved understanding of the entire reservoir leads to better production and economic estimates.

  Country: North America > United States (1.00)
  Geologic Time: Phanerozoic > Paleozoic (0.47)
  Industry: Energy > Oil & Gas > Upstream (1.00)