Slide — 1 /18

1. Thesis Defense 27 May 2011 — College Station, Texas A Numerical Study of Nonideal and Secondary Fractures in Shale-gas Reservoirs using Voronoi Grids Olufemi OLORODE Department of Petroleum Engineering Texas A&M University College Station, TX 77843-3116 (USA) +1.803.397.7623 — olufemi.olorode@pe.tamu.edu Slide — 1/18

2. Objectives: • To present an unstructured mesh-maker that is used in gridding complex and non-ideal fracture geometries • To study the effects of nonplanar and nonorthogonal fractures on reservoir performance • To study the interaction between secondary and primary fractures • To assess the validity of single-fracture representation of multiply-fractured horizontal wells Slide — 2/18

3. Motivation: • Cartesian grids do not provide the flexibility of modeling irregular fracture geometries. • Cartesian grids require far more grid-blocks, many of which are unnecessary. • No consensus on the effect of nonideal fracture geometries on production. • Very little is known about the interaction between induced and hydraulic fractures (Houze et al. 2010). Unnecessary refinement X Y Cartesian Mesh showing 4 planar fractures X Y Voronoi grids showing nonplanar fractures

4. Approach: Develop Meshmaker Relative Sandstone Pore Diameter Construct Voronoi grids Visualize the grids Debug code Any bugs? Yes No 324m Perform simulation 260 x Magnification Base case? Yes Validate with Ecrin ~100μm No Analyze rates using log-log plots Relative Shale Pore Diameter Provide pressure maps where needed Slide — 4/18

5. SRV Gridding: Single-fracture Representation n=6 1 2 3 4 5 Y-axis Fractures xf Horizontal well Z Unstimulated Reservoir Volume Y X-axis X xmf= n*xf 3D View X Y 2D View Slide — 5/18

6. Results: Log-log Rate Profile 1 month 1 year 5 years 30 years Single fracture Representation 10 multi-stage fractures • Discussion: Single-fracture Representation of Multiple Fractures • Fracture interference is absent in single fracture case • Boundary-dominated flow is not seen in the single fracture case

7. Results: Distinguishing between kf and wf • Conductivity is kept constant at 492 md-ft (1.5x10-10 mm-m2). • Do we see distinct trends at early times? Table 1—Fracture parameters in field units Slide — 7/18

8. Results: (after porosity modification) Porosity Modification: where, • Porosity modification keeps mass accumulation constant • Bad news: • we cannot distinguish between kf and wf. • Good news: • we can represent very minute fracture cells with much bigger cells. Slide — 8/18

9. b Nonplanar fracture lt = b+c+d+e+f la = lt sin θ where, lt is total length, la is apparent length, All segments are inclined at angle θ to the horizontal. θ Nonorthogonal fracture lt= a la= a sin θ where, lt is total length, and la is apparent length c h = a sin θ d a e f Background: Nonplanar & Nonorthogonal Fractures Illustration of “Total” and “Apparent” Lengths 3D Schematic of a Nonplanar Fracture 2D Schematic of a Nonplanar Fracture 3D and 2D Schematics of a Nonorthogonal Fracture Slide — 9/18

10. Gridding 2D Aerial View of Nonplanar Fractures Y X 2D Aerial View of Nonorthogonal Fractures Y X Slide — 10/18

11. Results: Nonorthogonal and Nonplanar Fractures • Discussion: • Irregularities in the fracture geometry limits flow-regime analysis with diagnostic rate plots Slide — 11/18

12. Results: Nonorthogonal and Nonplanar fractures Nonorthogonalfrac xf=lt Nonplanarfrac xf=la • Discussion: • The cumulative production initially matches that of a planar fracture with xf=lt, but drops gradually over time. Slide — 12/18

13. Gridding: Secondary Fractures h/4 • Three secondary fracture configurations are studied: • A secondary fracture that intersects the primary fracture at height, h/4. • A centered secondary fracture. • Two secondary fractures at heights h/4 and 3h/4, respectively. h/2 h/4 Y Z X Slide — 13/18

14. Results: Secondary Fracture Flow Profile • Parallel half-slope lines depict linear flow into the SRV. • Increase in rates correspond to the increase in the SRV that is drained into the wells. • Change in slope at late times indicate outset of boundary-dominated flow. • NB:Secondary fractures were modeled with infinite conductivity, and are 0.05 mm (0.00016 ft) wide. Centered secondary frac 2 secondary fracs Secondary frac at h/4 Primary fracs only Slide — 14/18

15. Results: Effect of Secondary Fracture Conductivity Table 2—Secondary fracture conductivity parameters • Discussion: • Dimensionless rate profiles show a reduction in the linear half-slope when the dimensionless conductivity of the secondary fractures becomes less than 10 (finite conductivity) • This may be useful in optimizing fracture design Slide — 15/18

16. Results: Effect of Primary Fracture Conductivity Table 3—Primary fracture conductivity parameters • Discussion: • Dimensionless rate profiles show a drop in production as the primary fracture conductivity drops • Results match those published by Freeman et al. (2010) Slide — 16/18

17. Conclusions: • Irregularities in fracture geometry can limit the analysis of these reservoirs with diagnostic plots. • Production increases as SRV increases for infinite-conductivity secondary fractures. • All infinite-conductivity secondary fractures with the same SRV have identical flow behavior, while finite-conductivity secondary fractures show a reduction in magnitude of the half-slope line.

18. Thesis Defense 27 May 2011 — College Station, Texas A Numerical Study of Nonideal and Secondary Fractures in Shale-gas Reservoirs using Voronoi Grids End of Presentation Olufemi OLORODE Department of Petroleum Engineering Texas A&M University College Station, TX 77843-3116 (USA) +1.803.397.7623 — olufemi.olorode@pe.tamu.edu Slide — 18/18

19. Questions?

20. Table 1.1—Representative Barnett Shale-gas Parameters