Slide — 1 /18

27 May, 2011 — College Station, TX Study of Nonideal and Secondary Fractures

O.M. Olorode Sli

de

— 1

/18

A Numerical Study of Nonideal and Secondary Fractures in Shale-gas Reservoirs using Voronoi Grids

Thesis Defense 27 May 2011 — College Station, Texas

Olufemi OLORODEDepartment of Petroleum Engineering

Texas A&M UniversityCollege Station, TX 77843-3116 (USA)

+1.803.397.7623 — [email protected]


O.M. Olorode Sli

de

— 2

/18

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


O.M. Olorode Sli

de

— 3

/18

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).

Voronoi grids showing nonplanar fractures

Cartesian Mesh showing 4 planar fractures

Unnecessary refinement

X

Y

X

Y


O.M. Olorode Sli

de

— 4

/18

Approach:

324m260 x Magnification

~100μm

Relative Sandstone Pore Diameter

Relative Shale Pore Diameter

Visualize the grids

Develop Meshmaker

Construct Voronoi grids

Analyze rates using log-log plots

Perform simulation

Provide pressure maps where needed

Any bugs?

Base case?

Debug code

Yes

Yes

No

No

Validate with Ecrin


O.M. Olorode

xmf = n*xf

Sli

de

— 5

/18

Gridding: Single-fracture Representation

3D ViewSRV

Unstimulated Reservoir Volume

X-axis

Y-axis

1 2 3 4 5

xf

n=6

2D View

Horizontal well

Fractures

Y

Z

X

Y

X


O.M. Olorode Sli

de

— 6

/18

Results: Log-log Rate Profile

●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

Single fracture Representation

10 multi-stage fractures

1 month 5 years1 year

30 years


O.M. Olorode Sli

de

— 7

/18

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?

wf, ft kfrac, md modified

0.010 50,000 0.33

0.049 10,000 0.066

0.098 5,000 0.033

0.328 1,500 0.0099

Table 1—Fracture parameters in field units


O.M. Olorode Sli

de

— 8

/18

Results: (after porosity modification)

●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.

fified mod

newref wwf /

where,

Porosity Modification:


O.M. Olorode Sli

de

— 9

/18

Background: Nonplanar & Nonorthogonal Fractures

Nonorthogonal fracturelt = ala = a sin θwhere, lt is total length, andla is apparent length

f

e

d

b

c

θ

a

h = a sin θ

Nonplanar fracturelt = b+c+d+e+fla = lt sin θwhere, lt is total length,la is apparent length,

All segments are inclined at angle θ to the horizontal.

3D Schematic of a Nonplanar Fracture

2D Schematic of a Nonplanar Fracture

3D and 2D Schematics of a Nonorthogonal Fracture

Illustration of “Total” and “Apparent” Lengths


O.M. Olorode Sli

de

— 1

0/18

Gridding2D Aerial View of Nonplanar Fractures

2D Aerial View of Nonorthogonal Fractures

Y

X

Y

X


O.M. Olorode Sli

de

— 1

1/18

Results: Nonorthogonal and Nonplanar Fractures

●Discussion:■ Irregularities in the fracture geometry limits flow-regime analysis with

diagnostic rate plots


O.M. Olorode Sli

de

— 1

2/18

Results: Nonorthogonal and Nonplanar fractures

●Discussion:■ The cumulative production initially matches that of a planar fracture

with xf=lt, but drops gradually over time.

xf=lt

xf=la

Nonplanar frac

Nonorthogonal frac


O.M. Olorode Sli

de

— 1

3/18

Gridding: Secondary Fractures

■ 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.

Y

X

Z

h/4

h/2

h/4


O.M. Olorode Sli

de

— 1

4/18

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.

2 secondary fracsCentered secondary frac

Secondary frac at h/4

Primary fracs only


O.M. Olorode Sli

de

— 1

5/18

Results: Effect of Secondary Fracture Conductivity

●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

kfrac, md Cf, md-ft CfD

3x106 4.92x102 1.67x104

2x106 3.28x102 1.11x104

1x106 1.64x102 5.56x103

2x105 3.28x101 1.11x103

1x105 * 1.64x101 5.56x102

1x104 1.64x100 5.56x101

2x103 3.28x10-1 1.11x101

1x103 1.64x10-1 5.56x100

2x102 3.28x10-2 1.11x100

Table 2—Secondary fracture conductivity parameters


O.M. Olorode Sli

de

— 1

6/18

Results: Effect of Primary Fracture Conductivity

●Discussion:■ Dimensionless rate profiles show a drop in production as the primary

fracture conductivity drops■ Results match those published by Freeman et al. (2010)

kfrac, md Cf, md-ft CfD

5.00x106 4.92x104 1.67x106

5.00x105 4.92x103 1.67x105

5.00x104 4.92x102 1.67x104

5.00x103 4.92x101 1.67x103

5.00x102 4.92x100 1.67x102

1.67x102 1.64x100 5.56x101

3.33x101 3.28x10-1 1.11x101

1.67x101 1.64x10-1 5.56x100

3.33x100 3.28x10-2 1.11x100

Table 3—Primary fracture conductivity parameters


O.M. Olorode Sli

de

— 1

7/18

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.


O.M. Olorode Sli

de

— 1

8/18

A Numerical Study of Nonideal and Secondary Fractures in Shale-gas Reservoirs using Voronoi Grids

End of Presentation

Thesis Defense 27 May 2011 — College Station, Texas

Olufemi OLORODEDepartment of Petroleum Engineering

Texas A&M UniversityCollege Station, TX 77843-3116 (USA)

+1.803.397.7623 — [email protected]


O.M. Olorode

Questions?

Sli

de

— 1

9/18


O.M. Olorode Sli

de

— 2

0/18

Parameters SI Unit Field Unit Frac half-length, xf 90 m 300 ftFrac width, wf 3 mm 0.00984 ftFrac spacing, df 100 m 328 ftWell length, Lw 1200 m 4000 ftNumber of fracs 12 12

Reservoir thickness, h 100 m 330 ftPermeability, kshale 1.0x10-19 m2 1.0 x10-4 ft

Frac permeability, kfrac 5.0x10-11 m2 5.0 x104 ftPorosity, 4 % 4 %Frac porosity, frac 33 % 33 %Temperature, T 93.33 0C 200 0FWell radius, rw 0.1 m 0.32 ft

Reservoir pressure, pi 3.45x107 Pa 5000 psiaWell pressure, pwf 3.45x106 Pa 500 psia

Table 1.1—Representative Barnett Shale-gas Parameters


O.M. Olorode Sli

de

— 2

1/18


O.M. Olorode Sli

de

— 2

2/18


O.M. Olorode Sli

de

— 2

3/18

wfiD ppkh

qq

2.141

2

0002637.0

ftD xc

ktt

Slide — 1 /18

Documents

Transcript of Slide — 1 /18