Marco R. Thiele and Rod P. Batycky

1
SPE 84040
SPE ATCE Denver, CO October 2003
Water Injection Optimization Using a
Streamline-Based Workflow
Marco R. Thiele and Rod P. Batycky StreamSim
Technologies Special thanks you to Nobuo
Nishikiori, Arabian Oil Company, for field data
set.
2
Outline

1. Objectives and introduction to streamline
   simulation.
2. The seed additional data supplied by
   streamlines.
3. Waterflood optimization.
4. A workflow.
5. Example application.

3
Introduction Biz Objectives
Develop/obtain knowledge
Synthesis and analysis
Performance Tools
(simulation)
Obtain and use key data
People
Performance Tools

• Analogues
• Decline Curve Analysis
• Material Balance/Type Curve Analysis
• Streamline Simulation
• Finite Difference Simulation
• Integrated Production Modeling

Increasing complexity,time, skill, knowledge
(after Pande 2003)
4
IntroductionStreamline Sim.

• Solve transport problem along 1D streamlines
  rather than on an underlying 3D computational
  grid.
• Efficient solution of transport problem?finer
  grids, speed but also visualization new data.

Streamlines
Compositions
Geology
Pressure
(From Y. Gautier et al., 1999)
5
Modern Streamline Simulation

1. SLs are traced in 3D.
2. Conservation equations along SLs are in terms
   of TOF.
3. SLs are updated in time to reflect changing
   reservoirconditions.
4. Numerical 1D solutions along SLs.
5. Gravity
6. Compressibility

6
The seedNew Data from SLs

• Streamlines are able to quantify
  injector/producer relationships

7
The seedFlux Pattern Map
8
The seed Injection Efficiency

• Injection Efficiency

• Ieff for injector or for each producer-injector
  pair changes with time.
• Works as a ratio of rates (instantaneous) or
  cumulative volumes (average)
• Can be used for any injection volume type.

9
The seed Injection Efficiency

• Injection efficiencies can be used to
  automatically identify well pairs with extreme
  water cycling.

Oil Produced
Water Injected
10
Waterflood Optimization

• Objective?best use of injected volumes
• for displacement purposes, not
  re-pressurization.
• Increase rates in high efficiency connections,
  decrease rates in low efficiency connections.
• Must increase/decrease rates at both producers
• and injectors.

11
Waterflood Optimization

• Once well pairs are identified, use a weighting
  function to decide how much volume to shift from
  low to high efficiency injectors.

12
Workflow
t,q
Calculate efficiencies using last know rates.
t,qn1
Re-allocate rates using weights from equation.
tDt,qn1
Simulate forward by Dt using new rates.
13
Waterflood Optimization
14
Middle Eastern Limestone Res.

• Compressible SL simulation
• 8 producers / 5 injectors
• Acceptable HM
• Objective plan future water injection.

15
Middle Eastern Limestone Res.
SLs by Producers
SLs by Injectors
16
Middle Eastern Limestone Res.
Flux Pattern Map by Injectors(beginning of
optimization)
17
Middle Eastern Limestone Res.
Flux Pattern Map by Injectors(end of
optimization 5 years)
18
Middle Eastern Limestone Res.
T
10 optimization steps (6 months /step)
T 5 Years
19
Middle Eastern Limestone Res.

• Streamline simulation a complementary tool to
  finite difference simulation ?
• Take optimized reservoir rates from SLs and feed
  to finite-difference simulator

Performance Tools

• Analogues
• Decline Curve Analysis
• Material Balance/Type Curve Analysis
• Streamline Simulation
• Finite Difference Simulation
• Integrated Production Modeling

20
Middle Eastern Limestone Res.
Optimized Period
Not optimized
Oil
Optimized
Water
21
Conclusions

• Streamlines offer a powerful new way to optimize
  flooding operations
• Well allocation data
• Flux Pattern Map
• Streamlines complementary to more traditional
  simulation approaches.
• Procedure is automatic?ability to optimize large
  fields/many wells.

22
Thank you.