Streamline-Based History Matching

The advantage of streamlines is that they relate reservoir regions to individual well production mismatches throughout each well's history. Starting with this simple idea, Streamsim Technologies invested +10 years of R&D in a joint industry project involving Streamsim, Stanford, and multiple production companies. We have a novel workflow which addresses the difficult problem of individual well-level history matching to modify static gridblock properties between wells.  Thus we also assume that the mismatch at the well-level can be corrected by modifying geological properties between wells. Our workflow replaces tedious and time-consuming manual history matching. No more need for using BOX keyword multipliers, which has become standard practice for history matching, but is universally recognized as having no relation to reservoir geology.

 

More Info: Pricing

Support: Knowledge BaseTutorialsTraining


The History Matching Workflow

Field-Level History Match. The starting point is to have a flow simulation run over the complete history and that has a good field-level history match.  In general a model with a good field-level match will have many wells with poor well-level matches.

Well Selection. The next step is to quantify the individual well history matches in a single history match error plot then apply a selection criteria such as 'worst 10 wells' or 'most important wells' or a combination of criteria.
Time Interval Selection.  Because grid property corrections due to early timesteps can also impact the history matching results at later timesteps, the overall strategy is to match the early timesteps first and progress to matching later timesteps.  By default, time intervals of equal history match error are identified.
Static Properties Selection. Any geological property keyword that is defined on the grid, such as porosity, directional permeabilities, NTG's, TMult's, etc, can be modified within a time interval. Once a property is modified, the flow simulation must be repeated from TIME=0 and the history matching errors recomputed.
Example Correction Factors. For a time interval, studioSL will visualize the values of the multipliers. The locations of the multipliers are based on the streamlines supporting the wells at the particular time interval of mismatch.  The values of the multipliers are based on the amount of mismatch at the selected wells for the particular time interval.
Automatic Running. With the wells, static properties, and time intervals selected, studioSL will start the history matching workflow.  After each run is completed, new corrections will be computed, applied to a new run, and restarted from time zero.  A summary plot of the history matching errors (objective function) or well-by-well mismatch on rates can be created to track the progress of the run.
Well-level History Match Comparison. Once all time intervals have been processed, it is easy to compare the history match improvement for all wells or individual wells.
Static Grid Properties Comparison. Use studioSL 3D grid property comparison feature to quantify where changes were made and by how much, compared to the original field-level history matched model.

Additional Features

  • High resolution grids with 100's of wells and 100's of timesteps.
  • Works with Eclipse .DATA models.
  • Compare well vs field level matches among multiple runs.
 
 

Contact in USA

Corporate Headquarters 
StreamSim Technologies, Inc.
865 25th Avenue
San Francisco, CA 94121
U.S.A.

Tel: (415) 386-0165

Contact in Canada

Canada Office 
StreamSim Technologies, Inc.
Suite 102A - 625 14th Street N.W.
Calgary, Alberta T2N 2A1
Canada

Tel: (403) 270-3945

Important Links