Video: Dynamic vs. Static Ranking: Comparison and Contrast in Application to Geo-cellular Models
- Mohan Gajanan Kelkar (University of Tulsa) | Saikiran Pochampally (Kelkar and Associates Inc.) | Asnul Bahar (Kelkar and Associates Inc.) | Mohammad Sharifi (Amirkabir University of Technology)
- Document ID
- Society of Petroleum Engineers
- Publication Date
- Document Type
- 2014. Copyright is retained by the author. This document is distributed by SPE with the permission of the author. Contact the author for permission to use material from this document.
- upscaling, ranking, dynamic model, geo-cellular models, realizations
- 0 in the last 30 days
- 5 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 12.00|
With increasing computational power, large geo-cellular models are constructed comprising of multi-million cells. Using uncertainty work flows, large number of geo-cellular models (realizations and scenarios) are constructed which can incorporate uncertainties in static parameters. The typical number of geo-cellular models can range between as few as thirty to as high as hundreds or even thousands. Once these models are constructed, the geo-modeler as well as simulation engineer realizes that all these models cannot be flow simulated without significant upscaling. Therefore, a search begins to find a method which would cost effectively select just a few of these geo-cellular models which can bracket the uncertainties present in these models.
Static methodologies, which rank these realizations based on static properties such as Hydrocarbon Pore Volume (HPV) or STOIIP/GIP may not truly represent the ranking based on actual performance of the reservoirs. For example, a reservoir which contains large HPV but is not well connected will ultimately produce less oil/gas than a reservoir which is well connected but contains less HPV. We, therefore, need a ranking based on dynamic characterization of the reservoir.
We have developed methodology based on Eikonal equation which can rank multiple realizations extremely efficiently. The method determines the time it takes for a traveling of a pressure “wave” to reach a particular grid block from a given well and assumes that time it takes to “tag” a grid block will reflect how quickly that grid block can be drained from a given well. Unlike finite difference or streamline simulation, the method does not require solving any matrix; hence, it is unconditionally stable. In matter of few minutes, a geo-cellular model containing more than fifty million cells can be interrogated.
Using both synthetic and field examples, we demonstrate that the ranking based on our methodology is consistent with the ranking one would obtain using finite difference simulator. Further, ranking based on static properties does not correlate well with the dynamic response of the reservoir. By using the proposed methodology, limited number of geo-cellular realizations can be selected which can capture true dynamic uncertainty in reservoir performance.
The standard operating procedure for conducting modern reservoir modeling consists of steps of building a detail base case containing multi-million cells static model followed by uncertainty analysis. The analysis includes various uncertain variables that may originate from geophysical, geological, petrophysical and engineering disciplines. The combination of these uncertain variables will generate hundreds or even thousands of models through random realizations and/or scenarios.