ISI Logo

Top-Down Modeling: Frequently Asked Questions (FAQ)


Please communicate ANY questions that you might have on our technology and you will receive a prompt reply from ISI.

Q: What is Top-Down Modeling (TDM)?

A: Top-Down Modeling is a new reservoir simulation and modeling technology, invented and introduced to the industry by Intelligent Solutions, Inc.

Q: How is TDM different from existing Reservoir Simulation and Modeling?

A: The existing reservoir simulation and modeling technology is based on the numerical solution of the diffusivity equation. As such, it is based on our current understanding of storage and transport phenomena in the porous media that has been formulated and then solved numerically.

During Top-Down Modeling we make no assumptions about our understanding of the fluid flow in porous media. Instead, we let the data do all the talking. Incorporating the pattern recognition capabilities of Artificial Intelligence and Data Mining, TDM discovers patterns from the data collected from the field and presents them in the form of a reservoir model. The reservoir model is presented in the common format that reservoir engineers are comfortable in interacting with.

Q: Since TDM is developed based on data, what type of data is required?

A: Production/injection history, well logs, core analysis, well tests, seismic, etc. TDM has a completely open architecture when it comes to data. It can and will take advantage of any and all types of data that is available from the field.

Q: But the above data comes in different scale, can TDM use data of different scale?

A: Yes it can. TDM has been developed by reservoir engineers (not mathematicians, statisticians or AI developers) that have a solid understanding of the nature of the data that is available in the oilfield. TDM has been designed to include facilities that understands the scale of data and uses this information accordingly.

Q: How much data is required for TDM to be applicable?

A: Since TDM works with historical and measured data from the field in order to construct the reservoir model, it can only be applied with certain amount of history. The quality of the reservoir model will always be a function of the asset's maturity. The amount of data required for TDM is a function of time and space (just like the diffusivity equation). In other words, it is a function of number of the wells and life of each well. At the lower end of data availability, there must be a balance between number of wells and how long they have been producing as well as the resolution of the production data that is available (higher resolution of production data is more desirable). For example, we have developed successful TDM for an asset in the North Sea with less than 30 production wells and about 6 years of production.

Q: How is Top-Down Modeling (TDM) defined?

A: : Top-Down Modeling is a formal, comprehensive, full field, multi-variant, empirical reservoir simulation model. TDM is developed entirely based on measured data from the field.

  • It is Formal, since a complete and comprehensive workflow has been developed that guides the modeler through the entire modeling process from A to Z.
  • It is Comprehensive, since it has the capability of tracking reservoir pressure and phase saturations throughout the reservoir as a function of time in addition to modeling and history matching oil, gas and water production at each individual well.
  • It is Full Field; since it covers the entire asset and takes into account the interaction and interference between producers and producer-injectors.
  • It is Multi-Variant, since it models production, pressure and saturation as a function of all available data, including the seismic, if available.
  • It is empirical, since it does not start with any pre-conceived notions about our state of understanding of fluid flow in a given field. Instead, it relies on the field measurements in order to develop a version of the reservoir model.

Q: How fast is a TDM run?

A: Once, trained, history matched and validated, TDM has a small computational footprint. Each TDM run takes a few minutes (at most) while reservoir simulation model runs may take hours or days.

Q: What are the outputs of TDM?

A: Output of TDM is very much the same that are expected from a reservoir simulation model. They are pressure or rate profile at the well (depending on the operational constraints) or pressure and saturation distribution as a function of time (each time step) at each grid blocks in the simulation model. Please note that for pressure and saturation distribution, some measurements of static reservoir pressure and time-lapse saturations are required.

Q: Has TDM been tested in the field?

A: YES. TDMs have been successfully applied to fields in the United States, the Middle East, the Gulf of Mexico, and the North Sea.

Q: Has TDM been validated? How can such model be validated to make sure that it does accurately represent the asset that is being developed for?

A: YES. TDM has been validated. To validate such a new technology (and to debug some of its features to make sure how one can enhance its capabilities) the most common technique is as follows:

  1. Develop a reservoir model using a numerical simulator. Include in the model all complexities that you wish to test the TDM on.
  2. Use the numerical model and generate the type of data that is usually available from an asset.
  3. Add some noise to the data (since data from the field is hardly ever as clean as you can get from a numerical model).
  4. Use the TDM workflow and the data from the numerical model and build a Top-Down Model for the asset.
  5. Test the TDM and see if you can predict some of the features of the asset that was incorporated in the numerical model, such as production profile in some future time, Production from a new well that is drilled after the model is completed, optimum injection of water to increase production, etc.

Q: What is the science and technology behind TDM?

A: Artificial Intelligence and Data Mining is the technology behind TDM. Specifically; the main science behind TDM is a proprietary implementation and integration of fuzzy set theory and artificial neural networks that resulted from more than two decades of research and development.

Q: What are some of the limitations of TDM?

A: TDM cannot be developed for green fields (new fields). If the main mechanism of the production changes after the development of TDM, it cannot predict those behaviors, for example, a TDM that is developed before a water flooding process has been initiated, cannot predict what would happen if water flooding is implemented in the field.

Q: What are other applications and uses of TDM?

A: Top-Down Model has as many uses as any reservoir model can have. It can be a used as a tool for Master Development Plan, any type of production optimization such as identification of optimum infill drilling locations, optimizing water flooding, gas injection, etc.