Most of the reservoirs in the world contain diffuse or small scale fractures. When these small scale fractures have an impact on the oil or gas flow, they have to be included in the reservoir model.

Exactly like for the matrix modeling issues, you have to understand perfectly the fracture network organization as well as the fracture properties to be able to create a precise fracture network model.

With the Fault network characterization module, called “fault analysis”, you may identify the main fault orientations and dippings. You will understand the impact of the throw and the fractal dimension. Finally you may group the faults by lengths trends.

The module even gives you the possibility to create Fault density grids or maps.

When fracture sets have been identified, you would like to understand how dense they are. You would like to understand if they are denser in one region of the reservoir than another.

By using the Fracture density log computation module, you may compute fracture density logs for all the well, fracture set by fracture set.

As a consequence, the fracture network density is known and completely described everywhere there is a well nearby.

The following workflow step (fracture driver computation), gives you reliable workflow to interpolate this fracture set density between the wells.

Taking into account the fracture density logs previously computed, this module enables you to populate the grid by creating a driver that combines all the 3D properties that have a correlation with the fracture density logs (discriminant analysis).

Then a linear combination of the relevant properties is done to compute the 3D fracture density driver. As this process is done for each fracture set, you may repeat these steps for all the sets with the relevant fracture density logs.

The Fractures and faults objects are described by sets and characterized by numerical properties static (geological) or dynamic (reservoir). This module enables building the static fracture network with any dynamic aspect for the moment.

Thanks to the numerous modeling possibilities you are able to build the most geological numerical fracture networks on the market. Small scale fractures are combined with large scale faults, all constrained by direction, matrix conditions, and density at wells.

When a fracture statistical simulation is not enough to model accurately an over complex field or a human modified fracture network, you may use the Deterministic fracture modeling tool.

Manyaly include the parameters to the object by yourself, and include it within the geological network. It will be taken into account during the secondary porosity, secondary permeability calculation.

When it comes to dynamic parameters, like aperture and conductivity, it is clearly more difficult to appraise numerical values. In general, people enter a long and difficult trials and errors process which is often not concluding, and working well by well : TIME CONSUMING!

The KH calibration is an advanced automatic iterative process, working on a multi well basis, that compares the KH measured at well and the one simulated with the numerical fractured model.

In addition, you may still use flowmeter or well test matching traditional model optimization, as the functionalities are recognized for being leader on the market.

Even if the fracture network is perfectly modeled, a simulator is not able to use it to perform flow simulation.

The only understood format is a grid filled with porosity, permeability and shape factor (or block size, a, b and c).

Use the analytical method to perform a precise upscaling except in low connection cases. For an even more advanced result, use the numerical method, using real discrete numerical models.

In a naturally fractured reservoir study, you generally not only want to characterize and model the fracture network, you also want to simulate fluid flow in the model.

The aim of the simulation is of course to perform better production forecasts and help the decision making process.

The dual medium simulation is particularly efficient with PumaFlow. Not only because PumaFlow 2011 does not have any keyword to write anymore, not only because PumaFlow 2011 is faster, single proc or multiproc, than any other simulator on the market for dual medium, not only because there is no data transfert anymore from the fracture network modeling step to the simulation step, but simply because PumaFlow dual medium physics is the best one on the market!