To main content

Flow diagnostics on stratigraphic and unstructured grids

Modern reservoir simulators provide detailed forecasts of hydrocarbon recovery based on a description of the reservoir, the fluid dynamics, well controls, and couplings to surface facilities. In model-building workflows it is often desirable to determine volumetric connections and the main flow paths at a stage in the process where all the data necessary to perform a full simulation is not yet established. Likewise, even if a simulation model is available, the complexity of a full simulation is not necessary to make certain reservoir management decision or the associated computational cost is too high

Contact person

Conceptual illustration of flow diagnostics. Time-of-flight gives travel time along flow paths, whereas influence regions provide partition of unity for the reservoir volume. Both can be computed by standard finite-volume methods and give residence times of flow paths, drainage and sweep regions, well-pair regions, flux allocations, as well as generalized measures from classical sweep theory expressing dynamic heterogeneity in flow paths. The F-Φ diagram shows how F % of the flow can be attributed to Φ% of the flow volume. The Lorenz coefficient, defined as twice the area between F (Φ) and the straight line F = Φ, correlates well with hydrocarbon recovery


Flow diagnostics are simple and controlled numerical flow experiments that are run to probe a reservoir model, establish connections and basic volume estimates, and quickly provide a qualitative picture of the flow patterns in the reservoir. The methods can also be used to compute quantitative information about the recovery process in settings somewhat simpler than what would be encountered in an actual field. As such, these methods offer a computationally inexpensive alternative to performing full-featured multiphase simulations.

Traditionally, flow diagnostics has been associated with streamline methods. The idea of using finite-volume methods instead stems from a collaboration between SINTEF and Technoguide AS, developer of worldleading seismic-to-simulation software Petrel, in 2000-2002. Seeing the value of the idea, SINTEF continued to perform research on numerical methods for computing time-of-flight and influence regions, which were later picked up by Chevron and put into a reservoir-modeling perspective. In the project, Chevron and SINTEF have teamed up to further develop the flow diagnostics technology.


The primary objective for the project is to develop tools for fast computation of flow diagnostics on complex grids based on finite-volume discretizations.

Secondary objectives:

  1. Develop improved flow-diagnostic solvers on stratigraphic and unstructured grids.
  2. Extend the flow-diagnostic tools to drainage processes and problems with buoyancy.
  3. Demonstrate the utility of the new tools for through case studies and benchmarks.
  4. Develop a prototype tool for use in pressure tests.

Project results

We have demonstrated that flow diagnostics, computed using standard finite-volume discretizations, has a large potential to speed up existing workflows for reservoir modelling and reservoir management. We have also developed examples of new and innovative workflows.

Industry partner Chevron has demonstrated the utility of flow diagnostics for a large number of geomodeling studies, and the use of flow diagnostics has become a standard practice within Chevron’s reservoir characterization workflows. Specifically, flow diagnostics are used to test the sensitivity of the flow response to static modeling parameter choices early in the assessment phase of a project.

SINTEF has primarily focused on workflows related to optimization of oil recovery and netpresent value. Flow-diagnostics do not produce these objectives directly, but for optimization purposes it suffices to compute measures that correlate well with the actual objective. We have shown how simple heterogeneity measures like the Lorenz coefficient provides a very efficient means to suggest injection rates, well placement, and well scheduling that give substantial improvements in recovery factors. Likewise, we have used time-of-flight to define an accurate proxy for optimizing net-present value.

To accelerate the adoption of the new methods, they have been released a new module in MRST, our open-source research tool. The module  computes basic flow diagnostics like time-of-flight (travel time from the nearest injector to a point in the reservoir, or from a point in the reservoir to the nearest producer), tracer distributions and volumetric partitions. From these quantities, one can compute drainage volumes, sweep regions, well-pair connections, well allocation factors, and various heterogeneity measures like flow and storage capacity, sweep efficiency, Lorenz coefficient. The module has an interactive viewer that loads a model and computes all these quantities, as well as examples that demonstrate use of flow diagnostics to assess the accuracy of model upscaling and optimize well placement.

Basic functionality for computing time-of-flight and influence regions are also available in OPM (see GitHub), and various flow diagnostic quantities have been implemented in ResInsight.


Key Factors

Project duration

01/06/2012 - 31/12/2014

Project type

Collaborative and knowledge-building project


The Research Council of Norway, grant no. 215665


Chevron (San Ramon, CA)

Relevant links

Selected literature

Project team