[Image Courtesy: KAUST.]
The oil industry must continually and dependably meet the hydrocarbon demand as most of the easy oil is gone and future production will come from more challenging reservoirs that require complex technologies. Hydrocarbon producers must be cost-conscious, acknowledging that oil prices might not reach the high levels of the past, and as an industry ensure that oil prices do not drop significantly for long-term projects to remain sustainable. Also, in the age of ever increasing environmental scrutiny, the industry must be clever to minimize its environmental impact despite the associated added costs. To sum up these facts: In the next half a century or so, how do we maximize hydrocarbon recovery from challenging carbonate reservoirs while limiting our environmental footprint and keeping profit margins sufficiently high?
In early February, porous media experts from all over the world assembled at the King Abdullah University of Science and Technology (KAUST) in Saudi Arabia for a research conference to find answers. They shared their latest research and outlook trying to shed light on the issue through five different disciplines: geology, multiscale imaging, multiphase flow, geomechanics, and numerical modeling.
Geology
Led by KAUST Professor Volker Vahrenkamp, the Geology discussion focused on description and characterization of multiscale heterogeneities in naturally fractured carbonate outcrops and reservoirs and how to make meaningful translations of geological findings into models that can be used for fluid-flow by reservoir engineers.
Professor Giovanni Bertotti from TU Delft described how conventional workflows commonly rely on frequency distributions, stochastic tools, and other weakly constrained geological drivers (i.e., faulting and folding). He then proposed an alternative approach that aims at representing fracture networks by a limiting number of end members that can be identified in the subsurface from borehole data. Based on this knowledge, he argues that we can build more geologically realistic fracture models.
Pascal Richard from Shell spelled out an integrated workflow used for the characterization and modeling of naturally fractured reservoirs comprising several components such as core analysis, borehole imaging, and well testing. This workflow allows them to construct small-scale discrete fracture network (DFN) models as well as full-field DFN to support well planning, production history matching, and forecasting.
Multiscale Imaging
The session on multiscale imaging, led by KAUST Professor and Center Director Tad Patzek, focused on the interplay between increasing technical capabilities that allow us to obtain higher resolution and details for every scientific phenomenon of interest and the practical need for such details at the industrial scale. In other words, what do we really need to know to successfully increase oil recovery at the field scale?
Professor Michael Marder from University of Texas at Austin provided insight into a diffusion imaging approach to understand the flow of shale gas through the complex networks generated during hydraulic fracturing. Some useful quantities that can be extracted from this work are the volume accessed by the fractures and the characteristic timescale of depletion.
Patzek took part in the discussion and shared some of his recent work on a physics-inspired data mining technique that was applied to the prediction of oil production from the Bakken. The Bakken is a particularly challenging play because most wells are operated using artificial lift, and the fact that some wells are operated above bubble point pressure while others are operated below bubble point pressure, resulting in significant changes to gas/oil ratio. Patzek’s approach handles this difficulty easily by adjusting the total system’s compressibility and oil mobility.
Multiphase Fluid Flow
The session on multiphase fluid flow in heterogeneous systems, led by KAUST Professor Carlos Santamarina, explored the topic with an emphasis in pore-scale approaches. Steffen Berg from Shell presented a pore-scale simulation study trying to connect the displacement events at this scale to the macroscopic observations at the Darcy scale for consistent upscaling. He highlighted the Euler characteristic as a missing state variable that controls relative permeability and can be used to eliminate the concept of capillary hysteresis.
Professor Christopher MacMinn from Oxford University offered insight into the process of fracture creation by systematically studying the complex gas-migration mechanisms through sediments. Depending on the competition between buoyancy, capillarity, and stiffness the migration of gas can take place through fluidization, pathway opening (mimicking the formation of fractures), or pore invasion.
Geomechanics
The session on geomechanics, led by KAUST senior research scientist Thomas Finkbeiner, focused on the multiphysical interactions at the reservoir scale. Professor Nicholas Espinoza from University of Texas at Austin gave a presentation featuring the impact of gas desorption induced stresses, which were shown to have a noticeable effect on the evolution of stresses at the reservoir scale and hence modify the permeability of natural and induced fractures.
Leonardo Cruz from Baker Hughes presented results from an investigation into the interactions between hydraulic fractures and pre-existing joints, which can be used to optimize design parameters related to the fracturing fluid.
Numerical Methods
In the final session on numerical methods, KAUST Professor Hussein Hoteit focused on emerging trends in simulation techniques that may be required for the accurate forecast of displacement mechanisms that rely on complex physics in fractured media.
Professor Abbas Firoozabadi, from the Reservoir Engineering Research Institute, presented experimental phase behavior observations for the flow in shale media. Particularly, he emphasized the need for molecular modeling for the proper interpretation of experimental results for these systems including pressure-induced adsorption hysteresis. Hoteit also took part in the discussion by providing a high-level view of different modeling approaches and the need for each depending on the application. For the modeling of fractures, the standard dual-porosity models are commonly used in field application due to its attractive low computational demand. However, their overly simplified geometric assumptions and lack of clarity in the definition of shape functions make them unreliable for cases requiring complex physics (e.g., enhanced oil recovery). Alternatively, discrete-fracture type models are regaining interest due to widespread applications in naturally fractured carbonate reservoirs and unconventional oil and gas reservoirs. These models, though physically more representative, usually come with a hefty computational price tag.
In closing, the conference brought together some of the world’s prominent fluid flow in porous media researchers and pointed a way forward toward possible solutions for some of our industry’s most challenging dilemmas. This gathering made a significant dent at the issues in question, mainly by inspiring a large group of petroleum engineering students, which was the core of the audience. It highlighted the need for interdisciplinary thought and collaboration as a prerequisite for finding the way ahead.
