Press Esc to close

WorldOil


JUNE 2014

Vol. 235 No.6

CORE ANALYSIS

Digital core flow simulations accelerate evaluation of multiple recovery scenarios

Digital core flow simulations can now replicate familiar lab results. Consequently, operators will be able to dramatically shorten traditional cycle times, better understand increasingly complex reservoirs before making costly field decisions, and maximize both short-term production and long-term recovery.

WO0614_Anderson_SLB_Core_Analysis_Fig_00.jpg
Using digital rock and digital fluid models, CoreFlow simulation delivers flow and oil recovery information to help optimize reservoir recovery (image courtesy of Schlumberger).

Analyses of rock samples in laboratories worldwide provide a wealth of information about key rock and fluid properties, including lithology, porosity, permeability, fluid saturation and multiphase flow, enabling operators to better characterize pore systems in oil and gas reservoirs. Physical core measurements allow geoscientists and reservoir engineers to better understand depositional environments, evaluate exploratory wells, assess field redevelopment potential, and investigate early production strategies, as well as improved recovery scenarios.

To boost recovery in conventional sandstone and carbonate reservoirs, it is essential to identify which fluid types, injection rates, volumes, and other engineering parameters are likely to have the greatest impact on performance. In unconventional shales and mudstones with complex vertical and lateral heterogeneities, it is vital to determine which subtle rock textures exhibit the greatest producibility. Traditional core analysis can help, of course. However, as reservoir engineering becomes more complex to deal with current challenges, core analysts must understand and realistically model pore geometries and fluid behaviors at pore scales. In addition, obtaining physical core measurements under different boundary conditions can take months, even years, to complete.

Emerging digital core technologies and pore-scale flow simulations are beginning to complement—and enhance—traditional laboratory techniques. Integrated physical and digital rock and fluid workflows promise to accelerate the evaluation of multiple scenarios, improve decision-making, and maximize production and ultimate recovery from highly complex reservoirs.

DIGITAL ROCK MODELING

Traditional core analysis can be time-consuming. Consider the steps involved in a typical relative permeability test. First, rock samples are cut from the whole core and cleaned, which can take from days to weeks and, in some cases, more than a month. Often, fluids must be recombined, which can also take several weeks. Saturating rock samples with fluids, and allowing time for core and fluids to reach equilibrium requires approximately 40 days. Most relative permeability tests analyze roughly six equilibrium points, each of which takes three to five days, or an additional 18 to 30 days to obtain the initial set of curves. All told, a single test can take three months or more. To run a second test on the same core, the sample must be cleaned again, which may require less time, but still, roughly a week. The core must be resaturated with fluid and equilibrated, which takes another 40 days. Finally, obtaining a second set of relative permeability curves requires several more weeks to a month.

If an operator wanted to investigate multiple production strategies and improved recovery scenarios, it could literally take years to run them all, using traditional laboratory techniques. Since few companies have the time or budget, most settle for a base-case scenario—an engineering evaluation to estimate the best-case scenario, and a second lab test. Frequently, this is the end of core analysis, at which point, an executive decision may be made—based on just two tests—either to implement the recovery scenario, or to suspend the project. In some cases, risky trial-and-error experiments are conducted in the field, injecting various fluids under differing boundary conditions to see what happens. The danger, of course, is that permanent reservoir damage can occur, which has serious, long-term economic consequences.

Ideally, most oil and gas companies would prefer to test many more scenarios in far less time. For several decades, therefore, researchers have been investigating ways of using digital images of core samples to extract useful reservoir information. In the 1980s, for example, universities began digitizing thin sections from cores, trying to calculate capillary pressure and relative permeability, without impressive results. Later, research laboratories began using medical imaging technology to scan cores and create various digital rock models for core analysis. Only in the past five years, however, have digital rock technologies been commercialized successfully. And only within the last year or so, have core analysts, petrophysicists and operators begun to accept the potential viability of digital approaches.

Why has there been such skepticism until recently? First, vendors initially overstated the capabilities of digital rock modeling. Claims were made that digital core analysis would replace traditional lab techniques, which has never been feasible. Second, digital rock models lacked sufficient validation from physical measurements to convince specialists that they could realistically simulate multi-phase fluid flow through complex pore systems. In fact, most providers focused primarily on modeling the rocks, not the fluids, hence the term “digital rock.” In addition, no one could actually perform both digital and physical core analyses on the same rock sample in the same facility. Third parties, different core plugs, and inevitable delays in obtaining final lab results were not uncommon. Discrepancies—often significant ones—between the digital and physical measurements were inevitable, especially in heterogeneous reservoirs.

Finally, digital modeling suffered either from oversimplified pore geometries, oversimplified fluid properties, or both. For example, one technique retained the overall complexity of the rock’s pore spaces, while reducing their actual geometries to a network of artificial tubes and spheres. A newer method successfully retained the true geometry of the pores, but fluid dynamics remained overly simplistic. Neither digital rock approach could realistically simulate the highly complex chemistry and physics of fine, pore-scale rock-fluid and fluid-fluid interactions.

All these shortcomings raised genuine concerns in the minds of astute observers. Uncertainties at the beginning of a project could lead to unreliable results farther down the road. Therefore, to take digital core analysis to the next level, more integrated processes and more realistic flow simulations were necessary.

INTEGRATING PHYSICAL AND DIGITAL ANALYSIS

In December 2013, Schlumberger introduced CoreFlow digital rock and fluid analytics services, Fig. 1. This new, integrated approach is unique in two ways. For one thing, physical and digital core techniques are fully integrated in the same laboratory, using the same rock sample for both types of analysis. Physical measurements refine the digital model, while digital flow scenarios guide subsequent lab tests in an iterative manner, rather than in a sequential series of steps.

WO0614_Anderson_SLB_Core_Analysis_Fig_01.jpg
Fig. 1. Digital rock and fluid analytics services use results of laboratory evaluations, such as fluid properties tests, as input to digital core analysis. Running multiple digital scenarios provides guidance for additional physical laboratory measurements, such as flow studies (image courtesy of Schlumberger).

More importantly, complex multiphase pore-scale flow simulations are carried out using a new breakthrough technology known as direct hydrodynamic (DHD) simulation, Fig. 2. Unlike conventional digital rock approaches, DHD simulations correctly represent real pore geometries, real fluid properties, and real rock-fluid/fluid-fluid behaviors—without oversimplifying any of these vital components. Thus, digital simulations enable geoscientists and engineers to evaluate numerous production or improved recovery scenarios in a fraction of the time required by traditional lab methods, alone.

WO0614_Anderson_SLB_Core_Analysis_Fig_02.jpg
Fig. 2. Flow through the pore spaces of rock samples can be visualized after running a scenario in the DHD simulator. In the image, the grains are rendered invisible so the fluid saturation locations can more easily be seen (image courtesy of Schlumberger).

Workflow integration. In either a conventional or unconventional reservoir, the integrated workflow typically begins by imaging the whole 4-in.-diameter core, using a helical dual-energy computed tomography (CT) scanner, leaving the rock and fluids sealed inside the original sleeve. Cross-sectional images of rock, spaced as close as 0.5 mm apart, are used to construct a slice-by-slice 3D “movie” of the whole core. Then, heterogeneous rock types and flow units can be color-coded to help asset geologists identify and prioritize reservoir zones for more detailed physical and digital core analysis.

Next, sample plugs of sufficient size are taken from selected locations to allow core analysts both to physically measure rock properties and to digitally image the same plug. By cutting a 4×8-mm sample plug from the whole core, it is no longer necessary to test two, slightly different samples. This enables a one-to-one comparison of physical and digital results, minimizing or eliminating the uncertainties that arise from subtle differences between even closely spaced plugs. Once a sample’s physical properties have been measured, if it is a conventional sandstone or carbonate, micron-scale or microCT imaging comes next. In an unconventional shale or mudstone, nano-scale imaging with a scanning electron microscope (SEM) is more appropriate, Figs. 3 and 4.

WO0614_Anderson_SLB_Core_Analysis_Fig_03.jpg
Fig. 3. An adsorption isotherm test determines the volume of gas that can adsorb onto the pores of a shale or coal sample. These pore spaces in unconventional materials can be evaluated using SEM digital imaging and analysis techniques (image courtesy of Schlumberger).

WO0614_Anderson_SLB_Core_Analysis_Fig_04.jpg
Fig. 4. Scanning electron microscopy image shows quartz overgrowths that have formed on quartz grains in a core sample (image courtesy of Schlumberger).

MicroCT imaging effectively resolves clastic and carbonate grains and pores in samples with micron-scale resolution. Technicians shine X-rays through each sample to obtain a backscatter image, rotate the sample a 10th of a degree, take another image, and continue rotating until 3,600 backscatter images provide a full 360° view. Running a mathematical inversion on the microCT data enables them to construct a high-resolution, 3D digital image of the rock. Additional processing creates a digital rock model through which various core flow simulations can be run.

SEM analysis of shales and mudstones provides an order-of-magnitude higher magnification, imaging grains, pores and capillaries at nanometer scales. First, the sample surface is flattened, using a focused argon-ion beam, and a 2D image is captured. Then, by repeatedly milling away the surface, and taking an additional 200 to 500 images of the slices, a 3D volume, about the diameter of a human hair (less than 100 microns), is created. Flow in unconventional reservoirs is controlled by pores associated with minerals, which are extremely tiny compared with conventional rocks, as well as pores associated with organic material, which can be as small as 20 nanometers in diameter.

After imaging by microCT or SEM, the static 3D digital rock model is entered into the DHD simulator, along with a dynamic digital fluid model, derived directly from physical fluid analysis.

DHD simulation. The theory and physical principles underlying direct hydrodynamic simulation date back to the 19th century, originally in the field of thermodynamics. In the early-to-mid-20th century, the theory was utilized more broadly in physics and chemistry. In the 1990s, it was finally applied to compositional hydrodynamics by a Russian scientist, who subsequently joined the Schlumberger Research Center in Moscow. Formal development of the DHD simulator, itself, began about five years ago. It was commercially introduced in late 2013. This unique technology accurately simulates real multiphase, multi-component fluid flow through digital rock models that honor real pore geometries, even at very fine scales. Its technical capabilities have been documented extensively in more than 20 academic papers and one book.

A combined digital and physical core analysis study can begin by establishing an initial base-case from rigorous laboratory analysis, and confirming that the DHD core flow simulation is a good match with the physical measurements. At this point, the true power of the digital approach becomes apparent. In the time that it would normally take to run one more physical core test—to evaluate a presumed best-case scenario—analysts can perform several—perhaps four to eight—complicated digital rock and multiphase fluid simulations.

Scenario evaluation. To plan improved recovery operations in conventional reservoirs, operators could rapidly compare, for example, the relative sweep efficiencies of gas or waterflooding under different initial injection rates, wettability conditions, and pressure drop scenarios, at varying distances from the wellbore. Digitally investigating multiple flow scenarios in less time empowers the asset teams to make more confident technical and financial decisions, to focus on the geological and engineering parameters that represent the greatest risks, and to optimize those risks with the greatest potential to increase ultimate recovery. Even in the primary production stage, DHD simulation enables operators to assess the economic impact of moving from single-phase to two-phase fluid flow.

The value of digital DHD rock and fluid simulation in unconventional reservoirs lies in its ability to quantitatively distinguish the relative producibility of different rock textures. Traditional core analysis often cannot differentiate subtle heterogeneities in shales, due to extremely small sample sizes. Thus, physical measurements tend to average flow characteristics across several distinct textures. Arming engineers with the relative producibility of specific argillaceous, calcareous and siliceous zones would allow them to design better hydraulic fracture completions. Once petrophysicists know which textures matter, these textures can be located elsewhere in the reservoir through optical examination of drill cuttings.

DIGITAL ROCK AND FLUID MODEL VALIDATION

Prior to commercialization, DHD core flow simulations through pore-scale digital rock models were rigorously benchmarked against both physical models and real rocks. Comparisons have been made with experimental flow studies using 2D micromodels, standard reference curves from core analysis in the classic literature, and pore flooding of actual rocks, imaged in real time with advanced X-ray microtomography.

The actual behavior of multi-phase fluids through various pore and capillary geometries is well known from experimental studies flooding etched glass micromodels in the lab. Digital DHD simulations have properly replicated pore filling and displacement behaviors of wetting and non-wetting phases in simple models, including crossing flow paths, narrow capillaries connecting two pores, pore doublets or capillaries that split into channels of two different sizes before reconnecting, and networks of pores and flow paths.

Analysis of improved recovery scenarios aims at achieving the lowest possible residual oil saturation, which depends, in part, on different capillary numbers—the ratio of viscous forces to capillary forces. To validate DHD simulations against data from the literature, waterflooding of an oil-saturated digital model of a real sandstone was modeled at different capillary numbers until residual oil saturation was negligible. The resulting capillary desaturation curve corresponded well with published curves, measured physically in very similar sandstone samples.

Finally, Shell researchers used fast, synchrotron-based X-ray microCT imaging to validate DHD modeling of digital two-phase fluid dynamics. First, the distribution of oil in a physical rock sample was imaged before, during and after a series of pore filling events. Next, the simulator reproduced selected events using the same initial conditions as the experiment—fluid properties, pore geometries and flow regimes. The resulting experimental and digital geometries of oil distribution, as well as displaced fluid volumes, were in quite good agreement. In fact, the digital simulation was able to visualize even more complexities, at more time steps during pore filling events, than fast microCT imaging.

FUTURE IMPLICATIONS

By combining rigorous physical testing in the laboratory with advanced 3D imaging and direct hydrodynamic simulation technologies, core analysts are finally demonstrating the validity of digital rock and fluid analytics. There is no reason, however, to believe that digital core analysis can, or ever will, replace physical analysis. Laboratory testing will always be essential to anchor digital models in the ground truth, found only in real rocks and fluids.

Nevertheless, as more petrophysicists, geologists and reservoir engineers see that digital core flow simulations can actually replicate familiar lab results, they may decide to forego further, time-consuming physical tests, when they need to evaluate additional recovery strategies. As a result, operators will begin to dramatically shorten traditional cycle times, better understand increasingly complex reservoirs before making costly field decisions, and maximize both short-term production and long-term recovery from oil and gas assets worldwide. wo-box_blue.gif

The author


MARK A. ANDERSEN is Schlumberger’s domain head for core physics and manager of the digital core analysis group at Schlumberger Reservoir Laboratory in Houston. Dr. Andersen is the author of many technical papers focused on wettability and production rock mechanics, and is the author of a book on reservoir properties of North Sea chalks. He earned a BS degree in engineering physics from the University of Oklahoma at Norman, Okla., and MS and PhD degrees in physics from The Johns Hopkins University in Baltimore, Md.

Bookmark and Share


WO DATA HUB

EngineeringTablesIcon-Large

Engineering Data Tables

World Oil's specialized upstream Engineering Data Tables featuring the Drill Bit Classifier, Tubing Tables and more. Get Total Access today.

WO SUPPLEMENTS

2013 Fracturing Technology

2013 Fracturing Technology

ENERGY EVENTS

Upcoming Events