It must be great being a company like Statoil. Its brilliant researchers come up with new technologies, and Statoil then spins them off to establish companies which can develop innovative and groundbreaking products. Then the parent company profits from being an “early adopter” of these new technologies, usually with great success.

So it is that a company called Numerical Rocks came into being. Beginning as a research and development project at Statoil, it was decided after eight years to spin the technology out via a separate technology company. Now the spin-off is helping the parent on one of its trickier fields, Heidrun in the Norwegian Sea.

The technology
Numerical Rocks’ goal is to provide technical services and software solutions which enable

Figure 1. The Numerical Rocks workflow begins with a micro-CT image (top), which is modeled in 3-D to represent a rock model (middle) and ultimately becomes a pore network image (bottom). (Images courtesy of Numerical Rocks)
fast and reliable prediction of reservoir rock properties without needing traditional laboratory testing. To that end the company has developed e-Core, described as “an electronic core laboratory for modeling petrophysical properties and simulating fluid flow in the pore space of reservoir rocks.” e-Core simulates Nature’s own processes of sedimentary rock formation, including sedimentation, compaction and diagenesis.

Input data are high-resolution images of thin rock sections, and the resulting 3-D rock models are used for the calculation of petrophysical parameters such as permeability and formation factor and simulation of multiphase flow properties to obtain capillary pressure curves and relative permeabilities.

The workflow is shown in Figure 1.

According to Ivar Erdal, chief executive officer at Numerical Rocks, the predictive nature of the software is what gives the technology its edge. “We only need a small piece of the rock, but that tiny piece is very important,” he said. “We don’t make this up off the top of our heads. It’s based on pure physics and a thorough geological understanding of the material.”

That being said, the tiny slice of rock combined with the software produces everything from a three-dimensional rock model to the flow properties mentioned above. “This is what is so intriguing to the oil companies,” Erdal said. “They like what they see and its potential impact on the bottom line.” The “digital core lab” saves months over lab work, which also requires large amounts of core samples. Erdal said that any chunk of rock from the field or formation of interest that gets circulated back up with the drilling mud can, in principle, be used for this type of analysis.

“We can see the potential for modeling and simulating the reservoir behavior at a much earlier stage,” he said.

The concept has been tested on reference sandstone rocks such as Berea, Bentheimer and Fontainbleau as well as reservoir rocks from fields in the Norwegian and North Seas, Algeria
and Germany.

The field
Heidrun was discovered in 1985 and declared commercial in 1986. Statoil took over in 1995 as production operator. It was Statoil’s first foray into the Norwegian Sea.

The field has been developed with a concrete-hulled tension-leg platform (TLP) moored to the seabed by 16 steel tethers. It ranks as the world’s largest TLP and the only one with a concrete hull.

Heidrun produces a mix of oil, gas and water, separated on the TLP. Oil is transported by
Figure 2. The micro-CT image (left) is compared to a random cut through the digital rock model (right).
tanker, and gas travels through the Haltenpipe line to the Tjeldbergodden industrial complex or to customers in Europe through the Åsgard Transport trunk line. Ultimate recovery is estimated at 1.1 billion bbl.

The application
Multiple realizations of the microstructure of a number of different lithofacies in the Fangst and Bat Groups were reconstructed from thin sections obtained from core material. Reservoir rock samples from Heidrun are, in general, poorly consolidated due to shallow burial, which makes them difficult to handle with traditional lab measurements. But that is no obstacle for applying the e-Core technology.

The reservoir rocks are heterogeneous because of complex diagenetic alterations such as the formation of authigenic clay minerals and patchy carbonate cementation. This is a challenge for the digital modeling process.

A mosaic of electron microscope images taken from the thin sections was used to extract the input parameters for the digital reconstruction of the samples. High-resolution micro-CT
Figure 3. This figure shows the transport properties of two core plugs from the same reservoir. The measured relative permeability at various water saturations (filled symbols) are compared to simulated data (open symbols). The krw and kro curves are average values for simulated data.
(computed tomography) images were also acquired for one rock type for comparison with the geological process-based model (PBM). The resolution of the micro-CT data is 2.62 microns for an extracted sample size of 10 mm in diameter. Cross-sections of the micro-CT image and the corresponding PBM model are compared in Figure 2.

Important properties were calculated and compared for these two models:
• Extracting the pore network. The connectivity of the pore network is determined by extracting the “skeleton” of the pore space by an ultimate dilation of the grains followed by detailed measurements along the entire skeleton network to acquire size, volume and shape for every pore body and pore throat;
• Multiphase flow simulations. In all the multi-phase flow simulations, it is assumed that capillary forces dominate at the pore scale. For two-phase flow, the equilibrium fluid distribution is governed by wettability and capillary pressure;
• Transport properties. Relationships such as capillary pressure and relative permeability curves are determined by simulating two-phase displacements such as primary drainage, waterflood and secondary drainage on the pore network representation of the reconstructed rocks, and
• Special Core Analysis (SCAL) data for comparison. Available SCAL data for the rock type includes Amott wettability measurements, centrifuge-measured oil/water relative permeabilities and steady-state relative permeabilities (Figure 3).

The e-Core technology allows Numerical Rocks personnel to carry out sensitivity studies for important reservoir parameters. In this case average relative permeability and capillary pressure curves for waterflooding were calculated using four different Amott wettability indices. The resulting data were compared with SCAL data (Figure 4) for evaluating the effects of wettability on relative permeabilities, capillary pressures and residual oil saturations. This is crucial information for the evaluation and simulation of reservoir performance in Statoil. This study also illustrates the importance of having correct input of wettability in the modeling.

According to Statoil representatives, the flow parameters generated from the pore scale
Figure 4. This figure illustrates the laboratory measurements of capillary pressure curves at three different wettability conditions: Amott Wettability indices 0.14, 0.57 and 0.8, respectively, compared to calculations by Numerical Rocks.
modeling were used for two main purposes: a) to increase the understanding of multiphase flow in the Heidrun field by providing data for a wider range of rock types than was available from SCAL analyses, and b) to use the data in reservoir simulation models to improve the spatial distribution of flow parameters in accordance with variations in the geology of the field.

This study demonstrates the potential of combining computer-generated rock models with numerical calculations to predict rock and flow properties over a wide range of porosities.
The technology offers exciting possibilities for bridging the gap between detailed geologic reservoir models and the lack of associated reservoir parameters — relative permeabilities and capillary pressure curves.

Erdal added that his ultimate goal is not to be a replacement for detailed core analysis but rather to give companies more data than they’re used to having to populate their models and to get those data to them in a more streamlined fashion. “I believe this technology will become a standard tool for the oil companies to describe their reservoirs in the future,” he said.