The Importance of Computed Tomography in the Creation of Carbonate Reservoir Databases Based on the Study of Natural Analogues

R. Swennen, S. Claes, and A. Foubert
Geology, Earth & Environmental Sciences, KU Leuven, Heverlee, Belgium

It is common knowledge that pore networks in carbonate reservoirs are more complex than in sandstones. This not only relates to the very wide variety of rock constituents but also is caused by diagenetic overprinting, especially the development of secondary porosity (e.g. karstification, solution enlargement of fractures, dolomitisation, …) with porosity development ranging from micro- to several meter scale. However, the lack of quantitative geometrical data from key carbonate geobodies, including sedimentological, diagenetical and petrophysical data, provides a serious uncertainty in reservoir volumetric calculations and prediction of production behaviour. Data derived from reservoir analogues, where specific reservoir rock types are studied at subseismic scale, can reduce this uncertainty, but requires intensive field work in well exposed areas. However, the most important drawback at this moment is the lack of any coordinated action. Many companies carry out reservoir analogue studies “à la carte” if suddenly a certain industrial need is emerging. A real long-term strategy is, however, missing. Also from academia, no systematic approach exists, on the one hand because most research groups are rather opportunistic (and thus are happy if they can carry out a reservoir analogue study sponsored by industry), on the other hand most research groups are too small, and finally most groups lack the full range of expertise to carry out a full reservoir study.

Nevertheless, a large number of good reservoir studies have been carried out over the last couple of years, but there are a number of potential issues that need to be raised, before one can come to a solid database. The first issue relates to the definition of what are the criteria to define a good analogue. We are aware of the existence of major reservoir systems, be it related to specific reservoir levels like the Khuff, the Arab, the Smackover, the Thamama, … or with respect to specific sedimentary or diagenetic settings such as the Lower Carboniferous reefal reservoir systems, hydrothermal dolomite reservoirs, karst reservoirs, sub-salt continental carbonate systems, … These analogues should possess similar sedimentological, diagenetical, structural and petrophysical characteristics as their reservoir counterparts, and they should expose these features at sub-seismic scale (several kilometres by several kilometres and over a sufficient vertical extend). However it is clear that due to their present-day exposure of these analogues, some near-surface related diagenetic alteration possibly hampers their straightforward use. All reservoir analogues should be scanned by high resolution LIDAR (eventually also by hyperspectral analysis) and georeferenced sampling should be carried out. In conclusion, a number of these key-reservoir analogues should be studied in a systematic way by a consortium of research teams, following the same research protocol for each case added to the database.

Another key problem concerning many studies, is the tradition of defining sedimentary and diagenetic characteristics in a descriptive way, despite that we all know that we should follow a quantitative and reproducible approach. Furthermore geologists are used to study rocks by the use of thin-sections, which are in fact 2D representations of features that are relevant in 3D (e.g. mineralogical distributions, porosity, permeability, etc. …). For example, presently porosity classifications only consider information in 2D. A real breakthrough in this respect is the systematic use of Computed Tomography (CT) in reservoir analogue studies, in conjunction with techniques like MICP, NMR, etc. But in contrast to many of the other techniques, CT allows to acquire quantitative data on porosity as well as mineralogy (applying for example dual energy scanning; Remeysen and Swennen, 2009). CT allows to work out a new porosity classification in 3D where several different aspects of shape can be identified, where form, sphericity and roundness can be considered as key parameters. There are however a number of intrinsic problems related to CT, such as resolution and artefact reduction. Resolution can be as low as 0.5µm in three dimensions, however, as resolution improves the size of the sample becomes very small (2mm in diameter) questioning the representativeness of what is analysed especially in macro-porous carbonate reservoir rocks. This leads to research addressing the criterion of “Representative Elementary Volume”, i.e. the smallest value that can be taken as a representation for the entire sample area/volume with respect to a certain parameter that does not respond to small changes in volume or location (Bear, 1972). Note that it is very likely that the REV will be pore-type dependent. Artifacts cause some errors in the CT data (e.g. beam hardening, thresholding (where double thresholding is recommended)…), but noteworthy is that these errors will be systematic at least if a uniform protocol with regard to data treatment is applied. The latter should also allow to better understand the linkage with permeability, since data on pore throat characteristics will become available. It is obvious that the CT data can be compared with a large number of additional measurements carried out on the rocks, such as acoustic properties, permeability, NMR, … Of prime interest is also that based on the recent developments in CT research samples can be scanned at low as well as very high resolution. The latter data can then be projected into the lower resolution scans, allowing to work out an upscaling strategy. In addition based on pattern recognition or the application of Multiple Point Geostatistics it will be possible in the future to quantify lithotypes in a very objective way, and this together with the quantification of mineralogy and its distribution will lead to a better understanding of porosity – permeability – acoustic properties relationships. Based on the scanned data artificial FMI simulations can be calculated allowing linking the CT data to downhole logging. Obviously, the acquisition of the data should be done is a systematic way, and all research teams involved should store the data as agreed in the protocol. The interpretation of the data, however, should be the responsibility of each research or industrial group. Apart from the petrography, CT, petrophysical measurements, a number of geochemical analysis should also be stored in the database. The latter should especially address the reservoir modifying diagenetic processes.

Of interest would also be the creation of a rock library of representative reservoir rocktypes. This will allow all partners involved in the study to carry out additional analysis in function of proper scientific or industrial needs. General principles on how this rock repository could be managed can be learned from IODP for example.

How such a database can be created will be illustrated based on two case studies, namely one with regard to continental carbonates (subsalt exploration scenario) and a second on hydrothermal dolomites that likely formed due to fluid convection in platform carbonates in relation to magmatic intrusion that acted as heat and possible Mg source.

It is clear that the creation of such a structured database will on long-term be very beneficial for academia and industry. It will force to some extend to work out standard research protocols, and to consider the storage of data and reservoir rocks. Obviously, this will be a challenging task for all research and industry groups involved.

References

Bear, J., 1972, Dynamics of fluids in porous media. American Elsevier, New York.
Remeysen K. and Swennen R., 2008, Application of microfocus computed tomography in carbonate reservoir sedimentology: possibilities and limitations. Marine and Petroleum Geology, 25, 486-499.

AAPG Search and Discovery Article #120034©2012 AAPG Hedberg Conference Fundamental Controls on Flow in Carbonates, Saint-Cyr Sur Mer, Provence, France, July 8-13, 2012