Many reservoir properties can be semi-quantified using a suite of conventional wireline logs to calculate porosity and the application of Archies Law to calculate water saturation. Carbonate reservoirs tend to be petrophysically more complicated than sandstone reservoirs because of influence of life processes on carbonates at deposition and the ongoing overprints of either diffuse or more focused diagenetic fluid crossflows in near-surface and subsurface realms.
Many carbonate reservoirs contain more than one porosity style with geometries in the reservoir volume that are not stratified (e.g. interparticle versus isolated vug and touching vug porosity. Even with reasonable porosity values showing up in the conventional well logs readings we use to estimate porosity (density, neutron and sonic), a carbonate host with high levels of isolated vug porosity can lack free-flowing pore fluids and have negligible effective porosity. Our relative inability to quantify in a carbonate compared to a sandstones complicates well log interpretation and the application of Archies law in storage and volumetric saturation estimates.
Application of Archies Law to siliciclastic reservoirs is much simpler than carbonate reservoirs, as porosity style is intergranular and reservoir volumetrics are stratiform. Sandstone reservoirs can be reliably modelled by an understanding of stratabound depositional sand geometry, in combination with knowledge of styles of clay distribution in the reservoir sands. A Vclay based on a gamma output is usually sufficient to define reservoir net to gross volumetrics. Dealing with most carbonate reservoirs worldwide requires additional understandings in terms of quantified diagenetic geometries, porosity types and structuration histories (Figure 1).
1. Petrophysical classifications of carbonate pore types as first defined by Archie (1952) compared with the fabric-selectivity concept of Choquette and Pray (1970) and Lucia (1983).
Many carbonate reservoir to nonreservoir layers remain largely undifferentiated using either a total or a spectral gamma log (Figure 2). This reflects the lack of radiogenic clays in most carbonate intervals and the ability of uranium to be mobilised in crossflows of subsurface waters. So, in this case, what exactly does a Vclay calculation mean in terms of reservoir determination (net to gross). In my experience, not much. Only when other more responsive well-logs are used, typically via a neutron-density overlay, do porosity-relevant reservoir layers become resolvable (Figure 2). Only then can Vclay be estimated using this responsive well log suite and so used to calculate a meaningful net to gross.
2. Conventional well-log signatures of a carbonate succession in the Permian Lower Clear Fork Fm. Texas. Note the lack of response in the gamma log to changing lithologies, as logged in the core. Density and neutron values better reflect lithology and measured core plug porosities (after Ruppel and Jones 2006).
Evaluation or quantification of porosity in any hydrocarbon reservoir varies according to the measurement technique and scaling of the sample volume (Figure 3). If the value comes from a well log (density, neutron or sonic), then we are measuring a response to nuclear bombardment or the passage of a sound wave. This samples a cylinder or ellipse of information collected in the vicinity of the wellbore. We then apply a set of assumptions as to matrix and fluid properties to convert that physical measurement into an estimate of porosity.
Porosity from a density log is derived from medium-energy gamma-ray bombardment, an understanding of the Compton effect at the atomic level and an indirect measurement of electron density (bulk density). Emitted gamma rays lose energy when they collide with the electrons in the formation (Compton effect). The number of collisions is related to the bulk electron density of the formation, and the bulk density is related to porosity and density of common minerals found in carbonate reservoirs. To determine porosity from a bulk density value requires knowledge of the likely matrix density and fluid density.
Porosity measurement from a neutron log is derived from a higher energy nuclear source in the logging tool, which is bombarding the formation with neutrons and the fact that highest energy attenuation occurs with neutron interaction with an atom of equal mass, namely hydrogen. Application of a hydrogen index understanding is then calibrated against a limestone standard of known porosity, as first calibrated in a measurement pit constructed in the 1950s in the grounds of the Petroleum Engineering Building at the University of Houston, Texas (Figures 4 and 5). This standardised neutron output is used to define the worldwide standard and is calibrated in limestone-equivalent-porosity-units. Before standardisation in the 1960s, early neutron tools measured in counts/sec. A neutron porosity measurement must be adjusted whenever the neutron tool is run in a dolomite or a sandstone interval.
Porosity measurement using a conventional sonic logging tool is based on measuring interval transit time (delta t, measured in microseconds per foot ), which is defined as the reciprocal of the sonic velocity measured in ft/sec (Figure 6). To calculate porosity from a measurement of interval transit time requires the input of the matrix velocity and the fluid velocity.
Porosity derived in this way from a conventional compensated sonic tool measures the velocity of first arrivals, and so gives the delta t of the soundwaves taking the fastest path between the sound source and receiver. Thus a porosity value calculated from a conventional sonic log tool does not see the return of sound waves that have passed through larger fluid-filled cavities such as vugs and open fractures. These sound waves will past back to the tool but are slower and so arrived later not detected. Detecting these slower sound waves is the basis for the shear-sonic tool (Figure 6).
In contrast, a porosity estimate based on values measured by a density or neutron tool is a calculation of total porosity in the sampled region. Unlike the sonic tool, which runs centred in the borehole and makes a global reading, density and neutron tools are mounted on the head of a spring-loaded arm, with a plough in front of the tool head. This tool set-up is designed to push aside the mud-cake and so maximise contact of the tool with the borehole wall. It also means the sampling volumes of the neutron and density tools are much smaller than the circum-borehole volume sampled by the sonic tool. So porosity calculations from the sonic tools are made on a larger volume than either the neutron or density tools.
In a homogenous sandstone reservoir, the difference in various porosity tool sampling volumes should be similar. With heterogenous polymodal porosity in a complex carbonate reservoir, the calculated porosities from the three tools can differ. In an in-gauge borehole opposite a sandstone reservoir, the porosity calculation usually comes from an averaging of the density and neutron derived porosities. In a more rugose borehole, where the density-neutron measurement arms may not maintain good contact with the borehole wall, a more reliable porosity calculation tends to come from the centred borehole sonic tool readings, not the neutron or density outputs.
Then there is a potential problem in comparing porosities calculated from the first arrivals using a conventional sonic tool, with the "total" porosities derived from the density and neutron tools. For example, a sonic tool porosity calculation may not see fluid-filled vuggy or fracture porosity. By calculating porosity on a common scale using both the sonic and density inputs, a zone of separation can be useful in identifying intervals of fluid-filled fracture or vuggy porosity (Figure 7). However, this approach does not differentiate between fracture (touching-vug) and vuggy (separate-vug porosities). The effects of the two porosity types (fracture versus vug) on fluid flow volumes in a carbonate reservoir are very different, and their quantification for a reservoir model usually requires an image-log (FMI or UBI).
In a study of separate vug oomouldic porosity in a Middle Eastern carbonate reservoir, Wafta and Nurmi, 1987 concluded separate-vug porosity levels could be estimated from the equation;
ϕvug = 2(ϕt – ϕs)
where ϕs and ϕt are the calculated sonic log porosity and the neutron–density (total) porosity log, respectively. Subtracting the level of oomouldic (vug) porosity from the total porosity leads to improved estimates of flowable reservoir storage volume.
A neutron tool sees hydrogen abundance. In a simple reservoir model for a clean sandstone reservoir, the assumption is that this hydrogen resides in water or hydrocarbons in the intergranular pores of the reservoir. But the neutron tool also sees as porosity, hydrogen that resides not in a pore, but in a clay mineral (OH), or as non-moveable water bound to clay surfaces, and non-moveable structural water in minerals like gypsum or carnallite (Figure 2). It is also why coals show high levels of neutron porosity, which is actually from hydrogen held in compressed organic matter, making up the majority of the coal matrix. A density log sees electron density that is converted to a bulk density value. A density log does not see OH radicals in a clay mineral as porosity, but it does see clay-bound water and structural water as porosity (Figure 2).
All three porosity tool outputs cannot differentiate microporosity from larger-scale porosity (Figure 2). Once again, this is not usually a problem in well-log porosity calculations (storage volume) in a clean sandstone reservoir as generally there are only minor levels of microporosity present. But, in some carbonate reservoirs, microporosity can contribute up to 50% of the total porosity seen by a well log.
3. Estimates of porosity can vary according to the measurement technique
4. Neutron and gamma log calibration pits, University of Houston, USA. A) general view in the grounds of the Petroleum Engineering Department. B) Placing of the limestone calibration blocks in the pit. C) View down onto the cavity before placement of the blocks and filling with water. D) Running the tool in order to calibrate.
5. Components and dimensions used in the construction of the neutron test pit, pit is located in the grounds of the Petroleum Engineering buiilding in the University of Houston.
6. The geophysical wavetrain received by a sonic log. V is the velocity in ft/sec or m/sec.
Archie, G. E., 1952, Classiﬁcation of carbonate reservoir rocks and petrophysical considervations: Bulletin American Association Petroleum Geologists, v. 36, p. 278-298.
Choquette, P. W., and L. C. Pray, 1970, Geologic Nomenclature and Classification of Porosity in Sedimentary Carbonates: Bulletin American Association Petroleum Geologists, v. 54, p. 207-250.
Lucia, F. J., 1983, Petrophysical parameters estimated from visual description of carbonate rocks: a field classification of carbonate pore space: Journal of Petroleum Technology, March, v. 35, p. 626-637.
Ruppel, S. C., and R. H. Jones, 2006, Key role of outcrops and cores in carbonate reservoir characterization and modeling, Lower Permian Fullerton field, Permian basin, United States, in P. M. Harris, and L. J. Weber, eds., Giant hydrocarbon reservoirs of the world: From rocks to reservoir characterization and modeling, AAPG Memoir 88/SEPM Special Publication, p. 355-394.
At Saltworks, the aim of all our training modules and workshops is two-fold. 1) give an understanding of the relevant process, 2) train participants in the application of the skill sets tied to the concept and prioritise the skill sets needed to apply this understanding. Below we illustrate the skills and knowledge necessary to recognise subsurface evaporite salts using a conventional suite of well logs.
If you want to know more, please download the relevant saline geosystems or carbonate geosystems catalogue and choose a combination of the various training modules that best suite your company needs.