Content of PetroWiki is intended for personal use only and to supplement, not replace, engineering judgment. SPE disclaims any and all liability for your use of such content. More information
The determination of porosity is paramount because it determines the ultimate volume of a rock type that can contain hydrocarbons. The value and distribution of porosity, along with permeability and saturation, are the parameters that dictate reservoir development and production plans.
Determination of porosity from a wireline log is only part of the problem, because the values determined in one well must be upscaled into the space between wells. To extrapolate correctly, the team must identify depositional environments and rock types and then have access to analog data sets. Only then can the correct statistical distributions be extrapolated across the reservoir. (See the article on geostatistics).
If one has access to an undamaged whole core from the reservoir in question, direct measurement of the porosity is possible, if care is taken. Some profess that all cores are damaged by the coring process, thus no accurate assessment of porosity is possible by coring. Others disagree. X-ray computerized tomography scans, thin sections, and scanning electron microscope examination can verify that grain contacts are unmoved, authigenic pore linings are undamaged, and heavy-weight drilling fluids and/or particles are absent. Thus, given a satisfactory core, porosity can be determined accurately by several different methods specified in API RP 40.
There are a number of pitfalls to avoid. One pitfall is that cleaning the core of crude and brine must be both thorough and gentle. One must remove all the heavy ends of the crude (typically asphaltenes) but not damage the authigenic minerals. NaCl crystals left from the in-situ brine when the water is removed must also be removed, but the authigenic clay minerals must not be removed or disturbed. Tar must not be removed if it occurs naturally in situ, and NaCl pore linings must not be removed if it occurs naturally in situ. Gypsum, when present, must not be dehydrated during the determination of porosity, and hydroxyl water of clay minerals must not be removed and counted as part of the pore space. Pore space must not be created during the cleaning process by flowing large volumes of fluid through rocks with soluble grains, such as gypsum, anhydrite, limestone, or salt.
Another pitfall to avoid is that some rock samples are mechanically weak and uncompact when brought to the surface and freed from the overburden load. These rock samples must be returned to in-situ conditions of effective stress to return the rock sample to its in-situ value of porosity. Correct procedures to make these measurements have been published by several authors. Typically, stressed porosity measurements are more time consuming and more costly. If the core is subdivided into rock types, empirical correlations between stressed and unstressed measurements on a given rock type can be used to extrapolate to a larger data set.
An additional pitfall is that some sandstones contain fibrous pore-bridging clay minerals. These clay types are fragile and are damaged by most routine methods used to clean the core before porosity determination. When this type of clay is determined to be present, all subsequent core cleaning must be done using critical point drying methods. In addition to the methods defined in API RP 40, nuclear magnetic resonance (NMR) spectroscopy methods can be used to determine porosity on small core samples. When care is taken to ensure that the core sample is 100% saturated with water and or liquid hydrocarbon, the NMR method yields an accurate value of porosity. The NMR method for porosity determination can also be used in the borehole using wireline NMR tools with excellent results. The article on NMR logging shows examples.
Direct determination of porosity by core analysis is the "gold standard" and is used where available to calibrate all indirect measurements. Indirect methods allow leveraging of limited core data to provide more information on areal and vertical variations in porosity when coring is too expensive and/or only partial cores are recovered.
Patchett and Coalson determined that the density log is the most accurate method to determine porosity when one has knowledge of grain density and fluid density. While this method is standard in production wells, these parameters are often unknown for wildcats. Grain density can change rapidly along the borehole as lithology changes. Fluid types and saturations change more slowly, except at fluid contacts. Thus, we have four unknowns: porosity, grain density, hydrocarbon saturation, and water saturation, and one measurement: bulk density. A statistical method would be used to combine the environmentally corrected log readings from the density, neutron, acoustic, and GR to solve for the four unknowns. Often, a shallow resistivity log is included in the mix, and the acoustic log is dropped. Every logging vendor provides an answer product that uses this type of method, and most larger oil companies have published their own method. If we are certain that we are below the free water level and if we assume that we know the rock type is sand with a grain density of 2.65 g/cm3, the measured density log reading of bulk density can be converted into porosity, as the article on nuclear logging shows.
Another simplified method is to use the density-neutron crossplot provided by each of the logging vendors. Patchett and Coalson found no benefit to using the density-neutron crossplot over a density log if one used known and variable grain density. The novice often uses a simple method with little regard for changes in rock type, fluid type, or borehole conditions, and the result is considerable error in the determination of porosity. The density log can be quite accurate when logged in ideal to semi-ideal borehole conditions. However, in rugose boreholes, extremely thick mudcakes, or unusual weighting materials in the mudcake (e.g., hematite), the bulk density readings seen on the log will no longer reflect those of the borehole wall, and these readings must be discarded and the value of porosity determined by other methods. The borehole caliper and density correction curves are used to validate the quality of the bulk density readings.
The foregoing discussion on sands provides an answer for porosity that is correct, but it reflects the average porosity over the depth of resolution of the tools (i.e., approximately 3 ft). Sophisticated digital processing can increase the resolution to approximately 1 ft. Thus, when the reservoir is heterogeneous on a scale smaller than 1 ft, one must use other methods to deconvolve the resulting averages into values that reflect the true porosity of the individual rock types. One of the most common heterogeneous reservoirs is the laminated sand-shale sequence, in which the shale layers are often less than 1 in. thick. One published method used to determine the porosity of the sand layers free of the unresolved shale layers is the Thomas-Stieber method. In a sand-shale reservoir in which the shale laminations have a porosity lower than the sand layers, one will consistently understate the value of reservoir porosity if the unresolved shale laminations are not properly accounted for. Furthermore, if one is using porosity-to-permeability transforms, the value of permeability will be underpredicted.
Determination of porosity in carbonates is generally straightforward unless the rock type is one with large vugs (i.e., fist-sized or larger) or fractures. Density-neutron and neutron-acoustic crossplot have been historically useful and accurate when calibrated to core measurements. When the rock types become complex and numerous, then statistical, multiple-log methods that match the number of unknowns to independent log measurements are required. Every logging vendor provides an answer product to produce a reasonable value of porosity when all logs are environmentally corrected and validated. Large vugs can be spotted with borehole image logs (see specialized logging topics) and with large diameter cores (see relative permeability and capillary pressure). With appropriate sampling, the borehole readings can be corrected for the effects of large vugs. Logging tools that investigate larger volumes are given higher weights in the analysis.
The dual-porosity system that exists in a fractured matrix reservoir provides a challenge in the opposite direction in that often the overall value of porosity is quite low (2 to 3%). Because most wireline porosity logs have random statistical error of 1 to 2%, the error is as big as the value being measured. Under these conditions, reservoir simulation and history match is the most reliable method to determine storage capacity and reserves, and porosity becomes moot. Borehole image logs are used to locate the fractures and provide probable production intervals.
Application of various logging tools to porosity calculation
Different logging tools can provide different types of insight to assist in calculating porosity. The best method may be to rely on a combination of approaches where the information is available.
- API RP 40, Recommended Practices for Core Analysis, second edition. 1998. Washington, DC: API.
- Swanson, B.F. and Thomas, E.C. 1980. The Measurement of Petrophysical Properties of Unconsolidated Sand Cores. The Log Analyst (September–October): 22.
- Wei, K.K., Morrow, N.R., and Brower, K.R. 1986. Effect of Fluid, Confining Pressure, and Temperature on Absolute Permeabilities of Low- Permeability Sandstones. SPE Form Eval 1 (4): 413-423. SPE-13093-PA. http://dx.doi.org/10.2118/13093-PA
- Wawak, B.E. and Campbell, W.L. 1986. Characterization of Clay Fabric Using Critical Point Drying to Preserve Clay Texture and Morphology. Scanning Electron Microscopy 4: 1323.
- Patchett, J.G. and Coalson, E.B. 1982. The Determination of Porosity in Sandstone: Part Two, Effects of Complex Mineralogy and Hydrocarbons. Paper T presented at the 1982 Annual Soc. of Professional Well Log Analysts Symposium, Corpus Christi, Texas, 6–9 July.
- Peeters, M. and Visser, R. 1991. A Comparison of Petrophysical Evaluation Packages: LOGIC, FLAME, ELAN, OPTIMA and ULTRA. The Log Analyst 32 (4): 350.
- Thomas, E.C. and Stieber, S.J. 1975. The Distribution of Shale in Sandstones and Its Effect on Porosity. Paper presented at the 1975 Annual Soc. of Professional Well Log Analysts Symposium, New Orleans, 4–7 June.
Noteworthy papers in OnePetro
Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read
Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro