|
|
PfEFFER ConceptsGeneral |
In most cases, either a deep-reading resistivity log or a computed estimate (corrected for invasion characteristics) of true formation resistivity, Rt, should be used, so that saturations are good estimates of those in the undisturbed reservoir. However, shallow-reading resistivity devices (such as microresistivity tools) can also be used, but their saturations will reflect those of the flushed zone, close to the borehole wall. This information can be useful, particularly when compared with deep-resistivity saturation data, as a quantitative indicator of moveable hydrocarbons.
In PfEFFER, the Rt selection button is the site of the user-selected resistivity log. Values in these cells are combined with porosity values in the adjacent PHI column to automatically generate water saturation and other variables, through application of the Archie equations and parameter values in the PARAMETER tab. The choice of resistivity log dictates the meaning of the computed fluid saturations. For a resistivity tool whose depth of investigation reaches beyond the invasion front, the reading will be an estimate of Rt, the undisturbed formation resistivity. If a shallow-focused tool such as a microresistivity device is selected, then the saturations in the flushed zone are estimated. However, in this case, remember to substitute a mud filtrate resistivity value (at formation temperature) in the RW parameter cell. PfEFFER also provides a movable hydrocarbon option for users who wish to evaluate residual, movable, and total hydrocarbon saturation simultaneously.
The porosity of any zone can be estimated either from a single "porosity log" (sonic, density, neutron, or magnetic resonance log) or a combination of porosity logs, in order to correct for variable lithology effects in complex reservoirs. When using a single porosity log, the true porosity is calculated from interpolation between the values for the matrix mineral and the pore fluid (usually equated with mud filtrate, because of the shallow investigation of the porosity tools). So, if a bulk density log is used, then porosity is calculated from the relationship:
where rb is the bulk density, f is the porosity, rma is the matrix density, and rf is the pore fluid density. If a sonic log is used for porosity estimation, the equivalent relationship is:
where Dt is the zone transit time, f is the porosity, Dtma is the matrix transit time, and Dtf is the pore fluid transit time. Modern neutron logs are recorded in apparent porosity units with respect to a given mineralogy. Calcite is commonly chosen as a default mineral, in which case the porosity values will be true porosities in limestone zones. Where zones are not limestone, the limestone-equivalent neutron log should be rescaled to the zone matrix mineral or combined with a density limestone-equivalent porosity in an estimate of the true porosity.
The combination of density and neutron logs is now used commonly as a means to determine porosity that is free of lithology effects. Each individual log records an apparent porosity that is only true when the zone lithology matches that used by the logging engineer to scale the log. The true porosity may be estimated either by taking an average of the two log readings or by applying the equation:
where fn and fd are neutron and density porosities.
The alternative options of calculating porosity from sonic, density, neutron, or the density/neutron log combination are provided by PfEFFER and the results are stored in the PHI column of the Computation section. In some situations, users may want to differentiate between "primary" porosity (intergranular and intercrystalline) and "secondary" porosity (vugs and fractures). It is widely perceived that the sonic log is sensitive only to the primary porosity, while the neutron and density logs reflect total porosity. PfEFFER provides a secondary porosity option to apply this concept for interested users.
The calculations presented for porosity estimation from a single log assume that shale effects are not significant or that they can be disregarded. The correction for shale involves an expansion of the porosity estimation equations to:
| for density |  |
|
| for sonic |  |
|
| for neutron log |  |
where Vsh is the fraction of shale estimated from log(s) and rsh, Dtsh, and fnsh are readings of representative shales on the density, sonic, or neutron log.
There are a number of methods to estimate the proportion of shale, Vsh, from wireline logs. When using a single log (typically a gamma-ray or SP), Vsh of a zone is estimated by interpolating the zone log reading, L, between the extremes of a log reading of a "clean" (shale-free) formation, C, and a reading of a typical shale, S. The equation is then:
This estimate is often known as the "gamma-ray index"
and represents the proportion by weight of radioactive material calibrated
to the reference shale. Some log analysts
apply one of a variety of functions to estimate the volume of clay, Vcl.
This relationship varies with the stratigraphic age and sedimentary
facies of the shale and so is not included in PfEFFER.
However, it has been reported in the literature (e.g. Yaalon, 1962)
that the "average" shale contains approximately 60% clay minerals. Therefore, as a first approximation: 
as suggested by Bhuyan and Passey (1994).
The spontaneous potential log can be used in the place of the gamma-ray log in this estimation procedure. However, there is not a linear relationship between shale content and the SP response (Griffiths, 1952), and calculated values are almost always overestimates.
If two logs are used (most commonly the neutron and density logs) and are calibrated to the clean reservoir lithology, then the estimate is given by the equation:
The analyst provides the boundary values from an inspection of the section itself because of the extreme variability of shale properties both laterally and with depth.
The estimate of Vsh is a pragmatic measure and it is common for several alternative measures to be computed together. Field experience has suggested that log measures of shale tend to be overestimates, so it is common practice to select the lowest figure as the best estimate. The values are stored in the VSH column of the Computation section.