Modern Carbon/Oxygen Logging: Hydrocarbon-Saturation-Determination Techniques
Modern Carbon/Oxygen Logging: Hydrocarbon-Saturation-Determination Techniques
50 MAY 2005
acquisition by calculating uncertainty in
the analysis of these variables.
A new method uses Monte Carlo simu-
lations to determine uncertainties in satu-
ration computations from resistivity data.
Errors in the resistivity analysis increase as
the formation water becomes fresher and
reaches a point at which the errors become
greater than those associated with C/O
logging. Because each evaluation
represents only one condition, several
simulations with varying porosity and
electrical parameters would be needed for
a complete evaluation.
Well Evaluations
Well A was a newly drilled well close to an
injector. The primary purpose of the well
was to determine the remaining oil satura-
Fig. 2Windowing and spectral-based techniques are used to compute the tion behind the flood front. Initially, a
relative abundance of carbon and oxygen in the formation. study was conducted in nearby wells to
determine the homogeneity of lithology,
porosity, and rock type as described by
compute hydrocarbon volumes. In the paper details tool designs, calibration petrophysical facies.
elastic mode, the neutron loses its energy techniques, and interpretation methods Samples from the Schlumberger MDT
until it is absorbed in thermal energy for the three tools. formation-tester tool were used in con-
levels. In these interactions, elemental junction with a tracer material to deter-
yields such as silicon, calcium, chlorine, C/O vs. Resistivity Logging mine formation-water resistivity, Rw, and
hydrogen, and iron are determined. Recent technological advancements have hydrocarbon mobility at several intervals.
A difficult aspect of this measurement greatly enhanced resistivity and PNS- Laboratory tests from these samples
is the differentiation between elemental logging tools. Slim-array-induction tools showed an Rw variation from 7,000 to
yields within these spectra. Energy peaks as well as newer cased-hole-logging tools 70,000 ppm NaCl. This variation in the Rw
from different elements often can cause also provide new options for reservoir profile produced large errors in resistivity-
background noise. Often, windowing monitoring. Both technologies have calculated saturations. Water saturations
and spectral techniques are used to merit, but limitations and aspects exist from resistivity analysis did not compare
overcome these problems. Fig. 2 shows that could make one measurement more well with the core data. In addition, an
how windowing is used to determine appropriate than the other under NMR log/inject/log method was used to
elemental yields. certain conditions. The following factors calculate oil volumes. There was an excel-
In the inelastic mode, the most impor- should be considered with C/O and lent match between the NMR and core-
tant measurement is the C/O ratio. Test- resistivity logging. measured values. NMR and core-measured
pit characterizations provide charts that Formation-water-salinity range and values were used as benchmarks.
relate this value to the fluid saturations. profile. The Schlumberger C/O tools then were
A saturation point is determined by cross- Borehole conditions. run in conjunction with production-
plotting porosity and C/O ratio. In Reinvasion of borehole fluids. logging tools. Several passes were made
addition to these charts, service and Resistivity contrast. with both shut-in and flowing conditions.
oil companies developed local relation- Porosity range. The tools provided saturation results that
ships by use of Monte Carlo simulations Scaling and acidizing. were in agreement, except for the small-
and local reservoir knowledge. Casing condition. diameter tool (RST-C) when used in flow-
Cement condition. ing conditions (a method that is not
Tool Design/Characterization Completion restrictions. recommended by the service company).
During the last decade, several new These factors should be evaluated in Because the small tool does not have
pulsed-spectroscopy tools have been every well, and consideration should be shielding, both near and far detectors are
introduced. In this study, Schlumbergers taken in the planning phase of every sensitive to the borehole environment.
Reservoir-Saturation Tool (RSTPro), logging job. This planning will minimize This sensitivity adds uncertainty to the
Halliburtons Reservoir Monitoring Tool systematic errors in the final answers. In calculated saturations because of a smaller
(RMT-Elite), and Baker Atlas Reservoir addition, random errors inherent in the dynamic range in the interpretations.
Performance Monitoring tool (RPM) tool measurements also should be
were evaluated in two wells in this considered. Service companies provide Well B. The objective in Well A was to
carbonate reservoir. Even though all C/O-job planners that compute random determine remaining oil saturations. To
these tools operate on the same princi- errors on the basis of porosity, borehole determine the effects of borehole fluid and
ples, they are vastly different in design condition, and acquisition speed. These changes in formation saturation, a well
and characterization. The full-length planners are used to optimize data with 50% water cut was selected. Well B
MAY 2005 51
was a candidate for propellant stimulation. RMT-Elite data showed reasonable
The logging objective was to optimize the results in most of the sections except in
treatment interval, therefore the oil/water zone A, for which it overestimated the
contact or unflushed zones needed to be water saturation. In this zone, production
determined. One of the primary objectives logs measured high water-production rates.
was to determine the oil saturation in In Zones B and C, benchmark satura-
Zone B. tion values from resistivity and sigma data
Well B was completed open hole. supported C/O-log results from RST-D and
Again, a complete geological and petro- RPM, but did not provide clear evidence of
physical analysis of nearby wells showed unflushed intervals, and the stimulation
homogeneous rock types. Core data were treatment was cancelled. An assessment
not available in this well, so an alternative based on RMT-Elite data alone would have
benchmark was needed. left at least two zones as candidates for
A test program was developed to com- propellant stimulation.
pare C/O data from the three service com- Borehole holdup measurements from
panies against saturations computed from the RST-D compared well with produc-
resistivity and pulsed-neutron-capture tion-logging results. The results were key
sigma-based data. in sigma-logging measurements.
The following observations were made RPM does not provide good holdup
from this well study: results.
Borehole-fluid sampling and sigma
wellbore measurements showed good Conclusion
matches with water salinity greater than One of the key elements in successful
100,000 ppm NaCl. At these high salini- C/O-log interpretation is planning.
ties, robust saturation measurements can Formation-water salinity, borehole condi-
be expected from both resistivity and tions, reinvasion of borehole fluids, resis-
pulsed-neutron-capture sigma logs. tivity contrast, porosity, scaling/acidizing,
Normalized sigma-measured values casing condition, cement condition, and
for the three companies were identical, completion restrictions must be taken into
thus providing similar saturation profiles. consideration before any job. In addition,
Hydrocarbon saturations measured by planning should involve a sensitivity
resistivity and sigma data compared well analysis. Monte Carlo simulations on
across the entire logged interval. resistivity-based saturation data and C/O-
In a comparison between the flowing job planners were used to optimize the
and shut-in passes, the large-diameter-tool saturation results. JPT
(RST-D) data provided reasonable reinva-
sion profiles. RPM data showed few rein-
vasion effects, while a large difference was
noticed with the RMT-Elite.
For a limited time, the full-length paper is
RST-D inelastic data matched well
available free to SPE members at
with both sigma and resistivity data.
www.spe.org/jpt. The paper has not
RPM data overestimated water vol-
been peer reviewed.
umes with respect to references.
52 MAY 2005