FORMATION EVALUATION

Whole Core vs. Plugs: Integrating Log and

Core Data to Decrease Uncertainty in Petrophysical
Interpretation and Oil-In-Place Calculations

Most carbonate reservoirs are char- Φ=22% Φ=50%

acterized by multiple-porosity systems Φ=11%
that impart petrophysical heterogene- Φ=10 % Φ=60%
ity to the gross reservoir interval. The
specific types and relative percentages
of pores present, and their distribution
Φ= 18% Φ=24%
within the rocks, exert strong control
on production and stimulation char-
acteristics of carbonate reservoirs. The Φ =22%
Φ=
Φ= 22%
effects of heterogeneity on core and
log measurements are presented. The
challenge is to determine the reliability
Φ= 22% Φ =23%
Φ=
of applying relatively small-scale prop-
erties measured by a log or core to the
large-scale reservoir property.

Introduction
Most sandstone reservoirs are single-
porosity systems (i.e., interparticle Frequency
pores) of relatively uniform (homo-
geneous) nature. However, most car- 0 Porosity, % 60
60

bonate reservoirs are characterized by

multiple-porosity systems. Reservoir
heterogeneity results from physical
and chemical reorganization processes Fig. 1—Varying plug porosities taken from a heterogeneous carbonate
over time (e.g., compaction, solution, core.
dolomitization, and cementation).
Heterogeneity complicates the reser- as a function of spatial locations in ver- orders of magnitude has been reported
voir-description task, in which reservoir tical and areal directions. Vertical het- over a distance of a few centimeters in a
properties (e.g., permeability, porosity, erogeneity has shown large effects on carbonate-core plug. In heterogeneous
saturation, and rock types) tend to vary sweep efficiencies of waterfloods and reservoirs, the real challenge is pre-
conventional gas injection into mixed- dicting reservoir performance through
This article, written by Senior Technology to oil-wet carbonate reservoirs. These the acquisition of representative meas­
Editor Dennis Denney, contains highlights effects are a result of large variations in urements. The challenge would be to
of paper SPE 137679, “Whole Core vs. permeability between different strata. reconcile the variability seen in high-
Plugs: Integrating Log and Core Data Permeability and other rock properties resolution (e.g., trim or plug samples)
To Decrease Uncertainty in Petrophysical can vary along the areal (lateral) direc- and low-resolution (e.g., wireline logs)
Interpretation and STOIP Calculations,” tion of the reservoir, which necessitates measurements and to quantify the tech-
by S. Serag El Din, SPE, ADCO; establishing an accurate and detailed nical and physical considerations at the
M.R. Dernaika, SPE, Weatherford understanding of geological heteroge- various scales.
Laboratories; I. Al Hosani, ADCO; neities and their effect on petrophysics Routine and special-core-analysis-
L. Hannon, Weatherford Laboratories; and reservoir engineering. laboratory (SCAL) measurements were
S.M. Skjæveland, SPE, University of The effect of rock heterogeneity made on carbonate-core samples of
Stavanger; and M.Z. Kalam, ADCO, pre- becomes greater when petrophysical various sizes ranging from small trims
pared for the 2010 Abu Dhabi International properties are measured at scales rang- to full-diameter whole-core samples.
Petroleum Exhibition & Conference, Abu ing from a small trim sample in the The measurements were analyzed to
Dhabi, UAE, 1–4 November. The paper laboratory to log measurements in the detect possible variability in rock prop-
has not been peer reviewed. field. Permeability variation of three erties influenced by the scale issue. The

For a limited time, the full-length paper is available free to SPE members at www.jptonline.org.

58 JPT • AUGUST 2011

 Φ=22% Φ=22%
results then were interpreted with log Φ=22%
data to study the effect of the measured Φ=22% Φ=23%
volume on the results, as a step toward
reconciling laboratory data with logs.

Scale of Measurements Φ=23% Φ=23%

Laboratory measurements can be con-
ducted on multiple sample sizes with
varying techniques and procedures. Each Φ=22% Φ=22%
technique has advantages and disadvan-
tages, with underlying estimations and
assumptions. In this paper, “trim” refers Φ=23% Φ=23%
to any small-sized, usually irregular, rock
sample taken from the main core plug
for the purpose of side analysis.
The 1-in.-diameter×1-in.-long plug
sample should yield more-accurate
mercury-derived porosity/permeability
data and capillary pressure data than
the trim sample because of higher pore
Frequency
volumes and, in turn, higher accuracy,
especially in heterogeneous or vuggy
rocks. It is a standard practice to meas­ 0 Porosity, % 60
60

ure helium porosity on this sample

size. Absolute permeability is measured Fig. 2—Similar plug porosities taken from a heterogeneous carbonate
core.
accurately on 1-in.-diameter plugs, if
the sample length is longer than 1 in., to on contact-angle hysteresis, wettability, variability seen between high-resolu-
minimize end effects. This sample size and irreducible fluid saturations. The tion measurements (e.g., core plugs)
could be used to obtain accurate forma- relative presence and distribution of and relatively low-resolution measure-
tion resistivity factor and cementation concave and convex pores are critical ments, such as whole-core or density-
exponent m. However, resistivity-index in determining wetting-film thickness, log measurements.
data and any other multiphase-flow which is a direct cause of wettability
experiment are questionable because alteration and, hence, an influence on X-Ray Scanning. X-ray computed-
of the high uncertainty involved in the fluid-flow behavior. tomography (CT) scanning enables
calculation of water saturations in such Variations in the distribution of pore viewing the sealed core, which is cru-
small-pore-volume samples. and throat sizes within the sample cial when core exposure must be avoid-
The 1.5-in.-diameter×3-in.-long create statistically different porosities ed or minimized. This technique allows
plugs are used in SCAL experiments. and permeabilities at different loca- examining the internal structure of the
This sample size could be used to obtain tions. The distribution of different pore core at different scales and identifying
many routine and SCAL data in mul- and throat sizes also could be viewed the degree and distribution of hetero-
tiphase-flow experiments. Composite as variations in the distribution of vugs geneity in the core. Often, it is used in
samples composed of several core plugs and/or microporosity, which can cause the core-characterization phase of core
are used to increase the sample pore severe degrees of heterogeneity. analysis for the selection of representa-
volume, especially in multiphase-flow Fig. 1 shows a series of 11 core plugs tive samples.
tests involving miscible- and immisci- (and their measured porosity values)
ble-gas injections. taken from a heterogeneous carbonate. Water-Saturation Uncertainty
The measured porosity at the core-plug Water-saturation values calculated
Reservoir Heterogeneity scale is highly variable, depending on from openhole resistivity measure-
Variations in the mineral composition whether an individual plug intersects a ments are needed for oil-in-place cal-
can yield different reactions between large hole (upper right of the inset) or culations. Fluid saturations depend on
the fluid(s) and the rock surface, which a low-porosity cemented region (light the accuracy of the Archie cementation
can have direct effect on the water-film colors). A porosity measurement with factor m and saturation exponent n.
thickness and, hence, on the distribu- a larger volume of investigation (such Archie’s equation provides a relation-
tion of contact angles and wettability. as a density tool) would read an aver- ship linking measured resistivity with
Performing measurements at different age of several plugs, resulting in a formation volume (i.e., porosity and
scales may alter the percentage of a smaller variance. Fig. 2 shows a series saturation). A crucial component of the
certain type of mineralogy over another of 11 core plugs (and their measured equation is m. This parameter can have
type of mineralogy present in the rock porosity values) taken from a homoge- a great influence on understanding the
sample, which may cause variations in neous carbonate (inset). The measured fluid distribution in the reservoir, and
the macroscopic results. Roughness of porosity at the core-plug scale varies use of unrepresentative m values can
the rock surface has a direct effect on little. The effect of heterogeneity on yield wrong estimations of oil in place
the water-film thickness and, hence, the core and log is concerned with the in the reservoir.

JPT • AUGUST 2011 59

 Uncertainty analysis was performed differences between whole cores and did the difference between whole-core
for the cored reservoir used in this plugs varied from sample to sample. and plug data. The match of poros-
study to determine which parameters However, very tight rocks showed ity values between whole-core and log
had the greatest effect on the water- higher whole-core permeabilities than data was excellent and was best among
saturation and oil-in-place calculations. plug permeabilities. Plugs with a low- all scale measurements. This result
The value of m had the greatest influ- er-permeability range (<10 md) had emphasizes the importance of whole-
ence on predictions of water saturation lower permeabilities than the whole core measurements to represent the
in the reservoir. The second-most-influ- cores, and plugs at the higher-permea- reservoir behavior best by capturing
ential parameter was n. Total-resistivity bility range (>10 md) showed higher the heterogeneity in the rock property.
and water-resistivity values were least permeabilities than the whole cores. All whole cores showed lower m values
important. These results indicated the This observation could be linked to the than the corresponding plug samples.
importance of having accurate core- degree and distribution of heterogeneity This result has a direct effect on satu-
analysis measurements performed in the rock samples. There are different rations and estimates of oil in place.
on representative core volumes (e.g., explanations for permeability variations Whole-core measurements yielded a
whole-core samples in heterogeneous seen with scale (sample size). Small lower composite-m value, and this, in
reservoirs) that can yield representative sample sizes (e.g., plugs) could overes- turn, would have a direct effect on the
m and n values. The greatest advan- timate permeability measurements by reservoir oil-in-place calculations.
tage of electrical measurements on het- short-circuiting flow through porosity
erogeneous whole cores is obtaining channels (e.g., vugs) with dimensions Conclusions
representative m values that lead to similar to the plug length. Flow paths in • Whole-core measurements are crit-
better reservoir characterization and larger whole-core volumes may exploit ical for obtaining core-analysis data in
more confidence in volumetric calcu- optimal connection pathways that tend heterogeneous and dual-porosity sys-
lations and reserves estimates and in to be unavailable within smaller subsets tems, such as fractured and vugular
simulation-model initialization. of the total rock volume. rocks in which the scale of local hetero-
In this case, the high-permeabili- geneities is greater than the dimensions
Experiment Measurements ty samples (>10 md) generally were of the plug samples.
and Interpretations more heterogeneous and vuggy than • Whole-core porosity tended to
Fit-for-purpose conventional core anal- the lower-permeability samples, as be systematically lower than porosity
ysis and SCAL testing were carried out revealed through corresponding CT determined for plugs.
on a carbonate core from the Middle images. Such a high degree of hetero- • Porosity from whole-core samples
East. The objective of the program was geneity would be more pronounced provided a better match with log poros-
to measure macroscopic rock proper- in smaller-volume rocks, which ity than with plug porosity.
ties at various scales to identify the could easily overestimate permeabil- • In low-permeability samples
effect of heterogeneity on different vol- ity through relatively larger-porosity (<10 md), whole cores tended to have
ume measurements. The full-length channels. In the lower-permeability higher permeability values than plug
paper details the testing and results. range of samples, larger rock volumes measurements had. However, in high-
could enhance 3D-connection path- permeability samples, the whole cores
Effect of Scale on Rock Properties ways that can promote flow paths and showed lower permeability values than
For this heterogeneous-carbonate core, yield higher permeability measure- did the plugs. This phenomenon could
the average plug permeabilities were ments. It was observed that whole-core be linked to the degree of heterogeneity
much lower than that determined porosities are at the lower end of plug- and the rock nature.
from whole-core data. Whole-core porosity distribution. Therefore, with • In the low-permeability samples,
measurement correctly averages high- higher-density measurements of plug larger rock volumes can enhance
and low-permeability zones into the porosity in the core, similar poros- 3D-connection pathways that can
overall volume, whereas plug samples ity measurements could be obtained promote flow paths and yield higher
do not capture the 3D connectivity between plugs and whole cores. This permeability measurements. In high-
among the highest and lowest perme- is the result of the high degree of permeability samples, heterogeneity
ability ­appropriately. heterogeneity in carbonate reservoirs. would be more pronounced in small-
In this study, porosity and perme- Nevertheless, whole-core-porosity data er-volume rocks, which could easily
ability data measured on 14 whole-core captured reservoir porosity more cor- overestimate permeability through rela-
samples were compared with measured rectly by averaging the different local tively larger porosity channels.
porosity and permeability data obtained rock properties in the larger whole- • The m values measured on whole
from vertical and horizontal 1.5-in.- core samples. It should be noted that cores were lower than measurements
diameter plugs taken adjacent to the averaging plug data will not yield cor- on corresponding plug samples. These
whole-core samples. Whole-core poros- rect porosity values. Comparing all data would lead to better reservoir
ities were lower than plug porosities. horizontal-plug-permeability data with characterization and greater confidence
The comparison was made from both whole-core horizontal-permeability in volumetric and reserves estimates.
horizontal and vertical plugs, which measurements confirmed the depen- • The more-representative measure-
were cut adjacent to each whole-core dence of the permeability-difference ments of porosity and electrical proper-
sample. The average porosity difference variations on permeability range (i.e., ties in whole cores decrease uncertainty
was 2.50 p.u., with maximum differ- heterogeneity). Moving down in the in petrophysical interpretation and oil-
ence as high as 5.92 p.u. Permeability reservoir, permeability decreased, as in-place calculations. JPT

60 JPT • AUGUST 2011