The Many Facets of

Pulsed Neutron Cased-Hole Logging

Ivanna Albertin ■ The multipurpose

Harold Darling RST service. Car-
bon-oxygen ratio,
Mehrzad Mahdavi inelastic and
Ron Plasek capture spectra,
Sugar Land, Texas, USA sigma, borehole
holdup, porosity,
Italo Cedeño water and oil
velocities, and
City Investing Company Ltd. borehole salinity
Quito, Ecuador are some of the
measurements that
Jim Hemingway can be made with
RST equipment.
Peter Richter
Bakersfield, California, USA

Marvin Markley
Bogota, Colombia

Jean-Rémy Olesen
Beijing, China

Brad Roscoe
Ridgefield, Connecticut, USA

Wenchong Zeng
Shengli Petroleum Administration Bureau
China National Petroleum Corporation Advanced neutron generator design and fast, efficient gamma ray
China detectors combine to make a reservoir saturation tool that is capable

For help in preparation of this article, thanks to Darrel of detailed formation evaluation through casing and more. Lithology
Cannon, Wireline &Testing, Sugar Land, Texas; Efrain
Cruz, GeoQuest, Quito, Ecuador; Steve Garcia, determination, reservoir saturations and flow profiles are some of the
GeoQuest, Bakersfield, California, USA; Michael Herron
and Susan Herron, Schlumberger-Doll Research, Ridge- comprehensive answers provided by this multipurpose tool.
field, Connecticut, USA; Chris Lenn and Colin Whittaker,
Schlumberger Cambridge Research, Cambridge, Eng-
land; and Chris Ovens, GeoQuest, Aberdeen, Scotland. To manage existing fields as effectively and A Multipurpose Service
In this article, CNL (Compensated Neutron Log), CPLT efficiently as possible, reservoir engineers The RST service was introduced in June,
(Combinable Production Logging Tool), ELAN (Elemental
Log Analysis), FloView, FloView Plus, FMI (Fullbore
monitor movement of formation fluids 1992 with a through-tubing pulsed neutron
Formation MicroImager), Phasor (Phasor Induction SFL), within the reservoir as well as production tool capable of providing both carbon-oxy-
RST (Reservoir Saturation Tool), SpectroLith, TDT from individual wells. Pressure measure- gen ratio (C/O) and sigma reservoir satura-
(Thermal Decay Time) and WFL (Water Flow Log) are
marks of Schlumberger. ments play a vital role in reservoir manage- tion measurements.1 Interpretation of either
1. For a detailed description of the RST tool hardware ment. However, these data need to be aug- measurement, under suitable formation and
and the latest scintillation detector technology: mented by other measurements to detect borehole conditions, provides quantitative
Adolph B, Stoller C, Brady J, Flaum C, Melcher C, fluid movement within the producing well oil saturation. The high-yield neutron gener-
Roscoe B, Vittachi A and Schnorr D: “Saturation
Monitoring With the RST Reservoir Saturation Tool,” and the surrounding formation. One ator and high-efficiency dual-detector sys-
Oilfield Review 6, no. 1 (January 1994): 29-39. recently introduced cased-hole logging tool, tem provide higher gamma ray count rates,
Sigma is a measure of the decay rate of thermal neu- the RST Reservoir Saturation Tool, provides and hence better statistics, than previous
trons as they are captured.
abundant single-well data to help reservoir generations of pulsed neutron devices. This
2. Holdup is a measure of the volumetric percentage of
each phase in the borehole. Water holdup plus oil engineers locate bypassed oil and detect has led to the development of many other
holdup plus gas holdup equals unity. Flow rate equals waterflood fronts, fine-tune formation evalu- applications, including spectroscopy mea-
holdup multiplied by area and by velocity. ation and monitor production profiles.

28 Oilfield Review
 Accurate Inaccurate
surements, accurate time-lapse reservoir
monitoring and evaluation in difficult log- Alpha processing Windows
■Accuracy and
ging environments such as variable forma- precision. Alpha
processing combines
tion water resistivity and complex lithology. the accuracy of the
Other features of the tool design allow elemental yields
several auxiliary measurements such as computation of oil
borehole salinity and thermal neutron Precise volume (bottom left)
with the precision of
porosity. The tool comes in two the windows
diameters—the 111/16-in. RST-A tool and approach (top right).
21/2-in. RST-B tool. Both use the same type The result is an oil
of neutron generator, detectors and electron- volume that is both
ics. However, the larger diameter RST-B tool accurate and pre-
cise (top left).
incorporates shielding to focus the near
detector towards the borehole and the far
detector towards the formation, allowing
logging in flowing and unknown borehole
fluids and also providing a borehole holdup Imprecise
measurement.2 More recent applications for
the RST-A tool include WFL Water Flow Log
measurements and separate oil and water
phase velocities in horizontal wells—Phase Yields
Velocity Log (PVL) measurements.
Essentially the RST service provides three 0.5
types of measurements:
• reservoir saturation from C/O or sigma
0.4
measurements
Sw=0%, Yo=100%
• lithology and elemental yields from
Far carbon/oxygen ratio

analysis of inelastic and capture gamma

0.3
ray spectra
• borehole fluid dynamics from holdup,
WFL and PVL measurements. Sw=0%, Yo=0%
0.2
This article summarizes the many facets of
RST logging and reviews several examples.
0.1
Sw=100%, Yo=100%
x
Reservoir Saturation xxx x
x
Reservoir saturation is derived from C/O or xxxxx
xxx
x xx x
0.0 xx
xx
inferred from sigma measurements (see “Sat-
uration Monitoring, South American Style,” Sw=100%, Yo=0%
next page ). Inelastic gamma ray spectra are -0.1
used to determine the relative concentration -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
of carbon and oxygen in the formation. A Near carbon/oxygen ratio
high C/O indicates oil-bearing formations; a
■ Water saturation, Sw, and borehole oil holdup, Yo, crossplot. Far car-
low ratio indicates water-bearing forma- bon-oxygen ratio (FCOR) is more influenced by formation carbon, and
tions. Sigma is derived from the rate of cap- near carbon-oxygen ratio (NCOR) is more influenced by borehole car-
ture of thermal neutrons—mainly by chlo- bon. A crossplot of FCOR versus NCOR (crosses) can, therefore, be used
rine—and is measured using capture to determine water saturation and borehole oil holdup. Overlying the
gamma rays. Saline water has a high value crossplot is a quadrilateral whose end points are determined from an
extensive data base that depends on environmental inputs such as
of sigma, and fresh water and hydrocarbon lithology, casing size and hydrocarbon carbon density. The corners
have low values of sigma. As long as forma- correspond to 0 and 100 % Sw and 0 and 100 % Yo. Interpolation pro-
tion water salinity is high, constant and vides Sw and Yo at each depth.
known, water saturation Sw may then be
calculated. cise than the C/O windows . Conversely, correction for the carbon density of the
Carbon-oxygen—Carbon-oxygen ratio is C/Owindows is often less accurate but has bet- hydrocarbon phase.
measured in two ways. A ratio (C/Oyields ) is ter statistics and so is more precise. Each Carbon-oxygen ratios are generated for
obtained from full spectral analysis of car- ratio is first transformed to give an oil vol- the near and far detectors. These two ratios
bon and oxygen elemental yields. A second ume, and then the two oil volumes are are used to give water saturation and bore-
C/O (C/O windows) is obtained by placing combined using an alpha processing hole oil holdup (above ).
broad windows over the carbon and oxygen method to give a final oil volume with good Sigma—Sigma is a measure of how fast
spectral peak regions of the inelastic spec- accuracy and good precision ( top ). The thermal neutrons are captured, a process
trum. The C/Oyields is the more accurate of transforms of C/O ratio to volume of oil use typically dominated by chlorine. Hence
the two ratios, but lower count rates and, an extensive data base covering multiple formation sigma may be considered a mea-
therefore, poorer statistics make it less pre- combinations of lithology, porosity, hole
size, casing size and weight, as well as a
Summer 1996 29
 Saturation Monitoring, South American Style

Fanny field, situated among the oil fields east of Sw RST<<SwOH

Sw RST<<Sw OH
the Andes mountains, in the Oriente basin,
Water
Ecuador, was discovered in 1972 and is presently Oil Water
operated by City Investing Company Ltd. (below). Bound water Oil
Lith. Caliper Sw from Bound water
Differential compaction of sands and shale Total
inelastic 6 in. 16 the RST Calcite
probably created the structural high that forms RST Sigma Near C/R 100 p.u. 0 Porosity Coal
the field. Primary production is from the M-1 0 c.u. 30 Far C/R Silt
Sand SP from OH Near C/R 100 0 25 0 Quartz
sandstones of the Upper Cretaceous Napo with Clay 120 mV 30 -0.10 -0.15 p.u. p.u. Clay
secondary production from the Lower U sand- Lime GR Far C/R Fluid Analysis Combined Model
stones of the Lower Cretaceous Napo. 10 API 110 0 0.25 50 p.u. 100 0 p.u. 100
Depth,
There are six wells in Fanny field and these are M-1 sand
ft
coupled to three others from the adjoining 18B
field drilled by the national oil company of
Ecuador, PetroProduction. Total output is 4000 7700
BOPD of 22.2° API oil with a fluctuating water cut
of between 37% and 91%. Production is by
hydraulic pump.
Fanny-1 was completed as a commingled pro-
ducer in 1978 and after 18 years it was still pro-
ducing about 150 BOPD with 90% water cut from
two zones in the M-1 sand body. The high water
cut prompted City Investing to investigate.
A 111/16-in. RST-A tool was run with the well shut- 7750
in to record carbon-oxygen ratio, formation
sigma, borehole sigma, thermal neutron porosity
and borehole salinity measurements.

Tumaco

Esmeraldas
Balao
Fanny Lower U sand
Quito
Tiputini
ECU AD O R 8400

Tigre ■ Fanny-1 RST log results. ELAN Elemental Log Analysis interpretation of Sw and lithology (track 3) shows the
original openhole water saturation. Since then the oil-water contact has risen to 7752 ft (track 2) shown by the
RST Sw of nearly 100% through the bottom section of the M-1 sand. The high carbon-oxygen ratio from 7702
to 7709 ft is a coal seam. Very little of M-1 above the oil-water contact is depleted and the Lower U sand also
shows high hydrocarbon saturation.

Formation sigma and thermal neutron porosity Tests on the interval 7710 to 7720 ft [2350 to
improved on the original formation evaluation by 2353 m] confirmed the RST results with a produc-
■ Fanny field location. providing a better estimation of shale volume in tion rate of 900 BOPD at only 10% water cut. The
South America the silty, sometimes radioactive, sandstones, two new zones were also tested and they pro-
and also more accurate lithology identification. duced 1300 BOPD at 4% water cut.
The final interpretation showed that high water The old perforations were cement squeezed
production was caused by a rise in the oil-water and the well, reperforated and recompleted, is
contact to 7752 ft [2363m] (above). It also now producing 1000 BOPD with low water cut—
showed that other sections of the M-1 sand were a sixfold production increase.
still at original water saturation and identified
two virgin oil zones.
30 Oilfield Review
sure of the chlorine content or salinity of Gamma Ray
the formation, and tracks openhole resistiv- Porosity from Core
0 API 300 SO from Core
ity curves. 100 p.u. 0
The raw sigma measurement contains con- 0 p.u. 100
tributions from the borehole as well as the SP SW (11/7/93) Clay
formation. To isolate the formation sigma, -90 mV 120 100 p.u. 0
the neutron generator is pulsed in a dual DCAL Quartz
burst pattern: a short burst followed by a -10 in. 0
long burst. Near-detector measurements are K-Feldspar
DIT-E SO (11/7/93)
strongly influenced by the borehole environ-
0 p.u. 100
ment and hence borehole sigma— espe- Bound Water
cially for the short neutron burst measure- RST SO (11/27/93)
ment. Far-detector measurements are 0 p.u. 100 Irreducible Water
influenced more by formation sigma—espe- RST SO (4/16/94)
cially the long neutron burst measurement. Formation Water
0 p.u. 100
Raw sigma measurements are also affected
RST SO (1/30/96) Phasor Oil Volume
by neutron diffusion and environmental
0 p.u. 100
variables related to the borehole, casing,
cement and formation. At the heart of the Steam/Air 1993
correction process for these effects is a data Depth,
base detailing thousands of combinations of ft Steam/Air 1995
borehole sizes, casing types, formations of
differing porosity and lithology, and bore- X100
hole and formation salinities. Instead of try-
ing to define the response to these variables
by a single set of equations with fixed
parameters, a dynamic parameterization
algorithm uses the data base to compute the
corrected response in real-time, during
acquisition (see “The Sigma Data Base,”
next page ).3
Time-lapse—Once carbon-oxygen mea-
surements or sigma measurements have
been interpreted to produce saturation logs,
these measurements may be repeated later to
monitor reservoir fluid movement such as
oil-water contacts, secondary recovery pro-
cesses or hydrocarbon depletion ( right ).
Good precision is important for time-lapse
(continued on page 34)

■ Time-lapse logging in California. This log

is from a well in the middle of a field that is X200
produced by heating the oil in place with
steam. Steam takes a narrow path from
one wellbore to another and will, therefore,
not flush out all the heavy oil. After some
time, the steam needs to be redirected to
produce bypassed oil. RST time-lapse data
are used to monitor steam location and
changes in oil saturation.
There has been little change in oil satura-
tion of the upper intervals X100 to X190 ft
(track 2). The lower interval, X200 to X270
ft, shows some oil movement. Steam has
been turned off in the zone X195 to X205 ft
which has resaturated with water (track 3).

3. For more on the dynamic parameterization algorithm

approach:
Plasek RE, Adolph RA, Stoller C, Willis DJ and Bordon
EE: “Improved Pulsed Neutron Capture Logging With
Slim Carbon-Oxygen Tools: Methodology,” paper SPE
30598, presented at the 70th SPE Annual Technical
Conference and Exhibition, Dallas, Texas, USA, Octo-
ber 22-25, 1995. X300

Summer 1996 31
 The Sigma Data Base

■The Schlumberger
Environmental Effects
Calibration Facility,
Houston, Texas, USA.
Over 4000 measure-
ments were made in
more than thirty forma-
tions of differing lithol-
ogy and porosity, with
different combinations
of formation salinities,
borehole salinities, and
completions to produce
the sigma data base.

Diffusion, borehole and lithology effects must be ■ EUROPA facility, Aberdeen, Scotland.
considered when transforming raw pulsed neu- umes of the rocks, fluids and tanks used. CNL
tron capture measurements to actual physical Compensated Neutron Log measurements veri-
quantities. These effects are difficult to account fied the porosity values and the homogeneity of
for in direct analytical approaches across the the formations.
entire range of oilfield conditions. Therefore, an Matrix sigma values were determined by gross
extensive data base of laboratory measurements macroscopic cross-section measurements pro-
is used to correct for these effects in real time.1 vided by commercial reactor facilities and by pro-
Over several years, the data base was acquired cessing complete elemental analyses through
for the RST-A, RST-B and TDT-P logging tools at Schlumberger Nuclear Parameter (SNUPAR)
the Schlumberger Environmental Effects Calibra- cross-section tables.2
tion Facility (EECF), Houston, Texas (above and Water salinity was determined by a calibrated
right). This enables raw tool measurements to be titration procedure and then converted into fluid
referenced to calibrated values of formation sigma again using SNUPAR cross-section tables.
sigma, borehole salinity and formation porosity
for a variety of environmental conditions. Each Algorithm—RST Sigma Processing
tool was run in over 30 formations of different eling was used to extend the range of available A three-step sequence is performed to translate
lithologies and porosities. Formation and bore- sandstone formations. To date, the data base con- raw log measurements into borehole salinity,
hole fluid salinities were varied and different tains over 4000 points. porosity, corrected near and far sigma and forma-
completions were introduced into the borehole The sigma values of the database formations tion sigma (next page, top).
representing different casing sizes and cement are calculated classically The first step is to correct the near and far
thicknesses. ∑ = (1-Φ ) ∑ ma + Φ S fl∑ fl detector time-decay spectra for losses in the
Altogether more than 1000 formation-borehole where Φ is the formation porosity, ∑ ma is detection and counting system, and for back-
combinations were measured for each tool. Mod- matrix sigma, Sfl is the formation fluid saturation
and ∑ fl is fluid sigma. 1. Plasek RE et al, reference 3, main text.
Porosity of the EECF tank formations was deter- 2. McKeon DC and Scott HD: “SNUPAR—A Nuclear
Parameter Code for Nuclear Geophysics Applications,”
mined by carefully measuring all weights and vol- Nuclear Physics 2, no. 4 (1988): 215-230.

32 Oilfield Review
 Input ground radiation. Typically the background is
Time decay spectra averaged to improve statistics.
The next step is to generate the apparent quan-
STEP 1
Correction to Spectra tities from the spectra, such as near and far
Counting loss corrections
Background adaptive filtering apparent formation sigmas. These quantities are
Background subtraction not environmentally corrected.
The third step is to apply transforms and envi-
STEP 2
Compute Apparent Quantities ronmental corrections to the apparent tool quanti-
Near apparent borehole sigma SBNA
Far apparent formation sigma SFFA
ties to arrive at borehole salinity, porosity and
Near/far capture count rate ratio TRAT formation sigma. The technique uses dynamic
database parameterization that handles both the
transformation and environmental corrections.
Environmental
Parameters External
Data Base Tool
Borehole size Knowledge
Calibration Accuracy
Casing size/weight (Optional)
Lithology Near/far ratio Porosity A series of benchmark measurements has been
Borehole salinity made to assess the accuracy of the algorithm
used with the data base to compute borehole
salinity, porosity and formation sigma (below).
STEP 3 These benchmark measurements include repro-
Transform from Apparent to
Corrected Quantities cessing the entire data base as well as logging in
industry standard facilities such as the EUROPA
sigma facility in Aberdeen, Scotland (previous
page, top right) and the API porosity test pit,
Outputs
Borehole salinity BSAL SIBF at the University of Houston, in Texas.
Porosity TPHI
Database points were reprocessed with the
Corrected near and far sigma SFNC SFFC
Formation sigma SIGM dynamic parameterization algorithm and the
results were compared with the assigned values.
■Simplified RST sigma processing.

60 35
Limestone
30 Sandstone 250
50 41 p.u.
Dolomite 18 p.u.
Measured sigma, c.u.

Borehole salinity, kppm NaCl

25
Measured sigma, c.u.

200 0 p.u.
40
20
30 150
15

20 100
10

10 5 50
-1.5 0.0 1.5
Deviation from assigned
sigma, c.u.
0
0 0
0 10 20 30 40 50 60 0 5 10 15 20 25 30 35 0 10 20 30 40 50
Assigned sigma, c.u. Assigned sigma, c.u. Sigma, c.u.

■ Processing accuracy. Benchmark measurements were made to assess the accuracy of the algorithm in computing formation and borehole sigma, porosity and bore-
hole salinity. Sigma measured with the RST-A tool versus assigned database sigma (left) shows average errors are small—0.22 c.u. Sigma measured at the EUROPA
facility in Aberdeen (middle) again shows excellent agreement with the assigned values. Comparison of RST-A tool sigma (right) versus borehole salinity shows that
corrected sigma is independent of borehole salinity—vital for time-lapse surveys or log-inject-log operations. In the crossover region (shaded area), formation sigma
approaches or even exceeds borehole sigma. Historically, pulsed neutron capture tools erroneously identify the borehole decay as formation sigma and formation decay
as borehole sigma in this region. However, the RST dynamic parameterization method solves this long-standing problem, correctly distinguishing between formation and
borehole sigma components.

Summer 1996 33
The algorithm does exceptionally well in match- 600 ■ Effect of clay and
30 p.u. calcite on perme-
ing the assigned values. For example, the aver-
ability. A small
age errors for formation sigma were 0.22 capture 500
20 p.u. percentage of clay
units (c.u.) for the RST-A tool and 0.20 c.u. for 10 p.u.
has a dramatic
effect on perme-
the RST-B tool. 400 20 p.u. 15% Calcite ability. Calcite also
reduces perme-

Permeability, md
The EUROPA facility is an independent sigma
ability. So to deter-
calibration facility partially funded by the UK mine a well’s pro-
300
Atomic Energy Authority with major support from ducibility or the
cause of any for-
a consortium of 15 oil companies and govern- mation damage, it
200
ment agencies. The RST-A tool was run in all the is important to
understand the
openhole formations and several cased-hole for-
mineralogy.
mations. A smaller number of measurements 100

were made with the RST-B tool. Both tools read

the true formation sigma over a wide range of 0
lithologies, porosities, formation and borehole 0 0. 2 0.4
fluids, borehole sizes and completions. Even in Dispersed clay, %
the difficult crossover region, where formation
techniques, which by definition look at dif- At high neutron energies, inelastic interac-
sigma approaches or exceeds borehole sigma, ferences from one log to another over a tions dominate. After a few collisions, neu-
the errors are small and the tool does not lock on period of several months. RST data can be tron energy is reduced below the threshold
to the wrong sigma component. gathered at logging speeds nearly three times for inelastic events. The probability of an
Both EUROPA and the University of Houston API those of previous-generation tools for the inelastic interaction occurring is also rea-
same precision.4 sonably constant for all major elements.
pits were used to check porosity readings. The
As neutrons slow to thermal energy levels,
agreement between the two sets of porosities Lithology capture interactions dominate. Some ele-
was excellent. Assessing reservoir deliverability and ments are more likely to capture neutrons
enhancing zone productivity rely on a thor- than others and so contribute more to the
Precision ough understanding of the rock matrix. For capture gamma ray spectrum.
Key to time-lapse monitoring techniques is
example, clay content dramatically affects Inelastic and capture gamma ray spectra
permeability ( above ). 5 Elemental yields are recorded by opening counting windows
repeatability or precision. Time-lapse uses differ-
derived from RST spectroscopy measure- at the appropriate time after a neutron burst
ences in measured quantities to monitor, for ments provide the input to determine clay from the RST neutron generator. Tool design
example, the progress of waterflooding, the and other mineral content and hence allows not only for much higher gamma ray
expansion of gas caps and the depletion of reser- improve understanding of the rock matrix. count rates than previous generation tools,
voirs. The RST tool has been benchmarked to log Elemental yields—Neutrons interact with but also for gain stabilization that enables
the formation in several ways. Inelastic and lower gamma ray energy levels to be
nearly three times faster than previous genera-
capture interactions produce spontaneous recorded for both inelastic and capture
tion tools for the same level of precision.3 release of gamma radiation at energy levels measurements. A major advantage of this is
that depend on the elements involved. Mea- the inclusion of the inelastic gamma ray
3. For examples of repeatability—precision—see:
Plasek et al, reference 3, main text. surement of the gamma ray spectra pro- peaks on the spectrum at 1.37 MeV for
duced by these interactions can then be magnesium and at 1.24 MeV and 1.33 MeV
used to quantify the abundance of elements for iron.6
in the formation. Elemental yields are often A library of standard elemental spectra,
used in various combinations or ratios to aid measured in the laboratory for each type of
complex lithology interpretation, to deter- tool, is used to determine individual ele-
mine shale volume or to augment incom- mental contributions (next page ).
plete openhole data (see “Making Full Use SpectroLith interpretation—SpectroLith
of RST Data in China,” page 36 ). processing is a quantitative mineral-based

4. For more details on time-lapse monitoring see sec- 7. Herron SL and Herron MM: “Quantitative Lithology:
tions on precision and auxiliary measurements: An Application for Open and Cased Hole Spec-
Plasek RE et al, reference 3. troscopy,” Transactions of the SPWLA 37th Annual
5. Herron M: “Estimating the Intrinsic Permeability of Logging Symposium, New Orleans, Louisiana, USA,
Clastic Sediments from Geochemical Data,” Transac- June 16-19, 1996, paper E.
tions of the SPWLA 28th Annual Logging Symposium, 8. See Roscoe B et al, reference 6.
London, England, June 29-July 2, 1987, paper HH.
6. Roscoe B, Grau J, Cao Minh C and Freeman D:
“Non-Conventional Applications of Through-Tubing
Carbon-Oxygen Logging Tools,” Transactions of the
SPWLA 36th Annual Logging Symposium, Paris,
France, June 26-29, 1995, paper QQ.

34 Oilfield Review
 Inelastic Spectra alone with the remainder of the formation
■ Elemental stan- being composed of quartz, feldspar and
dards for the RST-A mica minerals.
tool. Lower gamma SpectroLith interpretation involves three
Oxygen
ray energy levels
are recorded by the steps:
RST tools than by • production of elemental yields from
Silicon previous generation gamma ray spectra
pulsed neutron tools. • transformation of yields into concentra-
Magnesium Iron This allows mea-
Relative counts tion logs
surement of elemen-
Sulfur Calcium • conversion of concentration logs into
tal contributions
Background from elements such fractions of clay, carbonate and frame-
as magnesium and work minerals.
iron. Elemental
Carbon yields are processed
from standard spec- Borehole Fluid
tra obtained using The producing wellbore environment may
laboratory measure- include a combination of oil, water and gas
ments. Shown are phases in the borehole as well as flow
the standards for
inelastic (top) and behind casing. Borehole fluid interpretation
capture (bottom) is primarily based on fluid velocities and
spectra for the borehole holdup. The RST equipment
1 11/16-in. RST-A tool. makes these measurements using several
1 2 3 4 5 6 7 8
Energy, MeV independent methods, with enough redun-
dancy to provide a quality control cross
Capture Spectra
check:
• The WFL Water Flow Log measures water
velocity and water flow rate using the
Iron principle of oxygen activation. This
method detects water flowing inside and
Chlorine
Silicon outside pipe, and in up and down flow.
• The Phase Velocity Log (PVL) measures
oil and water velocities separately by
Relative counts

injecting a marker fluid, which mixes and

travels with the specified phase. This
Titanium method may be applied to up and down
Calcium flow, but only fluids in the pipe are
marked and therefore detected.
Sulfur • Two-phase—oil and water—borehole
holdup may be measured in continuous
Hydrogen logging mode with the RST-B tool.8
Gadolinium • Three-phase—oil, water and gas—bore-
hole holdup is currently an RST-A station
measurement based on a combination of
C/O and inelastic count rate ratio data.
1 2 3 4 5 6 7 8 • Borehole salinity is one of the computa-
Energy, MeV
tions made as part of the sigma and poros-
ity log and may be used to compute a
lithology interpretation derived from elemen- minum and total clay concentration. Of borehole water holdup with either the
tal yields. Traditional lithology interpretation more importance, it also showed that sili- RST-A or the RST-B tool.
relied on measurements of elements such as con, calcium and iron can be used to pro- (continued on page 39)
aluminum and potassium to determine clay duce an accurate estimation of clay without
content. Aluminum, especially, is difficult to knowledge of the aluminum concentration.7
measure and requires a combination of log- The concentrations of these three elements
ging tools; the interpretation is also complex. can be obtained from RST spectroscopy
A recent detailed study of cores showed measurements.
that a linear relationship exists between alu- In addition, carbonate concentrations—
defined as calcite plus dolomite—can be
determined from the calcium concentration

Summer 1996 35
 Making Full Use of RST Data in China

Gu Dao and Sheng Tuo are typical of the Shengli Sonic and gamma ray data do not provide
MONGOLIA
complex of oil fields about 200 km [125 miles] enough lithology information to account for matrix
CHINA
southeast of Beijing near the Bo Hai Gulf, China carbon. For example, carbonates cannot be distin-
(right).1 Both fields have a similar deltaic deposi- guished from tight siliclastic streaks. Sonic-
tional environment, with alternating sand-shale derived porosity may also be inaccurate if lithol-
Beijing
sequences. Thin, tight, calcareous streaks within ogy and formation fluids are unknown, and also, if
Beijing
the depositional sequences are common. Reser- the sands are unconsolidated and the compaction
voir layer thickness varies from more than 10 m Qingdao factor is unknown. The gamma ray curve alone is
Bo Hai Gulf
[31.2 ft] to less than 1 m [3.1 ft] and each layer is unsuitable for accurate shale volume evaluation
produced separately. because the reservoir sands are rich in micas and
Shanghai
For more than 30 years, many of these eastern Shengli Complex feldspars—both radioactive minerals.
Chinese oil fields have been under water injec- Gu Dao To augment the limited openhole data, an RST
Sheng Tuo
tion to maintain pressure and improve sweep of sigma-mode pass provided sigma for shale vol-
the heavy hydrocarbons. The water injection pro- ume estimation and thermal neutron porosity
TAIWAN
gram uses a mix of the low-salinity connate water (TPHI) for effective porosity evaluation. The
and fresh surface water, which has resulted in Hong Kong inelastic-capture data were analyzed in detail not
variable and unknown water resistivity in many only for the carbon-oxygen ratio (C/O), but also for
reservoirs. elemental yields to provide other ratios. For exam-
■Location of Gu Dao and Sheng Tuo fields.
In order to efficiently manage the waterflood ple, the ratio of iron to silicon (IIR) is indicative of
enhanced oil recovery program and maximize oil •Through-tubing logging, while the well was shale volume if kaolinite and heavy minerals are
recovery, it is essential to know the waterflood flowing, avoids formation damage and also not present; the ratio of silicon to silicon-plus-cal-
sweep efficiency, determine residual or remain- increases operational efficiency in a multiwell cium (LIR) may be used as a lithology indicator;
ing oil saturation, and pinpoint zones bypassed campaign. and the ratio of chlorine to hydrogen (SIR) gives a
by the recovery scheme. •The 5 1/2-in. casing inside 8 1/2-in. borehole formation salinity indicator.
Hydrocarbon saturation evaluation from open- completion produces a thick cement sheath The initial volume of oil was computed from the
hole resistivity logs, run in newly drilled infill that reduces measurement sensitivity. The RST openhole resistivity data in 1994 assuming that all
wells, is difficult because the formation water tool has a high-energy, high-yield neutron gen- sands were at connate water resistivity. The 1995
resistivity is variable and most of the time erator and an efficient detection system that RST carbon-oxygen evaluation computed remain-
unknown. Reservoir saturation monitoring with provide better statistics in thick cement than ing oil. A decrease in oil between the two may be
sigma measurements is impractical, as there is the previous-generation pulsed neutron tools. due to reservoir depletion, but could also be due
little contrast between the oil and water sigmas • An additional pass in sigma mode provides to an overly optimistic openhole evaluation if the
and, in any case, the water sigma is unknown. data useful to accurately evaluate shaliness, reservoir water was not at connate salinity, but at
These constraints leave carbon-oxygen measure- especially in wells with scarce openhole data. the fresher floodwater salinity.
ments as the only viable option. • Measurements such as neutron porosity and The additional RST data proved invaluable. For
The Shengli oilfield operators—Shengli count rates can also be recorded to aid inter- example, in the Gu Dao and Sheng Tuo fields in
Petroleum Administration Bureau, China National pretation when gas is present. general, sigma responds primarily to changes in
Petroleum Corporation (SPAB-CNPC)—decided to matrix sigma and therefore provides the best shale
run the 21/2-in. RST-B tool for many reasons: Evaluation with Scarce Openhole Data indicator. The lithology indicator ratio LIR was
•The shielded dual-detector system alleviates Key to the interpretation of carbon-oxygen data is used to identify the tight calcite streaks at X201 m
the effect of a changing or unknown borehole a knowledge of lithology to account for matrix and X218 m.
oil holdup, as well as the effect of waxy carbon, and effective porosity to calculate oil sat- Interpretation of the salinity indicator ratio (SIR)
deposits on the casing. uration. A typical Sheng Tuo well illustrates the is more complicated. However, when the forma-
benefits of additional data provided by the RST tion water volume remains constant, SIR responds
1. Olesen J-R, Chen Y, Zeng W, Zhu L and Zhang Z:
“Remaining Oil Saturation Evaluation in Water Flooded tool (next page). For this well the openhole data directly to formation fluid salinity and can be used
Fields Under Variable Formation Water Resistivity,” to be were limited to sonic and gamma ray logs. to determine the progress of injection water—
presented at the 1996 International Symposium on Well
Logging Techniques for Oilfield Development, Beijing, approximately the case in the large reservoir
Peoples Republic of China, September 17-21, 1996. between X220 m and X245 m.

36 Oilfield Review
 IIR LIR DT Openhole Analysis
0 2.5 0.625 1.25 150 µsec/ft 50 0 p.u. 100
SIGM SIR TPHI Openhole Fluid 1994
0 c.u. 50 -0.5 ppk 3.5 60 p.u. 0 100 p.u. 0
GR NPHI Shale
100 API 250 60 p.u. 0 Bound Water
• The inelastic count rate ratio (CRRA) from the
Quartz
near and far detector is sensitive to porosity
Calcite
and gas content.
Depth, RST Oil 1995
For example, in one Gu Dao well, the upper
m Water
sand body, X103 m to X109 m, shows the pres-
ence of gas (next page, top). Sigma and CRRA
X200
scales were chosen so that the curves overlay in
clean gas-free formations. In the upper sand they
show negative separation as both sigma and
CRRA are driven lower by the presence of gas.
Similarly, TPHI shows a reduced neutron porosity
when compared to the true formation porosity
taken from the openhole interpretation of 1990.
No gas was apparent on the 1990 openhole
logs, so it is assumed that reservoir pressure has
declined below bubblepoint allowing gas to come
out of solution. Tests indicate that this is a water-
bearing zone with some gas, confirming the RST
X250 interpretation.

Determining Water Resistivity and Flood Index

Interpreting openhole logs of newly drilled wells
in reservoirs that have been partially or fully
■ Formation evaluation with additional RST data. Volumetric analysis (track 4) shows remaining hydrocarbon
saturation determined from RST carbon/oxygen ratio. The 1994 openhole fluid curve indicates more oil due to flooded is challenging. Water resistivity, Rw ,
either depletion or an overly optimistic evaluation. A comparison of RST porosity (TPHI), cased hole CNL often varies continuously from the relatively high
Compensated Neutron Log porosity (NPHI), and sonic transit time (DT), shows good agreement (track 3),
value of fresh floodwater to the low value of the
especially when NPHI is put on a sandstone scale—3 to 4 p.u. shift to the left. The lithology indicator (LIR) is
about 1 for siliclastics and decreases for carbonates (track 2). Two tight calcite streaks can be seen at X201 more saline connate water. If connate water
and X218 m. The salinity indicator (SIR) responds to formation salinity if porosity and hydrocarbon saturation resistivity is used for Rw , then hydrocarbon satu-
are approximately constant (track 2). The iron indicator (IIR), gamma ray and sigma (track 1) follow the same
trend, and each may be used for shale volume calculation under the correct conditions. Gamma ray indication ration will be optimistic in partially flooded
of shale will be pessimistic if radioactive sands are present—for example, those containing micas and zones.
feldspars. Clays, except for kaolinite, contain iron. Sigma responds to formation matrix and fluids. Sigma fluid
is almost the same when oil and fresh water are present, so sigma responds primarily to changes in matrix. However, by combining openhole and RST data
In Gu Dao and Sheng Tuo, sigma has proved to be the best shale indicator. a continuously varying Rw may be calculated
leading to true hydrocarbon saturation. The eval-
In the shaly lower section of the reservoir, Identifying Gas-Bearing Zones uation may be taken further if floodwater resistiv-
salinity is high and probably at connate level, Carbon/oxygen ratio responds to the carbon con- ity is known and constant. In this case, the total
indicating minimal depletion. The middle section centration in pore space. In gas-bearing zones, volume of water may then be split into connate
is the cleanest, most permeable section and carbon concentration is low, so C/O is low. Low and floodwater.
shows a progressive drop in salinity. The water- C/O can easily be misinterpreted as a water-bear- Reservoir saturation acquisition timing is criti-
flood front has reached this section. The upper ing zone. However, several auxiliary measure- cal to the interpretation. It must be late enough
section shows an intermediate salinity and shale ments can help identify gas-bearing intervals: after well completion to allow drilling fluids to
content, and also a smaller discrepancy between • Gas sigma is much lower than water sigma or dissipate, but before significant hydrocarbon
RST saturation and openhole saturation. Flooding oil sigma; therefore, at comparable shale lev- depletion occurs. Four weeks has proven ade-
has reached this section, but is not complete. els, the RST sigma measurement will be lower quate for Gu Dao and Sheng Tuo fields.
Similar results have been seen with other RST in gas-bearing reservoirs. Water resistivity is computed using standard
logs in these fields. • Hydrogen index is also low in gas-bearing openhole interpretation methods. Openhole logs
zones. Therefore, neutron porosity measure- provide Rt , Rclay, Vclay and effective porosity,
ments such as RST porosity (TPHI) underesti- Φeff. Water saturation comes from RST interpre-
mate formation porosity.
Summer 1996 37
 Radius of Bit Openhole Sw 1990 Openhole Porosity Openhole Analysis tation. The flood index is determined as a linear
0 in. 10 100 p.u. 0 50 p.u. 0 0 p.u. 100 interpolation between floodwater resistivity and
Borehole Fluid Cased Hole Sw 1995 O.H. Fluid Volume 1990 Shale connate water resistivity.
Casing Wall 100 p.u. 0 50 p.u. 0 Bound Water In a Gu Dao field example, connate and floodwa-
Assumed Cement RST Gas Indicator RST Fluid Volume 1995 Quartz ter salinities are 8.5 ppk and 3 ppk, respectively
Sheath 5.75 1.75 50 p.u. 0 Calcite (below left). The lower section, X296 to X303 m,
Formation SIGM TPHI from Sigma mode RST Oil 1995 is shaly and water-bearing. The middle section,
Depth, -10.0 c.u. 30.0 0.5 p.u. 0 Water X287 m to X296 m, is the cleanest and is separated
m Gas RST Oil 1995 from the lower section by a thin, clean, sand streak
where the oil-water contact is situated.
X100
The clean midsection has the highest permeabil-
ity and provides a preferential conduit for water-
flooding. The discrepancy between RST-derived
and openhole hydrocarbon saturation is due to the
inadequate Rw estimation for the openhole evalua-
tion. True hydrocarbon saturation is 40% as shown
by RST data and not 60%. Water resistivity, com-
puted from a synthesis of RST and openhole data,
indicates that fresh waterflooding has increased
Rw from the connate water value of 0.35 ohm-m to
X125
about 1 ohm-m. The flood-index calculation con-
firms that the cleanest levels of this reservoir have
been heavily flooded.
The shalier upper sand section shows general
agreement between RST-derived and openhole
hydrocarbon saturation. Because of the increase in
■ Gas detection. Inelastic count rate ratios of near-to-far detector counts and sigma are both affected by gas shaliness and the related decrease in permeability,
(track 2). Negative separation of these curves indicates gas. RST porosity, TPHI, also reads lower in gas (track
3). Although no gas was shown on the openhole logs, it is assumed that solution gas has accumulated in the waterflooding essentially bypasses this section
fully depleted zone between X100 m to X109 m. Tests indicate that the layer is mainly water and gas. and little hydrocarbon sweep is achieved.

Radius of Bit RST-derived Rw Openhole Porosity Openhole Analysis Campaign Success

0 10 0 2 50 p.u. 0 0 p.u. 100 The Shengli oilfield RST campaign has shown that
Borehole Fluid Cased Hole RST Sw O.H. Fluid Volume 1994 Shale hydrocarbon monitoring in waterflooded fields with
Casing Wall 100 p.u. 0 50 p.u. 0 Bound water varying salinity is a viable procedure. In addition,
Assumed Flood Index RST Fluid Volume 1995 Quartz ancillary RST measurements complement open-
Cement Sheath 2 0 50 p.u. 0 Nonmovable oil hole information, improving both formation evalua-
Formation Nonmovable Oil Open Hole 1995
tion and detection of gas-bearing intervals. Also,
Remaining Oil RST1995 Movable RST Oil 1995
Depth, the combination of openhole and RST data
m Flood Water Water
acquired within one month is a powerful tool for
evaluating the waterflooding process. During the
course of the campaign, RST data contributed to
the achievement of the SPAB-CNPC engineers’ goal
of maintaining oil output while controlling water
production. RST results showed a large amount of
X290
remaining hydrocarbon, especially in the massive
sands of the Sheng Tuo oil field.

■Water resisitivity, Rw, and flood index. A flood

index can be calculated from variable Rw (track 2)
X300 computed from RST and openhole data collected
before any hydrocarbon depletion and after invasion
fluids have dissipated (track 3).

38 Oilfield Review
 WFL measurements—Water flow logging, Near Detector Far Detector Additional Detector
introduced with the last-generation TDT
Thermal Decay Time service several years
ago, is now available with the RST service.
The RST neutron generator provides
improved burst control, which allows detec-
tion of water velocities up to 500 ft/min
[150 m/min] with the far detector alone. In
Casing
addition, the introduction of energy discrimi- Minitron Oil
nation and shielding between neutron gener-
ator and detectors results in a significant
improvement in the signal-to-noise ratio, and
Water
extends sensitivity to low flow conditions.
Oxygen molecules in water are activated
by a burst of neutrons producing a radioac-
tive cloud. The cloud moves with the water
along the borehole, emitting gamma rays as 16
O+n p+16N β+16O* 16
O+γ Half-life ~7.1sec
activated oxygen decays back to its steady
state (top right ). As the cloud passes, gamma ■ WFL Water Flow Log service. A short burst of neutrons interacts
rays are first detected by the near detector with oxygen in the surrounding water forming an oxygen isotope
and then by the far detector of the RST with a half-life of 7.1 sec. As the activated oxygen decays back to
sonde, producing a characteristic peak in its steady state, gamma rays are emitted. In flowing water the
cloud of activated oxygen, and hence gamma rays, travels along
the count rate of each. The time between at the water velocity. Characteristic increases in count rate are
neutron burst and cloud detection—time-of- seen as the cloud passes the various detectors. The distance
flight—and the distance between neutron between neutron generator and detector and the time-of-flight
generator and detector give water velocity. give water velocity. The initial cloud volume is proportional to the
amount of oxygen present and hence volume of water. The area
Other detectors can be added farther away
under the gamma ray peak as the cloud passes a detector is,
in the tool string to detect extremely high therefore, also proportional to the volume of water flowing by
water velocities. The RST equipment can (water holdup)—allowing for effects of diffusion and decay rate.
also be turned upside-down to detect down- Combining water velocity and holdup gives water flow rate.
ward flow.
In addition, the volume of activated oxy- ■ Phase Velocity
Marker signal Logging (PVL).
gen is proportional to the volume of water A strong neutron
flowing by the detectors. The profile of the absorber is
detected signal carries information about injected into the
the mean water velocity, water holdup and appropriate phase
of producing fluid.
water flow rate. These quantities are related
This is subse-
in that the water velocity, water holdup and Start of injection quently detected,
effective cross-sectional area of the pipe can allowing a time-of-
be combined to compute the water flow 0 10 20 30 40 50 60 70 80 90 flight measure-
rate (see “Production Logging in the San Time, sec ment that gives
the velocity of that
Joaquin Basin,” next page ). phase.
PVL —Phase velocity logging has been Oil-miscible marker RST tool Oil
developed for horizontal wells where strati-
fied flow is present. Like WFL logging, the
Phase Velocity Log measures time-of-flight.
Gadolinium has a very high thermal neutron
capture cross section and is injected into the
producing borehole ( bottom right ). The
injection fluid is designed to mix with either
the water or oil phase only. Gadolinium acts Phase Velocity Sonde Water
as a sink, sucking in thermal neutrons and

Summer 1996 39
 Production Logging in the San Joaquin Basin

Elk Hills is one of the largest oil fields in the San Gas
Joaquin basin about 20 miles [32 km] west of Bak- Oil
ersfield, California, USA (below). The field forms Water
part of the Naval Petroleum Reserve No. 1 and is
Downhole Flow Rate, B/D
operated by Bechtel Petroleum Operations, Inc.
for the Department of Energy. Although Elk Hills 0 3000

was discovered in 1911, production was limited Pressure Temp Water Flow Stations
until the 1974 oil crisis resulted in opening up the Depth, Water Flow Log, B/D
field to full production in 1976. The field has pro- ft 1050 psi 1300 206 °F 211 0 3000

duced over 1.1 billion barrels of oil and a signifi-

cant quantity of gas, and now produces about
60,000 BOPD of medium-gravity crude.
Earlier this year, Bechtel wanted to determine X200
the flow profile and quantify the zonal contribu-
tions to oil, water and gas production from a well
in which production from a waterflooded sand
reservoir was commingled with production from a
shaly interval. A production log consisting of tem-
perature, pressure and spinner was run and sta-
tionary WFL Water Flow Log measurements were
X400 Thief zone
taken with the RST tool.
The flow profile turned out to be complex,
showing a zone of water recirculation near the
bottom and a thief zone above (right).1
A combination of spinner and WFL data located
the recirculation zone. The spinner indicated down
flow, while the WFL data indicated a small
X600
amount of water flowing up. The temperature log
Recirculating water zone
also showed a strong anomaly over this interval.
The flow profile shows a net flow of oil from this
zone simply because a recirculation zone requires
multiphase flow.
Both spinner and WFL data show an increase in
flow above the recirculation zone before an abrupt
X800
Fresno
San Andreas Fault ■WFL Water Flow Log. The flow profile indicates that most of the gas production is from X350 to X370 ft
(tracks 2 and 3). Below this depth is a complex profile of thief zone and water recirculation. WFL stationary read-
ings determined the water production profile, and temperature and pressure (track 1) aided the interpretation.
Coalinga
decrease at X430 ft. The temperature also drops The next significant event in the flow profile
C A L I F O R N I A at this point. The combination of decrease in flow occurs across the short perforated interval X350 to
rate and temperature can occur only if the forma- X370 ft. Here, a large increase in spinner flow rate
Bakersfield tion is taking fluid—a thief zone. Conventional and a change in slope of the pressure data indicate
Elk hills openhole logs and the mud log suggest that there an influx of gas. The WFL log shows doubling of the
Taft is a highly resistive, low porosity carbonate in water flow rate across the same interval.
this interval. The FMI Fullbore Formation
1. Water recirculation occurs, usually in deviated wells,
U S A MicroImager tool shows what has been inter- when water and oil are present. Water can flow up with
preted as a calcite healed fracture. This fracture the oil on the upper side of the well and down on the
■ Location of
lower side in a continuous cycle.
Elk Hills field, has most likely been opened by acid treatment
Kern County, A thief zone occurs when a perforated zone has a lower
California. and has created the thief zone. formation pressure than the borehole, causing flow
from borehole to formation.
40 Oilfield Review
 CPLT Combinable
Production Logging Tool RST Reservoir Saturation Tool FloView tool Fluid marker Total flow rate
Pressure and temperature Oil holdup Flow regime injector
Gas indicator Water holdup
Gamma ray
detector

CPLT GR RST FloView Plus tool

Spinner

WFL Water Flow Log Phase Velocity Log

Water velocity Marker injection for oil
Water holdup and/or water velocity
Water flow rate index

■ The next generation production logging tool string.

changing the borehole sigma. The detection The exact position of these points depends Comprehensive Cased-Hole Evaluation
of this change provides a time-of-flight mea- on lithology, porosity, hydrocarbon carbon Since commercialization of the RST service
surement for the marked phase. density, hole size, casing size, casing weight four years ago, many applications have
Two-phase borehole holdup —The two and sonde type—RST-A or RST-B sonde. been developed. With the addition of lithol-
detectors of the RST sonde provide two car- With the larger RST-B sonde, the quadrilat- ogy interpretation, phase velocity logging
bon-oxygen measurements that are suffi- eral is wide since the far detector is shielded and three-phase holdup, the tool is rapidly
cient to solve for formation water saturation to be more sensitive to the formation and becoming a comprehensive cased-hole
( S W ) and borehole oil holdup ( Y O ) (see the near detector shielded to be more sensi- evaluation service. 10 A future Oilfield
crossplot, page 29 ). Four points may be tive to the borehole. This provides good sep- Review article will explain in more detail
defined on a plot of far carbon-oxygen ratio aration of the signals and a good borehole some of these new services, including new
versus near carbon-oxygen ratio to give a oil holdup measurement in addition to a for- production logging combinations (above ).
quadrilateral: mation saturation measurement. The slim- —AM
• Water in the formation and water in mer RST-A sonde is not focused and, there-
the borehole (SW = 100, YO = 0 ) fore, requires knowledge of the borehole 9. For an alternative method of measuring borehole
• Oil in the formation and water in the fluids to separate the formation and bore- holdup with the RST-A tool: Roscoe B et al, refer-
borehole (SW = 0, YO = 0) hole signals.9 ence 6.
• Water in the formation and oil in Three-phase holdup—A combination of 10. Schnorr DR: “Determining Oil, Water and Gas
Saturations Simultaneously Through Casing by Com-
the borehole (SW = 100, YO = 100) RST measurements can be used to compute bining C/O and Sigma Measurements,” paper SPE
• Oil in the formation and oil in the three-phase holdup. Gas holdup is indicated 35682, presented at the SPE Western Regional Meet-
ing, Anchorage, Alaska, USA, May 22-24, 1996.
borehole (SW = 0, YO = 100). by the inelastic near-to-far count rate ratio.
The near and far C/Oyields depend on gas,
water and oil holdups. By combining these
measurements and applying two condi-
tions—the sum of the holdups must equal
unity and also the sum of the saturations
must equal unity—three-phase holdups may
be calculated. The RST measurement of
borehole sigma can also be combined with
this analysis to enhance the holdup calcula-
tion if the water salinity is known.

Summer 1996 41