Intraducing The RMT Elite"

Reservoir

Monitor Tool

The industryts most accurate and

r ep e

atable slim-h ol e

th

r ough -tubing

car b on / oxy gen I oggi n g sy stem

Halliburton's RMT Elite is a slim-hole pulsed neutron
iogging system for monitoring and managing the
production of hydrocarbon reserves, This unique throughtubing carbon / axygen (C/O) system has two to three
times higher measurement resolufion than other systems.
Its high-density Bismuth Germanium Oxide (BGO)
detectors lei the RMT Elite achieve resoiutions previously
available only with larger diameter C /O systems. The
RMT Elite can even be conveyed inio a well with fubing
completions unlike larger diameter C/O systems that can
only log through casing.

458

Energy (MeVl
RMT Elite lnelastic Spectra - highest spectral peak resolution of any
th rough+ubi ng QO system.

Increase productiory Save on Cost

Because the RMT Elite can accurateiy evaluate the timelapse performance of hydrocarbon producing reservoirs
without pulling tubing from the well, operators canr

.
r
.
.

Increase production more cost effectively

Monitor changing conditions and fluid movements

Tap into bypassed hydrocarbon reserves
Optimize, manage and produce reservoirs more

efficiently
Increase production to take advantage of increasing

.
.

oiI prices
Avoid production problems through enhanced diagnostics
Make faster decisions on workovers and completions

The RMT Elite also eliminates:

.
.
.

The cost of ki-tling the well

The cost of pulling tubing oui of the well
Operational cost and iost production revenue from

Potential production losses due to formation damage

additional workovers
from well kill fluids
The cost of recompleting the well by re-running tubing

Faster Logging speeds, More Accurate Results

Halliburton's RMT Elite is so accurate and precise it allows
operators to achieve iogging speeds two to five times faster
than any competing system. This blazing combination of
speed and precision allows the RMT ELte to:

.
'
.
r
.
r
o

Accurately determine oil and gas saturations in high

salinity or fresh water formations
Identify bypassed reserves
Pinpoint formation fluid contacts
Identify lithologies and mineralogies
Provi.de porosity information within the completion
interval
Evaluate gravel-packs and lithology via silicon aclivafion
Detect water flow inside or outside the pipe

Advanced Modular Design

Halliburton's RMT Eiite modular hardware desLgn provides a highly versatile system that has multiple operafing
modes and capabilities, allowing operators to make
simultaneous C/O, Sigma and water flow measurements,
Because the system is modul.ar, it can be combined with a
complete string of production logging tool sensors for
detailed production analysis.

Quality Logging For Optimum Reservoir Production

RMT Elite Primary Log Presentation - tack 1 of the display is used for
plotting basic correlation curves, ln this example the simultaneously
recorded formation Sigma (SOSM) and the potassium yield curve (YK)
are plotted. Also plotted in the track is the Oxygen activation curve
(OAl), which is used to detect water flow, Track 2 of the log is used to
display the raw Carbon to Oxygen ratio (COIR) and the Calcium to
Silican ratio (LlRl). The green shading between the curves is a quick look
representation of hydrocarbons, Track 3 of the log displays yield curves
computed from the capture spectra for Silicon (YSi), Calcium (YCa) and
Hydrogen (YH). Track 4 displays inelastic and capture near to far
detectar ratio curves. These cLtrves are used to identifi gas in the
formation (shaded in red).

RMT Elite Quality Log Presentation -Track I of the presentation are

curves that represent the accuracy of spectal gain stabilization measured
from ratios of the iran edge (FERC) and the hydrogen peak (HPLI),
Track 2 is a plot of the COIR and LlRl from the near space detector,
Track 3 is used to plot additional yield curves computed from the capture
spectra, Plotted on this example are the lron yield (YFe) and the
Chlorine yield (YCI). Tracks 4 and 5 are used to plot the total inelastic
and capture count rdtes br the near and far detectorc, T rack 6 is used to
ptot the simultaneous measured near formation sigma (SGFN) and the
far lormation signa (SOFF).

KernSat lnterpretation Example -This well

Iocated in Kern County, California in the Kern
River Field, is in an active steamJload
hydrocarbon recovery project, The log displayed
to the Ieft is an example of our customized
i nterp retati on m ode I K er n Sat,
Track 4 of the example displays the computed oil
saturation (shaded in green) and the gas
saturation (shaded in red). These saturatians
were computed using a combination of Carbon
Oxygen ratio and formation Signta,
Track 3 dkplays the Carbon Oxygen and the
Calcium Silicon ratio curves. The green shading
between the tyvo curves indicates hydrocarbons
in the forntation, Also displayed in the track are
the natural gamma ray measurement and the
s i m u I ta n eous reco rd ed fa rmatio n s igma.
Tracks I and 2 display a comparison of the open
hole density and neutron porosities and the
porosiry ratio indicators measured by the RMT
Elite. Track 1 is the open ho!e density neutron
porosity. Steant measured in the formation at the
time of the log is indicated by the gray shading
between the curves. Track 2 displays the inelastic
and capture ratios measured from the RMT Elite.
The red shading indicates the current location o(
steam in the reservoir This example indicates
that the steam chest has changed when
compared to the original lormation contacts,
The Depth track recorded at the far left side of
the log displays water flow nteasured by the RMT
Elite outside the casing,

Dimensions

And Ratings

Max OD:
Max Press:
Max Temp:

2.125 in.
1

Min Csg/Tbg lD:

Max Csg/fbg lD:
Weight:
Length:

Hardware Characteristics
Source Type:
Sensor Type:

Firing

Rate

5K psi

1.687

325"F
2.388 in.
9.625 in.
122 lb with Camma Ray and Telemetry
27.2 ft with Gamma Ray and Telemetry

lnstrument
Section

4-MeV Neutron Cenerator

2 BCO Scintillators

(QO):

One 30 ps burst every 100 ps; One 5 ms

background pause burst every 25 ms
One B0 ps burst every 1250 ps; One 5 ms
background pause burst every 25 ms
4 or 10 samples per ft
MuxB/82

FiringrRate (Sigma):

Sample Rate:
Combinability:
Measurement
Principle:

2.125 inch

Detector/
Generator

Neutron induced Camma Ray Spectroscopy

lnduced capture Camma Die-away

Vertical Resolution:
Depth of I nvesti gation :
Precision (C/O) ratio:
Precision (Ca/Si) ratio:
Precision (SCFF):
Primary Curves (C/O):

Section

(90%) 30 in.

(50%) 6 in. inelastic;12 in. capture

1.5% (l 5D) at 5 ftlmin, (QO) mode
1.5% (1 5D) at 5 ftlmin, (QO) mode
2% (1 SD) at 2O fUmin, Srgma mode
C/O ratio, CalSl ratio, Near/Far Capture

CR,

Near/Far lnelastic CR, lnelastic/Capture

ratio,Si yield, Ca yield, H yield, Cl yield,

yield, Fe yield
Pri

mary Curves (Sigma) :

Secondary Curves (C/O):

Seconda ry Cu rves

gi gma)

Bull-Nose

SCFF, SCBN, Near/Far Capture CR, Near/Far

Assembly

lnelastic CR, lnelastic/Capture ratio, Near CR,

Far CR, lnelastic CR
S yield, Ti yield, H peak ratio, Fe edge ratio,
C/O ratio uncertainty, Capture CR, lnelastic
CR, O activation CR, SCFF
SCFN, SCBF, Decay Curve fit error,
O activation CR, Near/Far Amplituoe ratio,
Near Amplitude

For more information about

Halliburton's revoiutionary new RMT

EliterM Reservoir Monitor Tool and its
unique combination of cost savings

and produclion benefits, contact your

local Halliburton representative-your
Solution Connection"-.

Calibration
Primary:

HES calibration pits, Houston

Secondary:
Max. Logging Speed:

Horizontal water tank

5 fumin (C/O mode), 30 ftlmin (Sigma mode)

www.halliburton.com

HALLIBURTON
Sales of Halliburton products and services

will be in accord solely with the terms and conditions

contained in the contract between Halliburton and the customer that is applicable to the sale.

H02621 09/oo

@ 2000

Halliburton Energy Services,

lnc. All Rights

Reserved

Printed in U.S.A

Society of Petroleum Engineers

sPE 53736
A New small-Diameter, High-Performance Reservoir Monitoring Tool

G. A. Simpson, L. A. Jacobson, /Halliburton Energy Services and Rodolfo Escobar/ HalliSurton de Venezuela

Copyright 1999. Society of Petroleum EngineeE Inc.

Thrs paper was prepared for presentalion at the 1999 SPE Latin Arcrin and Ca.ibbean
Pelroleum Engineering Conference held in Caras, Venezuela,21-23 April 1999.
Thrs paper was selected for presentation by an SPE Program Comrittee following .eview of
rnformation contained in an abslract submitted by the author(s). Contnls ot the papgr. as
presenled. have not been reviewed by the Sooety of Petroleum Enginee.s and are subject to

correction by the auihor(s). The material as presented, d@s nol re@ssarily reffect any
position of ihe Socrety of Petroleum Engineers_ ;ts officeE, or membe6. Papers presented al
SPE meetings ar subtecl 10 publication revrew by Editorial Committees of the Society of
Petroleum Engineers Eteckonic reprgduction. distribution, or storage of any part of this paper
for commercial purposes wilhout the written @nsent of the Socisly of pekoleum Engine6 is
prohibiled. Permission io reprodue in print is restricted to an abstEct of not more than 3OO
words; illustralions may not be copied. The abstract must @nlain @nspt@ous
acknowledgment of where and by whom the paper was prcsented. w'ite Librarian, SpE. p.O
Box 833836, Richardson. TX 75083-3836. U.S.A.. fax 01-972-952435.

Abstract
This paper drscusses a ncr.,, srnall-diameter, high pelfonn:nce
pulsed-neutron spectrometry tool that has been introduced

lecently for reservoir rnonitoring applications.

Basic

nleasurelnent principles are explained, and irnportant

opelating parantetels are presented. Field examples from
Western Venezr-rela and South Texas illustrate rnultiple
applications of the ncw clcvice and compare the performance
pulsed-neutron
spectrorretry to()l
Tlrc ne r.r, 2- l.8-rnch tool is designed for logging through 27/8-inch or largcr tubing. The tool is slightly larger than
traditional l- I l/ l6-inch thlough-tubing tools. This design

of the ne\\' instrurlcnt u'ith a larger

accornmodates

largel detectors, thereby improving

the

perfolrnance ol the carbon-oxygen (C/O) measurement. Two

bisu-ruth germarlatc detecrors yield high gamma ray count rares
u rth good s|cclr3] rcsolrrtron.

To optiniizc C,/O mcasurelnents, one mode of

opelation intcliaces 5nrs

of

tool
background measurements u.ith

20rns of l0-kHz neutron pulses. The tool records 256-channel

spectra fiorn each detector':

o
r
.

during the nellffon bursts (for analyzing inelastic

events ).

during the interpulse period (for analyzing capture

cvcnts
and

).

duling the background period (for anallzing acrivarion

and background events).

The tool also plovides

simultaneous neutron capture

cross section (sigma) uteasurement. Carbon-oxygen

and

calcium-silicon (Ca/Si) response characteristics for saturation

and lithology analysis are derived from

laboratory
measurements and are illustrated in the customary fan-chart
format. In a second mode of operation, the tool optimizes rhe
sigma measurement and provides high-quality neutron capfllre
spectra for quantitative lithology determination.

lntroduction
The evaluation of forrnation fluid saturations

in low

(or

unknown) salinity environments has always been problematic

in both openhole and cased-hole logging. The absence olsalt
in the water phase eliminates the contrast betx,s3n u,ater ::nr1
oil that is essential for either resistivity or caprure-crosssection tools to distinguish between rhe two fluids. In caseclhole applications, the carbon-oxygen measuremenr uslng
pulsed neutron spectrometry directly senses the presence of
carbon in the hydrocarbon. In circumstances of moderete ro
high porosity, this is a viable technique for quantitative
saruration evaluation, as has been demonstrated by a largediameter tool r. Recently, srnall-diarncter tools 2.3t capablJot'
logging through production tubing have becorne available. but
rvith some sacrifice of accuracy and logging speed cornpared

to

larger-diarneter tools. Because the C/O rneasuremenr ltas

inherent physical lirnitations, tool designers have struggled tt'r
rnaintain pelfbrmance while still addressing the operational
requirement of perfonning this rneasurernent through rubing.

This paper will describe one such eflbrt which has led to

thr.cugh-tubing tool capable of high perfonnance er.en thouqlr

it is of small diamerer.

The new small diarneter C/O logging tool, developed br

Halliburton and named the Reservoir Monitor Tool (RMTr\r).
is described in detail in the fbllowing section. Laborarorr
measurements quantiry the response of the tool in uellcharacterized environrnents through the use of the customarv
fan charts. It provides all the information generated by largerdiameter pulsed-neutron spectrometry tools. In addition. the
near-spaced detector of the new tool provides a C O
measurement dominated by the borehole fluid $,hen run
centralized in 5-inch or larger casing. Under these conditions.
the C/O measurement indicares rhe arnount of oil holdup.
When a gamma ray tool is added below the new tool, a silicon
activation rneasurement is provided sirnultaneously that is
used to detennine lithology and gravel-pack quality. Thus. this
single toolstring can
o provide formation fluid evaluation in moderate-rohigh porosiry for any f)uid salinity

G. A. Simpson, L. A. Jacobson, and Rodolfo Escobar

.
.
.
.
r

saturation from formation stgma

measure formation porosiry

derennine lithologic type

evaluate production fluid located in the casing
evaluategravel-packs

will tbcus, however, on the inelastic C/O

while these other aspects will be deferred to future

This paper
response

publications. Finally, several field logs

will

be shown; one

of

which will show a comparison to a high performance larger

diameter tool.

Tool Description

The RMT (shown schematicallf in Fig. l) is a dual detector,

2-ll8-inch (54 mm) diameter. pulsed neutron spectrometry
tool. The choice of a 2-llS-inch diameter tool size allows
larger diameter detectors to be used, which leads to higher
count rates and bettel spectral resolution as compared to a lI l/16-inch (43 mm) sized tool. This diameter allows passage
through all common tubing sizes except for 2-3l8-inch (60
mm), and the latter is not very corlmon in the fields where
C/O logging is often performed. The detectors are bismuth
gennanate (BGO) scintillators of 1.4 inch (35 mrn) diarneter.
The near-spaceri cietecror is I inch (25 mm) long, while the
far-spaced detector is 6 inches (152 rmn) long. They are
spaced (to crystal centerl ll.5 inches (29.2 cn) and 20.5
inches (52.1 cm) respectively frorn the soulce. The BGO
scintillators require protection from downhole telnperature, so
they are contained irr a Deu,ar rvhich maintains the detector
temperature in an acceptable range (<70'C.) for up to six
hours when exposed to a 150'C. arnbient temperature. The

A total capture cross section measurement is also obtained

after the last pulse early in the background pause (shown by
the box marked with t) by r-ecording a 6l time-interval decay
curve, identical to that obtained by the TMD-L tool. Since
only the last few neutron pulses will contribute to this signal.
the count rates are reduced significantly relative to the TMDL, but the much slower logging speed dictated by the C/O
measurement cornpensates somewhat for this signal loss_
Thus, a simultaneous formation sigrna (Ipy) is obtained
during the C/O mode of operation. Multiple passes (at least 3)

at 5 ft/min. are required for statistically useful inelastic

measurements. The actual number of passes depends on the
precision required in the oil saturation determination and the
porosity.

ray. and the 1-l l/16-inch diameter power supply section from
tool '. The total make-up length
(including cable head and bullnose) is 27 ft. (8.1 rn). The
portion of the tool that is 2-1,'8-inches (54 mm) in diameter is
14.2 ft. (4.3 rn) long.
The tool has trvo operating modes. The first is the C/O (or
inelastic) mode. In this nrode. the tool is pulsed at l0kHz with
30ps neutron bursts fbr a perio,J of 20rns, which is followed
by a pause to measure background for another 5rns. This
cyclic pattern, show'n in Fig. 2, is repeated continuously.

Halliburton's TMD-L

Spectra are obtained from each detector

in

three time

windows.

The first q,rndori is centered on the lattel part of the

to lecord the primarily inelastic

scattering gamma ray events, although some capntre
activity is also present ar this time.
The second tirnc ri indog' stafts 5ps after the end of the
neutron burst, and extends until the start of the next
pulse to record the caprure gamma ray activity.
The third tirne windorv records the background
activity measured during 3 ms at the end of the
background pause. This activity will include natural
radioactir,ity. as nell as any activation activity that
neutron burst

r
o

rnay be present.

At

passes at 5

rnodest porosities (15-20 porosiry units), 7-9

ftlmin. or 4-6 passes at 3 filrnin. may be necessan.

to obtain satisfactory precision in thc saturation calculation.

The second rnode of operarion replicates the pulsing and
gating sequence of the TMD-L so that a high quality caprure

cross section

poj,; is obtained.

Howevcr, since dete$or

spacings and sizes in the RMT ale slightly different compar.ecl

to the TMD-L, rhe response (Ipy, and near/far count rario.
etc.) is slightly different. A pulse rate of 800H2 is interlaced
every 20ms with a 5-rns pause to Ineasure background (as
shown in Fig.3). Three spectral rneasurernents are alstr
recorded.

One measuretnent duling the burst records inelastic

One measurement cotllrlences l70ps after the end tri

the burst and runs to the stafi of the next burst itr
record the fonnation capturu cvents.
One '' rneasurement dLrring the background paus.

tool uses an existing l-l l/16-inch diameter telernetry/gamma

sPE 53736

events (plus some capture).

records natural and activation events.

This mode provides all the rneasurements norrnalir

provided by a PNC 1og, but u,ith bctter. lithology infonrati(rn
from the capture spectrum. The norninal logging speed for rhis
rnode is 20 ftlrnin. if spectral infoln-ration is not essenrial. or j
ft/min. if high-quality spectral infor.rnarion is desired.
Data Processing
Data processing begins

with deadtime correcion of the decar-

curve data based on a tool-specif rc deadtime, *.hich .s

measured in the horizontal water'-calibration tank. Next, rhe
background is subtracted frout each curve ro obtain the tm:
decay curve. Fitting of a t\\/o-exllonential tunction to thri

corrected decay-curve has been extensively discussei

elsewhere and will not be further discussed here 5'6.
' The inelastic spectral data is flrst correcred for pile-ur,
While pile-up effects are impoltant during the inclasric gai:.
these effects are minimal for the capture and background dai.,
Next, a certain flaction of the capture spectrum is subtracr:c
from the inelastic spectrurn ro renrove the capru;.
contribution. This fraction is deterrnined fiorn laboratoi..
measurements by minirnizing the hydrogen peak in rl::

sPE 53736

A New Small-Diameter, High-Performance Reservoir Monitoring Tool

inelastic spectlum. For the C/O rnode of operation, this

fraction is reasonably constant as a function of changing
formation properties. The capture-corrected inelastic spectrum
is then reduced to count rates in broad energy windows (see
Fig. 4 - top) which primarily represent the oxygen, carbon,
calcium, and silicon inelastic tesponses. Ratios of C/O and
CalSi are computed and used in conjunction with fan charts
developed from laboratory measurements to estimate oil
saturation for a known porosity.

The capture spectrum is corrected for natural

activation background by subtracting a fraction of

and
the

background spectrum fi'om it. This fraction is straight-forward

in that it is merely the relative duty factor of the background

gate to the capfure gate. The corrected capture spectrum is

then reduced to l8 enelgy windows in a manner sirnilar to the
PSG tool7. The counts in these 18 energy windows are fit with
basis (pure elernent) spectra. obtained in the laboratory and
through computer modeling. using a weighted-least-squares
fitting procedure to detennine the relative contributions of
each of the basis spectra. The elements extracted in this
manner are: H. Ca, Si. Fe, Cl. K, S, and Ti. This capture
processing will be covered in rnore detail in a subsequent
publication but the basic plocedure has already been discussed
in Jacobson, et al., 1996.

Laboratory Characterization
Tool characterization is based on test-pit data. Three runs,
each of 300 seconds accurrulatior-r, were averaged to obtain
the response ofthe tool in a given tesrpit and casing/borehole
environment. The lithologies available are 2Yo, l2o/o, and 260/o
porous lirnestone with 6- and 10-inch borehole dianreters, a
freshwater sandpack of 35"h porosity with 6-, 10-, and l4-inch
boreholes, an oil sarurated sandpack of 36% porosity with 6-,
l0-, and l4-inch bbreholes. and a saltwater (150 kpprn)

6-. l0-, and l4-inch boreholes.

lnitial response measurelllents, discussed in this paper. were
obtained in the several test-pits u'ith 7-inch - 26 lb/ft casing
surrounded b,v freshwater cement in l0-inch boreholes. In
sandpack of 42oh porosity with

order to deterrnine the pulse pileup corrections applied by the

processing softwale. a few runs with much longer

accumulation times u,ere obtained at low neutron output to

detennine the shape of clcan specra (rninimal pileup).
After deadtirne correctiolt, pileup correction, and capture
background correctiou, foul windows are applied to the net
inelastic spectrurn, as shown in Fig. 4 (top). These window
count rates. sensitive to the oxygen, carbon, calcium, and
silicon concentlations. are formed into two ratios: C/O and
Ca/Si" For the 7-inch casing - lO-inch borehole environment,
the fan chart in Fig. 5 results, based on the 12oA and 26oh
limestone, the 35% fi'eshwater sandpack, the 36% oil
sandpack, and the 427o saltwater sandpack measurements.
Although only a few data points are used, the physics of the
process constrains the fan chart parameters. The rneasured
C/O .atio is mathematically modeled as:
Rc,,o =

Y..

o**

ro

pt-q)+t

" """""""""(l)

where

sensitivity factor

Y"

carbon concentration
oxygen concentration
scaling factor

Yo
B

porosity
offset due to downscattering ofhigher
energy gamma-ray into the lower windows.

The second tenn on the right-hand side accounts for the effecr
of formation density on the gamma-ray transport. The ratio of

the carbon-to-oxygen concentration, Yc/Yo, is related to the

matrix material, porosity and saturating fluid:
(l)soNCh +

":

':

1-

': 1."- I

(I_

(l))VcaNCca

: ll]:":i."'l

u':^o':
_

l]

,,,

where
the oil saturation
the limestone volurne fraction.
carbon atomic density
the oxygen atolnic density in material x

so

NC*
NO*

Table I shows the carbon and oxygen atomic densities for.

common formation materials and fluids. The lower case
subscript'h'refers to a typical oil hydrocarbon and'*-'refers
to water.

The Y"/Y. ratio defines the basic forrn of the C'O fan

chart. Thus, the only free parameters in equation I are a, p.

and y. The y?aramerer is essentially determined by the C,,O
value at O = I (frorn lneasurements in 100% water), This
leaves only two parameters to fit all the other measured data
shown on the left-hand side of Fig. 5 (shown as symbols). The
oil sandpack at 360/0 and the fieshrvater sandpack at 357o are
strongly favored in the fitting process. For the fan shown in

Fig.5:
Y

Rclo =6.173--916 93@+0.445 ....... .........

(3)

This value of q. is smaller than thar found for Halliburron's

large diameter C/O tool (PSC). This translates inro a reduced
logging speed compared to the larger tool, needed to achiere
the same level of precision in the cornputed fluid sarurarror:
(see below).
The CalSi ratio is modeled similarlv:

Rcysi

= (l

-o)lovcu *[]J* y

where

(I
o_
pI

scaling factor
scaling factor
an offSet factor evaluated at @ = l.

.(4)

G. A. Simpson, L. A. Jacobson, and Rodolfo Escobar

Evaluating the coefficients by fitting the measured data

shown (as syrnbols) on the right-hand side of Fig. 5 leads to:
RCa/Si = ( I - O)[0.29Vc,

- 0.221 + 1.52

...............................(5)

lnterpretation
The RMT measures RC/O and RCa/Si ratios. What the client
wants to know is So for a given porosiry (@). Equations 3 and
5 can be solved for Yc/Yo and VCa in terms of the measured
quantities and porosity. These (Yc/Yo and VCa) in turn can be
inserted in equation 2. which can be solved for So. Appendix

derives the follorving equations. which express this

sPE 53736

implies that a single logging pass ar 5 fVrnin.

will yield a

statistical uncertainty in S" of l2o/o. By comparison, the PSC

would obtain an uncertainty of 9o/o in the same circumstances.
Thus, it is necessary to increase the number of passes or slow
the RMT logging speed by -40o/o to achieve the same
performance as the larger tool.

Field Test Examples

Exarnple #1. The example log in Fig. 7 was run in a Southeast
Texas Gulf Coast Miocene formation. Openhole logs for rnost
of the wells in this area ale old E-Logs. Fresh and brackish
formation waters have rnade interpretation with pulsed neutron

AC/o = RC/o -0.2RCaS +0.02<D-0.185+k ....-.............(7)

capture logs difficult in these forrnations. Additionall,v.

drilling records at these depths were sketchy and somerinres
missing. It has long been thought that the Miocene sands had
the potential of producing oil if water sands could be
distinguished from oil sands. The operator of this rvell sas
rvorking over a lower producing sand and gar.e us the
opportunity to run carbon/oxygen logs over the Miocc.ne

The parameter' 'k' is a nurnber near zero that\weeps up all

the residual environmental etTects (borehole size, casing size,
cement type, etc.) which rnay lead to a AC/O value that is not
zero in a water saturated reservoir. The first two tenns on the

calcium

procedure:
So = 1.27

(1.-0.37o)AC/O

O(AC/O+0.l78pHC)

where,

right-hand side of equation 7 essentially represent what is

done on the log when C/O and CalSi cun'es are scaled and
offset to overlay in water zones and then separate in oil zones.
Fig.6 shows a plot of AC/O versus porosity. The symbols are
sanre data as shown on the left-hand side of Fig. 5 after
applying equation 7 (rvith k : 01.

Measurement Precision
One standard deviation of the C/O ratio measurement is given
by Jacobson. et ul.. 1993t'

section.
To evaluate the response charactcristics and resolution of
the RMT, the well rvas also logged with the 3 3/8" Pulsed
Spectra Gamrna Tool (PSCrNt). Track 2 of the exarnple in Fig.
7 contains a plot ofthe carbon to oxygen ratio in red. and the

to silicon ratio in blue. The curues provide a

qualitative indication of hydrocalbons when the C/O cune
crosses over the'Ca/Si curve. The green shading indicates
hydrocarbon. Track 3 shows tl:e salne curve set from the R\lT
measurement. While the absolute value of the RMT did not
rnatch the PSCT, the curves are. horvever. sinrilar in characrer.

Both logs indicate a potential oil bcaring sand at X560 r.r

X580-ft. The sand at X400 to X430-ft. indicates hydrocarbon.
but at a higher water saturation. Dlre Io other more lmpon,lnr
deeper pays these zones have yet to be tested. The results frtrnr
this test shorved that the RMT response is very similar to rhrr
of the PSGT.

oCtO = t.2RClO

.................(8)

Example #2. The second example log

Southeast Texas

where

C
O
S
D

=
=
=
=

the count rate in the carbon window

the count rate in the oxygen window
the logging speed in fi/min.
is the etfective depth averaging in ft.

accounts for the capture

background con'ection requiled. because C and O are gross
counts (not corrected for caprure background). For rypical
logging conditions, the total inelastic count rate (ITCR2 on the
log) will be -4000cps. For this siruation, C and O are tlpically

D is usually 3.25ft. The factor 1.2

-270cps and -570cps. respectively. For a logging speed of 5

ft/min. and a depth averaging of 3.25 ft: oc/o:0.007. The
dynamic range (DR) of the measurement is defined as the
change in R116 as the fluid saturation changes from water to
oil. From Fig. 5, the DR is 0.057 35oh porosity, which

^t

is also trorn ihc

Culf Coast region. but wr.s logged in the Fr:tr

formation. The Frio sands in this area have a Rw value of (t tt5

ohmm and corresponding Sigrrra water value ol 6l ;..r.
Typicat resenoir monitor logs usr"rally consist of PNC lc'.r.
due to the higher formation \\'ater salinities. Many of the zon:s
in this well have been produced tl'our other wellbores. Th.operators of this well rvanted Io run a reservoir monilor loc:..
determine what zones were suitable for re-completion. Ti-.:
field test of the RMT was recorded in both C/O and Captu:.'
modes. During the C,O mode. five passes at 5 feet per rnin;::
rvere recorded over the zones of interest. The specral d':-:

from the fir'e passes were surnrned together, and n:.,.

ratio and lithology ratio values \\ i::
recomputed. The CiO ratio {COIR) and the lithology ra:'.Carbon/Oxygen

(LIRI) are ploned in track 2 of Fig. 8. The formation captu:-:

Sigma (SGFF) from the Caprure rnode is also displayed :r
track 2. This data was combined u'ith openhole logging d;:;
consisting of deep induction. bulk density, neutron porosj:.

sPE 53736

A New Small-Diameter, High-Performance Reservoir Monitoring Tool

and garrurta lay to producc the cornputed water saturation log

represented in Fig. 9.

Three saturation curves are plofted

in track 2 of

the

example. The black curve shows water saturation based on a

shaly-sand analysis of the openhote logs. The green curve
shows the oil saturation computed from the C/O ratio and the
lithology ratio. The blue curve shows the water saturation

utilizing sigma fronr the RMT capture mode, which was

computed assurning gas in the formation. The gray shading
between the openhole watel saturation and C/O oil saturation

curve show the amount of hydrocarbons that have been

produced since the well was originally drilled. The red
shading is the difference berween the oil saturation computed
frorn the C/O ratio. and the water saturation computed frorn

the formation sigma assuming gas. Differences in the two

saturations are intel'preted as lesidual gas. Since gas has less
carbon than oil, the oil saruration curve from the C/O ratio will

zones for re-perforation. The log showed two sand intervals

from X530 to X560-ft, with an average oil saturation of 45o/o.
This interval was perforated as indicated in red on the figure.
After perforating, production was increased to 200 BOpD with
65%o water cut.

Summary

A new small

dianreter C/O logging tool, developed by

Halliburton and named the Reseruoir Monitor foot (RUf y,
was described. Laboratory measurements of the inelastic
response (CiO) of the tool in well-characterized environments
led to an interpretation procedure based on fan charts. This
tool provides all the inforrnation generated by larger-diameter
pulsed-neutron specrornetry tools. plus additional capabiliry
provided by a near-spaced detector.. A second n-rode of
operation provides a decay-tirne t.neasuretnent comparable to a
standard PNC log, and specral information for lithologl,,

be lower. Most of the gas seen in the pore spaces is not

evaluation.

producible.

reduction in precision of -20oh compared to a larger-diameter

tool can be achieved by logging -400% slower or by taking
-40%o more passes. Finally, several field logs demonstrated
applications of the tool, one of which was a comparison ro a
high perfonrance, larger-diamerer rool. Also, the tool has been
used successfully in Venezuela and has contributed to
significant increases in production.

The interr'al on the log tr-orn Y0l2 to Y025-ft. showed a

large difference between the two curves. This zone is
considered to be a gas cap in this interval. The oil saturarion
frorn Y025 to Y040-ft. acrually shows more hydrocarbon than
the water saturation curve, thus. this interval has more

oil

than

gas.

The interval atX947 to X962-ft. shows an oil saruration of

38% still left in the formarion. This interval should produce
substantial amounts of oil but will also produce some water.
The interval frorn X983 to X994-ft. shows saturations around
50%, which is near the original openhole saturation. This
interval has been chosen fol le-completion.

Venezuela Field Log Examples

Field Example

l.

The u,cll

in Fig. l0 is frorn

Western
Venezuela was originally drilled and cornpleted in 1938. The
RMT was run to evaluate potential intervals for recornpilation.
Production test two lnonths before running the RMT shorved
the measured production to be 200 BPD total fluid with 60%
water cut from a lower producing interval. Fig. l0 shows the
computed analysis of the RMT C/O data. Track 7 of the log is

a plot of the original electric log from 1938. The E-log

indicates a sand from X290 to X390-ft. The analysis of the
RMT data showed the intelval rvas composed of several interbedded layers of sand and barrier shale. Oil sarurations in

tlrese beds ranged florn 30oh to 600/o. The average oil

saturation over the entire inten,al was - 45ah. The sand rvas
perforated over the entire interval indicated in r.ed on the
figure. After pelforating, the u,ell was flowing 1836 BPD total
fluid with 580% rvater cut. Bets'een X050 and X200-ti. The log

indicates for,rr additional sands that have potential for

recompletion in the hrture.

Field Example 2. The well in Fig. tl is another rvell from

Western Venezuela. This sidetracked i'':ll was or;ginally
perforated in a lower interval that was now producing 25
BOPD with a 90oh water cut. Fig. I I is the cornputed analysis
of the C/O data from the RMT that was run to locate possible

An error propagation analysis showed that

Acknowledgments
The authors wish to thank the clienr companies for permission

to use the log examples shown here. Also the editorial

assistance of Steve Bollich, graphics assistance of Charles

Conley and Karen Campbell, ancl the programrning and

processing sllpport of Dennis Durbin are gratefulil.
acknowledged.

Significant portions

of this paper \\.ere exkacted trorn

SPWLA paper K presented ar thc ,.19'h Annual Log*ein_u

Symposium,26-29, May 1998 in Ker.stone, Colorado. The
authors wish to thank the SPWLA fbr perrnission ro use thc
information for this paper.

A New Small.Diameter, High'Pedormance Reservoir Monitoring Tool

sPE 53736

EH

.-{. l=

s$
EF
9o
5
cL
g$

J
u

o.g

{t

sd

t5

:E

AE

HE!

fi.EEC
i o

!i

F:3 r
* 5btbE
:

E9HE
e htl-c

>g- o
ttsF!,
o rD =,ts

9i st

igiir6
.srE
g o a_.cH
T*.E fi

e;

lf

fv'
trI
t9

EEg
EioT

r':'
l{l

5' (E(o
iuol
o

if1
lF

d' o-'cr

)tlr

g EEE

a
d'

EF"
3,+5?

z.

==

:tT
r'-,ts

*t

*;

3'= I
F;qrk
eI
'i
EE$

T
f-:'

Hi{E
E3

Eai

O s.! o

l-Eoc

EEE
otrcF
o= o

Fb-F
F O 6-

JEEg

.sT d

l!.;('uJ

g!--.%%*

s.LN :-ltf

3 fl 3r l1$l',lts*tl

ll'l nc{J'3 3 Il Tf I'l H'3*'J

sPE 53736

G. A. Simpson, L. A' Jacobson,.and Rodolfo Escobar

10

Corrected Ca/Si

Corrected GIO
0.580

1.650

r/

8/

1.600

0.560

/,

0.540

l.

,/'

0,520

a
(g
o

tr
X

,. 7/'10 FW

1,500

x 7/10 SW

r
1.450

1.400

0.480

0.460

0.1

0.2

0.3

PorositY

Oil

0.1

0.2 0.3

0.4

s,6/rq

f.

"il-l

0.5

PorositY

in a 1O'inch borehole' The

c/o (colR) and caisi (LlRl) ratios tor a 7'inch casino 3' The ca/si ratio fan is
Here is the inelastic fan chart response_.tor the
Fio. 5
svmuors using-eq'
r"l"rr"r"nt"'.hoiu-,iivih"
r),6
tI
iii"
paper,.and.fitted
c/b tan is generated by Eq.2 in the
size/horesize werrbore fluid type:
.oripri.b to tn" r.o ,i"u"ui..enti snown by the legend (casing
represented by Eq. 5,
"no ""-nFW = fresh water)
SW = salt water'

g.olt o swl

1.350

1.300
a.4

7/10

u 9.6/10 fw

0.440

a.sl6 sw
+.s/o rw

o
o
0.500

F+s% oir I

1.550

A New Small-Diameter, N-ligh'Ferformance Reservoir Monitoring Tool

sPE 53736

Aco
0.080

0.060
+

rl

4.516 FW

0.040

4.5/6 0il
4.5/6 SW

, 7110 FW

o
o

x 7110

0.020

sw

7110 Oil

9.6i 10 fw

9.6/10 sw

. 9.6110 oil
0.000

-0.020
0.1

0.2

0.3

0.4

0.5

PorositY

lab
c/o (Ac/o) fan obtained by applying- the procedure in Appendix A' The
water)'
fresh
FW
water'
salt
sw
tyiel
nuil
=
*"riror"
=

Fig. 6 _ This chart is a tithotogy-independent diflerential

u]Jioi"s'."
measurements, identified by the tegend symbots t"u"inl
were transformed using equation 7'

G. A. Simpson,

12

sPE 53736

L A. Jacobson, and Rodolfo Escobar

It Ir.:

8br{ lt,!st.

riR lrhlT

_t

-li-:::;::
li1.J

- ' '-

Fi
AT

:.,
,lr

t\j

4*

x450

-F
I

-..-l
.r'"

-r
t\
ft1r.

"l

."..\

'j

1a.p

"x.

t^-a'

,,TT

.Ei
-\l

tt.

3<

ti

rt\
IJ

t-r
*l

Hsfln

'--ll.

--'r

'.

'?r
),'
-^+'
a-i

tj

*?
tr-

tq
Fi
;t

{;-\l
\-2
At6

,t

{:*

-r

.{-

ai
'|

,-?

-t
(

.*3
.t 1
.f

-r
l-

I
'f

(!

.'1

"7

{'

;r

-J

L}

#f
-3d

--h

:..|

,ri

c:,
ii
,7,:

Hfi5O

rta.
,!!

t,ai\i
'tl

Fig.7_This

rog comparesthe

over a Miocene,ormation

,i I

duit co"st w"rr.

ri'"

.{.

']

-':l

tracks 2 and 3 respedtively)

rhe large dlameter-(3-3/8-inch) PSGTwith the RMT(in
PSG data'
the
from
derived
were
ca
and
si,
for
K,
lri"r""
vi"ra

c/o and calsi ratiosJrom

r"ii"

r;

".ri""

13

Tool
A New Small'Diameter, High-Performance Reservoir' Monltoring

riAl
l'..1

! rilH

G,'\\llt.{..\
,'ti

lL'tr"1'\
l-,.\1L

l{

I'ilt

Pl

l.l

rlt-l $.1.\"lllA

l,l

lr\l'l

.t

';:'
'"]',

a.

t'a
4*.:
.-\i

.>

i:t
r

5(

tt:
.1.,

-t

"4:

',

-l
t-_a

--\

1l
I

1 ttr

;l
';--

'L--

{--e

]
!

.tL
:1

il*

Y1CI0

.|.
t.-

I
l

l!p

i5S

passes run at 5 ft/min were averaged to produce this log'

Gulf coast welt in the Frio tormation, Five
Fig. g _ This RMT log was run in a Texas

sPE 53736

G. A. Simpson, L. A' Jacobson' and Rodolfo Escobar

.lt.t'l

-.,...,t.. . .....-..-

it;

l
!

,,

i,
,

r:

lir-.-.1 ,

l,i

i:' r ;rl
t":l
:lrrl

;:ir.!r
It:

rri,.
I

*'i

..,,

i
i

**:rr

Irt-5ii

14llk'"

7,-+-:-::
\'{ll

ll

,:

_"jjlr

*+

, *-, f,|{ flt*{l

;*'*"-

\ {jj rt

L1

w
,-

!i

i I lt.i

*
.-.. *--

*'
g
-rc..
.4*

Fig. 9

RMT data is combined with openhore data

while the green shading in track 3 is oil'

saiuiation,

red shading in tracks 2 and 3 is gas

to provide this interpretation of the log in Fig. I' The

A New Small-Diameter, High-Performance Reservoir Monitoring Tool

sPE 53736

Fig.10_TheseRMTcomputedres-ults^were.genelatedfromdatar:ecorded.in-3^,WesjernVenezuelawell'Newpedorationsincreased
fluid with 57% water cut'

priir"tion

rom zoO BPD todL fluid with 60% watar cut to 1836 BPD total

Rodollo Escobar
G, A. Simpson, L. A. Jacobson, and

Fio, 11

*ii"i

.ut

sPE 53736

25 B.PD with 90%

perforations increased production rrom
western Venezuera we'. New

RMT computed resurts rrom

to 200 BOPD with 60% water cut'