SOCIETY OF PETROLEUM ENGINEERS OF AIME PAPER

6200 North Central Expwy. NUMBER SPE 6542

Dallas, Texas 75206

ELECTROMAGIWTK PROPAGATION
,.. a New Dhnension in Logging
By
Thomas J. Calvort - Rama N. Rau
ScMumberger
Well Servkea, Houston, ‘J%xas
Larry E.Wells
Member SPE-AIME, Schlumberger Well Services, Houston, Texas

THIS PAPER IS SUBJECT TO CORRECTION

@Xopyright 1977
American Institute of Mining, Metallurgical and Petroleum Engineers, Inc.

This paper was prepared for the 1$)77 47th Annual elsewhere after publication in the JOURNAL OF PETRO-
California Regional Meeting of *he Society of Petroleum LEUM TECHNOLOGY or the SOCIETY OF PETRO.
Engineers of AIME, held in Bakersfield, California, April LEUM ENGINEERS JOURNAL is usually granted upon
13-I 5, 1977, Permission to copy is restricted to an abstract request to the Editor of the appropriate journal, provided
of not more than 300 words. lllust rations may not be copied. agreement to give proper credit is made. Discussion of this
The abstract should contain conspicuous acknowledgement paper is invited.
of where and by whom the paper is presented, Publication

ABSTRACT

One of the primary requirements of well logging water salhsity, except h ravges close to saturation.
is tbe ability to dis$hsguisb be~ween oil and water. Notable h the fact that tbe tsew too2 h most accurate
ResWvity/ConductMty and pulsed-neutron took can h fresh waters, an area that bas been troublesome.
do it ij the water h fairly s4&se, and if its salinity & An experimental tool has been h tbe &4d for
known. However, a method has long been needed that many months, running logs h eight states and a
& less depe~dent upon knowledge of water salinity, Canadian province. Field test results have exceeded
Scblumberger’s new Electromagnetic Propagation expectai~ons, and equipment A behsg built for corn-
Tool measures the phase shift and attenuation of merchsl seru~ce.
microwave-frequency energy propagated tbrougb tbe Tbh paper gives the o@ratAsg priciples of tbe tool,
formation near the borebote. From these measure- desctibes tLe hterpretatiors method that has been de-
ments, it is possible to derive values of porosity and veloped, and presents severs! jield examples to Wsstrate
water saturatio~ that are practically hdependent of tbe hnfiortance oj this new logghsg measurement.

Prinrcd in U.S.A.
ELMXROMAGNETICPROPAGATION. . . ANEW’ DIMENSION IN LOGGING SP 6542
SP 6542 THOMAS J. CALVERT— RAMA N. RAU — LARRY E. WELLS 3

‘ ELECTROMllGNETIC PROPMATION
.. .aiHew Dimension in Logging
By
Thomas J. Cakert - Rama N. Rau - Larry E.Wells

INTRODUCTION temperature, it is below 17 gigaHertz. With the exception

of water, there are very few materials abundantly found
The Electromagnetic Propagation Tool (EPT* ) meas-
in nature which have permanent electric dipoles. The
ures formation parameters not previously available. The
geometry of the water molecule is responsible for the
primary measurements are propagation time (tl,l) and at-
dipole, hence for the unique dielectric properties. A bore-
tenuation rate (Al) of an electromagnetic wave traveling
hole dielectric measurement in the 109-Hz frequency region,
through the formations. The propagation time of water
where the dipole polarization of water dominates, should
differs sharply from those of gas, oil, or matrix rock; and
lead to a measurement of water content which is indepen-
is moreover little affected by the salinity of the water. This
dent of salinity.
tool therefore opens up a way to evaluate water saturation
Table I gives laboratory-measured values of propaga-
that is relatively independent of water resistivity (salinity),
tion time and dielectric permittivity (relative to air) of
and in fact is most accurate in the fresher waters.
typical reservoir materials.
This progress report discloses the theory of measure-
ment, and some details of the experimental tools which
have been tested extensively in the field. A discussion is Table I
presented of the nature of high-frequency electromagnetic
RelativeDielectric Loss-Free
propagation in earth formations, together with an Appendix
Permittivity Propagation Time
giving a mathematical development of the measurements
MATERiAL (e) (tpe)
and their relationship to useful well parameters, Finally, a
tentative interpretation technique is described, supported Gas or air 1.0 3.3
by several field exampies showing the practical usefulness Gil 2.2 4.9
of this new logging tool in heavy-oil environments. Water 56-80 25-30
Quartz 4.7 7.2
THEORY OF OPERATION Limestone 7.5 9.1
Dolomite &9 8.7
Dielectric permittivity is one of the main factors de-
Anhydrite 6.5 8.4
termining electromagnetic propagation in a material. Di-
electric permittivity of any medium is proportional to the
electric dipole moment per unit volume. The electric dipole
moment is made up of one or more of the following ef- METHOD OF MEASUREMENT
fects: I ) Electronic, 2) Ionic, 3) Interracial, and 4) Various schemes for measuring dielectric permittivity
Dipolar. Since each of these dominates over a certain range in the Iaboratoryl can be categorized as reflection, trans-
of the electromagnetic spectrum, they can be separated mission, and resonance techniques. The last method does
experimentally. not lend itself to measurement in a imrehole. Of the
The electronic contribution is due to the displacement former two, the transmission method is superior for meas-
of electron clouds, and is the only one that operates at urement of dielectrics with high conductivity losses. The
optical frequencies. The ionic and interracial contributions borehole filled with water-based mud lends itself nicely to
come from displacement and movement of ions, hence are such a measurement.
confined to low frequencies. The dipolar contribution is due Fig. 1 represents schematically the transmission of an
to permanent electric dipoles which orient themselves in electromagnetic wave at the interface between a high-per-
the direction of an applied electric field. The frequency mittivity, lossy* medium (mud or mud cake) and a medium
region where this contribution dominates depends an the with lower permittivity and losses (formation). According
freedom of movement of the dipole. For example, in ice
it is below a few kiloHertz, while for water at room ● Electromagneticenergy losses in dielectric media arise primarily
from two phenomena: induced currents in conductive materials,
and dielectric losses caused by the energy consumed in reorienting
“Trademarkof Schlumberger. molecular dipoles in an alternating field.
4 ELE~OMAGNETIC PROPAGATION. . . ANEW DIMENSION IN LOGGING SP 6542

to the law of refraction, the transmitted wave is refracted

IA
away from the normal in the lower-permittivity formation.
,0
If the incident wave is at the critical angle, the refracted
s ,e~ wave will travel along the interface. It can be shown that
Formation E 0’ the velocity of propagation along the surface will be de-
i? ?@” termined by the permittivity of the formation. The surface
wave continucmly sheds energy into the mud. By placing
two receivers in the mud column, a differential measure-
ment of phase shift and attenuation (see Appendix) can
be made (Fig. 2). Since operating frequency is constant,
phase shift is easily converted to propagation time.
If the borehole propagation times are shorter than
those of the formation (such as the case of oil or gas-
filled holes), then the travel times recorded may be
unreliable.

DESCRIPTION OF TOOL
The tools now in the field carry two transmitters and
Fig, 1 — Diagram illustrating electromagnetic-
two receivers on a wall.contact pad, configured as shown
wave refraction at a mwi-borebole hster-
in Fig, 2 and Fig. 3, These ttammitters and receivers must

Fig. 3 — Tbe WT atstetsnapd.

Fig. 2 — Schematic of EPT afitctsnafud, shoul-

itsg @AscipJe of 2-rece4ver measurement
of transit time. Fig. 4- Tbe prototype tool&s operatwn.
SP 6542 THOMAS J. CALVERT— RAMA N. RAU — LARRYE. WELLS 5

receivers, similar in principle to the widely used method

of measuring At with a two-receiver sonic tool. In a
similar manner, the two receivers produce cancellation of
any effects caused by mudcake or variations in signal coup-
ling (so long as both receivers are affected equally).
To reduce any error caused by sonde tilt, the En
uses an antenna configuration similar to the transducer
array used in a borehole-compensated sonic tool (Fig. 5).
Transmitting antennas are placed above and below the
receiver pair, and are pulsed alternately. Simple geometric
considerations show that if these two transmission modes
are averaged, the first-order effects of pad tilt will be
eliminated (Ref. 2).
The block diagram of Fig. 6 shows the functional
layout used in the present EPT tools.
The basic principle of the tool involves launching a
surface or lateral electromagnetic wave along the surface
of a conducting pad. In the absence of mud cake, the
electromagnetic wave would move along the pad face past
two receiving antennas, but in the normal borehole case
with mud cake present, propagation takes place on the
surface between mud cake and formation. The phase shift
and attenuation per unit distance along the surface of the
pad are proportional to ~ and C (as shown in Appendix
1) for a plane wave.

ANTENNAS
The antennas used are cavity-backed slot antennas de.
signed to radiate electromagnetic energy efficiently at a
frequency of 1.I gHz under varying load conditions, and
Fig. 5 — Schematic O! EPT, showing borehole
also to withstand the extremes of temperature and pressure
competasatio~for wasboat or tilted sonde.
encounteredin the borehole.
Each receiver antenna sends its signal to suitable
processing circuitry, which culminate in measurements of

w“”” —
----
~ Signol
RF Signol

,*
‘\% /
%●.. /’
% --------
.“
T I%0s4shift, upper orroy
(6IJ)
1

I
Fig. 6 —Block scbematk of tbe prototype EPT,

be antennas, to operate as they do in the microwave fre-

quency range.
The antenna pad projects slightly from the body of
the tool, and is maintained in contact with the borehole
wall by a powered backup arm (Fig. 4) which is hydrau-
lically opened and closed on command from the surface.
This backup arm also provides a caliper reading.
Fig. 7 — Shsgk-cycle represeistation of tbe bore-
As mentioned earlier, the tool uses a differential bok compensated measurement of phase
measurement baaed on the signals detected by near and far shift.
6 ELBCL’BOMAGNBTIC
PROPAGATION. . . A NEW DIMENSION IN LOGGING SP 6542

Attenuation, E Travel Timem (tPl ), nsec/m .

dB/m
o 500,
~n Caliper 15.
----------------

8
0

Fig. 8 — Ats utsaveraged EFT log shows fine detail. Repeat section (faht curves)
of tbe Attenuation and tplcurues show excelktst repeatability.
SP 6542 THOMAS J. CALVBRT— RAMA N. RAU — LARRY E. WELLS 7

attenuation, in decibels per meter, and propagation time Tracks 2 and 3 are given to the principal measurement,
(%]) in nmoseconds per meter. travel time (tPl), in nanoseconds per meter.
Also in Track 3 are presented the signal levels from
THE PHASE MEASURE CIRCUITS the two receivers. The chief use of these curves is to monitor
Since the received signak can vary over a rangeof 60 the primary signal detection at the receivers, which provides .
dB or more, the two IF signals (Fig. 7) are amplified in an indication of the relative reliability of the log parameters
circuits designed so that the amplified output signal can be at any level,
sent to a zero-crossing detector without introducing a level- A self-evident and very real advantage may be in-
dependent phase shift. This is important, since the near ferred from the 4-cm spacing between receivers; the tool
and far levels can be quite different and the output pulses has excellent vertkal resolution. The log of Fig. 8 actually
of the zero-crossing detector will be decoded (decom- looks overactive in spots, but its repeatability (gray curve)
mutated) in synchronism with the switching of the trans- testifies that the recording is valid. In fact, the data recorded
mitters. Any level-dependent phase shift would introduce by the tool are tm detailed for direct merging with other
an error. logs via computer. Averaging (smoothing) subroutines are
Clearly, when Ti is transmitting Ri is the near receiver thus required preliminaries for programs using EPI’ data.
and Ra is the far receiver. When T2 is on, R2 is the near
receiver and RI is the far receiver. Therefore the resulting .INTERPRETATION
switched outputs RN and RF always correspond to the near Basically, the EPT log is interpreted to determine the
and far receiver signals. The zero-crossing pulses are used
bulk volume of water present in the formation. The bulk
to trigger a flip%op: on with the near signal and off with volume of hydrocarbons is obtained by subtracting this
the far signal. The percentage of the time that the flip.flop from the total pore volume. Given the narrow spacing and
is on (out of 360 degrees) corresponds to the measured consequent shallow depth of investigation, the measurements
phase shift of the signal traveling between RN and RF,
are normally related to the flushed zone.
and is proportional to tP]. Some of the characteristics and limitations of the
THE ATTENUATION MEASURINO CIRCUIT experimental tools are listed in Appendix 11.

“~hesametwo signak that go to the phase measure

APPARENT WATER-FILLEDPOROSITY
circuit are sent to wide-range logarithmic amplifiers. The
peak output of each amplifier is proportional (afer Q’com- The apparent water-filled porosity* (4EPT) is derived
mutation ) to the near and far received-signal ~e * ,, LN from log readings in a way similar to the derivation of
and LF, in decibels. These signals are sent to the surface, porosity from sonic At. The computation requires knowl-
where their difference is converted to attenuation in dB/ edge of the tool response to pure basic components of the
meter. In addition, the value of LF serves to warn when formation.
received signal Ievels are too low for reliable interpretation. If the log reading is tPl, the apparent water-filled
In the present tool, this can occur in conductive formations, porosi~ is:
such as unconsolidated shales and high porosity-high %1- tpm
#JEPT= t (1)
salinity formations, whose resistivity is less than about” *W- tpm,
2 to 3 ohm-m.
where tP~a is the response iri 100-percent matrix rock
TOOL CALIBRATION tPWis the response in water.

A stable source of reference signals is included in The validity of using a time-average formula of this form
the sonde. The calibration of the log measul’ements is for determining porosity has been well established in
achieved by introducing phase shifts and attenuations of sonic interpretation.
known values into the receiver circuits by meams of a The EPT measures the propagation time and attenua-
surface switch. The phase shifts used for ulibr~ tion are tion rtite of an electromagnetic wave through the forma-
keyed to propagation times over connecting cables of tion. The measurements are not only determined by the
known length, and are thus insensitive to circuit variation. dielectric permittivities of the matrix and fluids, but also
A further validation of tool operation is provided by a by the salinity of the conductive medium. However, under
tcx+in-air check prior to each job. fresh mud conditions, experience indicates that the effect
of a decrease in dielectric permittivity due to salinity is
THE LOO compensated for by the eflect of increased conductivity
Fig. 8 is an example of the log presentation cur- with salinity, Thus the -.propagation time of water remains
rently in use. relatively independent of salinity except in the more con-
Track 1 contains a conventional Caliper curve, taken centrated ranges.
from the motion of the backup pad, and the Attenuation
curve, scaled in dB/m. ●#EPT @~ a vw ~OSe esti~* of *C ~lk VOIWICof Water,
even when she fonnarionsare hydrocarbonbeariog.
8 ELECTROMAGNETIC
PROPAGATION. . . ANEW DIMENSION IN LOGGING SP 6542

vs terspature for average temperature and pressure

gradient conditions, Knowledge of water salinity is not
required for determination of tPW,

MATRIX PROPAGATION TIMEs tw

If the lithology is known, tpm, is used as taken from
the tw’s of Table I. With m-known or mixed Iithology,
28.@5
t~macan be estimated from the CNL* and FDC* logs.
From the CNL-FDC crossplo~ determine prosity,
$ND (Ref. 4 and 5). This porosity will be used
when computing water saturation.
Estimate pm, (apparent matrix density) by inter-
polating between the lithology lines, or: using #ND
from Step I and the bulk density reading from the
FDC, calculate p~a from the standard density-
porosity relationship:

● The tPm is read from Fig. 10 by entering with

p~. for a given Iithology.

NOTE: If the CNL and FDC are greatly

aflected by light hydrocarbons, other tech-
niques should be used to derive formation
porosity and pm,.

“Trsdcmark of Srhlumberser.

MATRIX PROPAGATION TIME - tplml

FORMATION TEMPERATURE - ‘F

Wg. 9 — Chart sbowhsg temperature and pressure

carrectbss, assndsg average temperature
and pressure gradknts.

This characteristic has been established by laboratory

measurements, and verified by log interpretation studies.
As with the Sonic log, the propagation times are used
without any corrections for attenuation.
The attenuation rate appears to be primarily a func-
tion of the shale quantity. Thk is supported by its good
correlation with the Gamma Ray and SP as well as with the
LL8 in sand-shaIe sequences. Investigations in this area are
continuing.

PROPAGATION TIMES
Table 1 lists the propagation times for fluids and
minerals most often encountered in logging.

PROPAGATION llME OF WATER: $).

The propagation time for water decreases with in-
Pig. 10 — Lwolo y Chrrt allows dktetmhation
creasing temperature, and very slightly with increasing of ~m ! rom pm and hsowkdge of
pressure. Fig. 9 is a plot of propagation time for water ltibology.
‘SP 6542 THOMAs J. CALvkxm— RAMA N. RAU — LARRYE. WELLS 9

tpmo tpmo
E P T SATURATION
:-,8, ?,? 7
* NOMOGRA~

-loo%

-90

-80

-70

/
-60

--------- -5s%
- so

c
,- I - -40
Fi
4-
-30

-20

-10

0
sb. Do-cI!M. AD. D&< h WATER
LITHOLOGY LITHOLOGY SATURATION

Fig. 11 — $=. nomograpb. Dashed lhse sbosus compntdwts fw Well A, Astext.

WATER SATURATION: (Smb value of bulk volume water is determined by:

A good estimate of S.o is given by: BVWEPT = &o . $N~ (5)
@EPT
&o=— (3)
. . GRAPHICAL SOLUTION OF S=.
AID
Fig. 11 is a nomograph for computing &o.
A more accurate value of & can be obtained if the Input Needed
type of hydrkarbon is known and characterized by its
● tpl Propagation time direct from log
propagation time tPh.
● #ND Formation porosity
The complete time-average equation is:
● tpma Matrix propagation time as derived from
tpl= #ND %0 tpw + 9%D (1 - SO) tph + (1 - #ND) tpma
pm, or known Iithology type
from which
● tpw Water propagation time determined from
SO= (tPl - tpm~)+ 4ND (tpma - tph) formation temperature
(4)
#ND (tpw - tph)
● tpll Hydrocarbon propagation time, or hydro-
With hydrocarbon taken into accoun$ a more accurate carbon type
10 ELBm0MAC3NBTIc nOPAGATION . . . A N8w DXMENSIONIN LOGGXNG SP 6542
.

t~,.
COMWTER PROOESSEDINTERPRETATION

r
WELL OATb
WITH GOREOATA
UAWA RN Lo s- - EP

EPT I B { Awressd fartwlty }

1
mm I soil - Cer
-------
MCW91JM
m
--l
.1-
s $N ,s Mo$?ora
.-. —-
Iehr .--! .---. *-

I.
’78

ml

Fig. 12. Logs d cwe-am ysis results from Well A,

Track 2 Saturation analysis, & from SARA-

BAND* (Ref. 6) and &o from Eq. 4
WEU A – Kern County, Ctslifornks Track 3 Buik Volume Amllysis from cores (cored
intervals indicated at left of track.
The EIW log, neutron-density porosities, CO1: an-
Track 4 Bulk Volume Analysis from EPT.
alysis and computed results are shown in Fig. 12.
Track 5 Buik Volume Analysis from SARA-
Examp!e Compfwtion BAND;
In order to use the nomograph or solve Eq. 4 for In foliations containing very heavy crude oil or tar,
&~, the following input is required: drilling fluids displace very little hydrocarbon, if any. When
invasion does occur, filtrate wili displace some formation
● $)1 Propagation time, direct from log. For the
water, but the bulk volume of water remains the same.
depth of 778, tpl = 10.9
The shallow-reading EPT will be desponding to formations
● 4 Formation porosity with the sme proportions of hydrocarbon-to-water as are
$N = 36%, #D= 33% (both s~dstone) : seen by d=pei reading devices. In such cases of “im-
results in #ND= 3470 and a pma of 2.68 movable” hydrocarbon, water saturation as computed from
the EPT should equal%.
● tpma Using ~.= 2.68, the quartz.dolomite lithol-
In this example, RW is a complicating factor for
ogy set is selected to give tprna= 7.4 (see
normal log interpretation due to the fresh formation waters.
Fig. 10)
(RWS 1 to 4 ohm-m). The RWused for the SARABAND
NOTE: This formation does not
computation was taken from measurement of DST samples
contain dolomite. However, the
and appears very reasonable over the lower zone (2,31 6 —
Quartz-Dolomite Line is nearly the
2,172 ft) and the known water zone immediately above,
same as the Quartz.Shale Line for
The interval above 2,000 ft is water-filled, but with an R~
the area.
much greater than the lower zone. The EPT interpretation,
● tpw Using a temperature of 85°F, tpw= 28.85 being es. mtiaily independent of Rw, correctly identifies
(see Fig. 9) the zone as water beadng.

c tpll Oil is assumed as the hydrocarbon type, with WELL C - Athabasca Tar Sand, Alberta, Canada
tph= 4.9 One of the important reservoir parameters needed to
The input of these data is shown in Fig. 11, result- evaluate the very shallow tar sands is the percent-bulk-
ing in an !&Oof 53 percent. weight tar. This is the ratio of the tar weight to the forma-
The comparison of core analysis with the computeo tion bulk weight expressed in percent (Ref. 7). No change
processed Em interpretation is quite good. in the basic interpretation technique is required for com-
puting the percent-buik.weight tar.
WELL B-California PlioceneTar Sand Fom~atiomporosity is computed from a neutron-density
crossplot. A plot of Prnn and tpma (apparent m~fix pro.
Computedresultsof thiswell are shown in Fig. 13.
Track 1 Saturation analysis from cores. ‘Tnsdemsrk of Schlumberser.
SP 6542 THOMAS J. CALVBRT— RAMAN. ILAU-IA8RYE. WBLMI 11

LnAtlw

I
3AMMARAY WATER PERCENT BULK WT
5ATURAT10h 3M.K Wt TAR ANALYSIS
% OPSULKWI
---a&- Ml D 5 Loos ~

.... B CORE o
ATER

I .
HYOROCMBOH

Il?l

moo

I
Fig. 13 — Comparison of S., from lWT soitb S.
from SARABAND, in varksbh. .fresh
‘waters(Well B). -

pagation time) is made and a trend established from the

clean.sand matrix point to sha!e. This method does not re-
quire the definition of the 100-percent shale values, or the
computation of shale volume. tpma is read from ~ma along
the trend of increasing shale content. The bulk-weight an-
. II
alysis is a function of formation porosity, & from EIW
processing, bulk density, and tar density.
Fig. 14 — Exampk of a tar-sand mtdysh compss-
Fig. 14 demonstrates such results together with core tation ssshssr#be Em. Core distiaare also
data for comparison. sbowts oswrtbe cored interval.
12 ELIKX’ROMAGNETICPROPAGATION . . . A NEWDIMENSION IN LOGGING SP 6542

Track 1 Gamma hy — A very good shale in- quire the use of empirical formulas of the Archie type, for
dicator in this area. formation factor and & derivation.
‘h& 2 &~ as computed with porosity and EIW Extensive field testing and interpretation studies have
data, shown that the EPT significantly improves our ability to
Track 3 Percent-bulk-weight tar from core am detect and evaluate hydrocarbon-bearing strata, particularly
alysis and computer processing of log in intervals of variably fresh formation waters.
data.
Track 4 Bulk weight analysis: percent-bulk” ACKNOWLEDGMENT’!’
weight tar and percent-bulk-weight
The field examples used in this paper were made
water.
available by the several oil companies who were involved
The comparison with core analysis is good, as the figure with the wells. Their courtesy and cooperation are grate-
shows. fully acknowledged.
Appreciation is also expressed for ScMumberger’s per-
CONCLUSION mission to publish this paper, and for the invaluable sup-
Measurements from the new Electromagnetic Propaga. port and cooperation of many Schlumberger employees over
tion Tool have been shown to provide a new method for several years. Particularly noteworthy were contributions
the evaluation of water saturation. This technique is rel. of G. Attali, C. Clavier, J. S. Gruss, W. R, Hoyle, J. H.
atively independent of water resistivity, and does not re. Moran and G. Samodai.

REFERENCES
1. Dielectric Materials and Ap/dications. Edited by Von
Hipple, A.R, Published jointly by The Technology
Press of M. LT. and John Wiley and Sons Inc., New
York, 19s4.
2. Calvert, T.J.: Patent, “Microwave Logging Appar?tus
Having Dual Processing Channels”, 11/19/74. United
States Patent No. 3,849,721.
3. Rau, R.N.: Patent, “Method and Ap aratus Utilizing
Microwave Electromagnetic Energy [ or Investigating
Earth Formations”, 3/1 6/76. United States Patent
No. 3,944,91o.
4. Schlumberger Log Interpretation Charts, Schlumberger
Limited, New York, 1972, 1973.
6. Poupon, A., Clavier, C., Dumanoir, J., Gaymard, R.,
and Misk, A.: “Log Analysis of Sand-Shale Sequences
— A Systematic Approach”, /. Pet. Tech,, July, 1970,
7, Sah, R,C., Chase, A.E., and Wells, LX.: “Evaluation
of Alberta Tar Sands”, presented at SPE of AIME
Annual Technical Conference and Exhibition} Houston,
Oct. 6-9,1974. Paper SPE 5034.
8. Aksel’rod, S.M. and Chernyak, G. Ya.: Razvedka i
okhrana nedr, 1959, 28.
9, Poley, J.P.: Patent, “Method of Determining the
Dielectric Properties of Formations Penetrated by a
Borehole”, Australian Patent No, 282,113, assigned
to Shell International Research Mattschappij N.V.,
published Oct. 20,1966.
10. Cihlar, J. and Ulaby, F,T.: “Dielectric Properties of
Soils as a Function of Moisture Content”. Retno:e
Seining Ltbowtory Technical Report 177-47, The
University of Kansas Center for Research, Inc., Nov.,
1974.
SP 6542 THOMAS J. ~LVBRT — RAMA N. RAU — LARRY E. WELLS 13

APPENDD( 1
The EPT measures the travel time and attenuation of formation. If the propagating wave is not a plane wave (a
an electromagnetic wave between two receivers. A relation. spherical wave is generated from a point source, a cylindrical
ship between these measurements and certain formation from a line source) suitable spreading-loss corrections to the
parameters is derived below. measured attenuation (Al ) are made before Eq, A-8 is
Assuming a plane wave varying sinusoidally in time, applied.
the electric field (E) at the second receiver is given by Based on good agreement with experiment in a
heterogeneous medium with fluid-filled pores, the composite
E= Eoe-7z+jd (A-1) travel time in the formation is given by
where EO is the electric field at the first receiver, z is the tPo= f#EPT tpf+ ( 1- 4EPT) tpma
distance between the two receivers, j is the vectorial operator (A-9)
~, o is the angular frequency, t is the time of travel
where tPm~ and tPf are the loss-free propagation times of
of the waves over a distance z in the formation, and y is
the complex propagation factor given by the matrix and fluid respectively. #EPT is the apparent
porosity measured by the EPT tool.
y=a+j~ (A.2) Based on Eq. A-9, Fig. AI-1 shows the variation of tpo
where a is the attenuation factor in nepers/meter, and /3 for various matrix and borehole fluids, For partially water-
filled formations, tpo is determined by Eq, A-8 from the
is the phase @or in radians/meter.
For a “lossless” formation, a = O. From Eq. A-1, the measured values of tP1 (electromagnetic propagation time
from the log) and Al (attenuation from the log), Know-
phase velocity VP. is given by
ing the values of tPf and tP~~, we can then compute an
z al 1 apparent EPT porosity, Since the tpe for hydrocarbon is
vpos —=—=— (A-3)
t yo tpo nearly the same as that of the matrix, the water saturation
of the formation can be determined by taking the ratio of
where the subscript “o” indicates lossless conditions.
apparent porosity of the EFT to the total porosity from
From Maxwell’s equations, it can be shown that
another soutce, such as a neutron. density crossplot,
y.= j.== jdPo (A.4)
CHOICE OF FREQUENCY
where p is the magnetic permeability (Henry/meter),
Since most formations of interest are nonmagnetic, p of Dividing ;q. A-6 by Eq, A-5, we define “loss tangent”
the formation is the same as that of free space (W = 47r c
x 10? Henry/meter), and e is the dielectric permitt ivity =—= —.2a/3 (A-1o)
we /32-c#
(Fwad/meter).
When the formation is 10SSY, y and e are complex. When the loss tangent is less than unity, Eq. A-5 and Eq,
Squaring Eq. A-2 and Eq: A-4 and equating the real A-(i can be solved for n and /3.
and imaginary terms, we have
~2#@=~2-02 (A-5)
(A.6)
a=+~[l-+(:)
2+...,...
]
~WC=2aP
(A-n)
where C is the equivalent conductivity (mhos/meter) of
the losses in the formation.
Dividing Eq. A-5 by az,
‘=”@[’ ++($- 1
we = —-—
Pz az
(A-7) As frequency increases, the phase factor p becomes
02 02
less depende~t on conductivity. H~wever, if the frequency
From Eq. A-4 we = tP#. Since #/ti( = tPl) is the travel exceeds 1 gHz, ”other losses (dipolar relaxation) come into
time in the Iossy medium, play. Based on these considerations, a frequency of I. I
gHz was chosen. For water conductivities up to 1 mho, @
(A-8) computed from Eq. A-9 is independent of the salinity of the
water. For higher water conductivities, the effect of salinity
Remembering that a is the attenuation factor, Eq. A-8 im- on tpr is corrected with experimentally determined curves.
plies that the actual propagation time in a conductive forma- Similarly, the effect of temperature on tPf is also corrected
tion is longer than that of a corresponding loss-free by means of experimental data.
14 ELWROMAGNETIC PROPAGATION . . . ANEW DIMBNSION IN LOGGING SP 6542

Interpretation Chart

I
49
E PT
45
based on:
40
20 36 tp. =’fhpf +H)tpm.
33
[ 30
10070
H@

Shale
r= / -—-. ~ 100% oil
LS
--- ~---
NOCI
“--- —
Ss

52 u 100?40G08 Oi I
or Air
I Air

/
01 1 I t t i 1 i I
o 10 20 30 40 —

Porosity, p.ue
Fig. AI.1 — Tet#ative interpretation chart.
SP 6542 THOMAS J. CALVE~ — ~MA N. MU — LARRYE. WELLS 15

APPENDIX II
Moro Promising Applications: ● Good pad.to-wall contact is rquired. Signals un-
Analysis of freshwater/heavy oil zones. reliable in mgose borehoks.
●

Evaluation of &. or S= in new wells drilled in ● Mud cake up to 78 inch ean be tolerated.
●
secondary recovcq projects. (Formation water sal- ● Hole-size limits for the experimental tools arc
inities are normally unknown because of variable 6% to 11 inches,
contamination by fled waters.)
● Best rdts have been obtained in sand formations.

Characteristics and limitations of Prosont Tool: ● Temperature limit of the experimental tools is
210°F.
● Tool is designed for fresh-mud applications. Signal
levels may be too low in salt muds, and signals are ● The Em quipment is available only in limited
not reliable in oil-or gas-filledboreholes. areas.