Академический Документы
Профессиональный Документы
Культура Документы
P 4 3 8 4 4 3 , 2 FIGURES, 2 TABLES
Zhi-Qiang Maol
I ABSTRACT I
The density and neutron logs are two important physi- different for gas-bearing zones and oil-bearing zones. The
cal measurements in formation evaluation. It was found correlation coefficient can be extended to an approach for
through theoretical derivation that the density log and the identifying natural gas formations since it is more sensi-
neutron response physically relate to each other in sedi- tive to gas.
mentary formations because of the existence of the ele- The example of one well from Tarim basin in north-
ment hydrogen. It is the physical dependence of the two west China demonstrates that an excellent correlation
logs that constitutes the usefulness of these two logs in between density and neutron logs does exist in log meas-
determining the porosity, identifying lithology of the for- urements. Thus, the physical dependence of those two
mation and detecting natural gas formation as widely used logs is shown not only by theoretical derivation but also
today. by field logs. The correlation polarity approach indicates a
Based on the dependence, a correlation coefficient gadoil contact more accurately than the conventional
between density porosity and neutron porosity has been overlay technique. This is verified with the RFT pressure
defined. The polarity (i.e, the sign) of the coefficient is data from the well.
I I
TABLE1 TABLE2
Z
Element A Z 2- Actual 2 C n 8 z
A
Compound Formula Density
H 1.008 1 1.9841
C 12.011 6 0.9991 Quartz Si02 2.654 0.9985
0 16.000 8 1.oooo Calcite CaC03 2.710 0.9991
Na 22.99 11 0.9569 Dolomite CaC03MgC03 2.870 0.9977
Mg 24.32 12 0.9868 Anhydrite CaS04 2.960 0.9990
A1 26.98 13 0.9637 Sylvite KCL 1.984 0.9657
Si 28.09 14 0.9968 Halite NaCL 2.165 0.9581
S 32.07 16 0.9978 Gypsum CaS042H20 2.320 1.0222
c1 35.46 17 0.9588 Klaolinite’ A14Si4010(OH)8 2.420 1.0102
K 39.10 19 0.9719 Mg-smectite’ MgA114si22060(OH)1~ 2.469 0.998 1
Ca 40.08 20 0.9980 Illite .~)020(OH)4 2.530
KI-I.sA~~(S~,-~.SA~I.I 0.9965
Fe 55.85 26 0.9980 Fresh Water H20 1.ooo 1.1101
Salt Water 200,000ppm 1.146 1.0797
“Oil” n(CH2) 0.850 1.1407
Methane CH4 1.247
number of the ith atom, and k is the total number of atoms “Gas” cl 1H4.2 1.238
per molecule.
‘Serra, 1984
For most of the elements and compounds found in sedi- ’Hepler et al., 1992
mentary formations, the factors apart from pb in equation
(1) and equation (2) very nearly evaluate to one (as shown
in Table 1 and Table 2).
In contrast, the neutron log measures the slowing-down In fact, the element hydrogen does exist in sedimentary for-
of thermal neutrons, mostly by the hydrogen in the forma- mations and the atomic weight of hydrogen (1.008) is very
tion. Thus, the neutron reading depends mainly on the close to unity, i.e. Ah = 2 h = 1 . So, we get
hydrogen index, HI, of the formation, which is proportional
to the quantity of hydrogen per unit volume
(3)
where n/,is the number of hydrogen atoms in the compound,
Mis the molecular weight, and c is a constant.
By convention, the hydrogen index of fresh water at
room conditions is taken as unity, so the constant c equals 9. where the last line employs
equation (4), we find
E. k
,=I
n, A; / M = 1 . Combining
Equation (3) can be thus rewritten as
(4)
1=I
and, / .
k
=
l A
I=)
POTENTIAL APPLICATION IN NATURAL GAS
DETECTION where k is the number of sampling points that represent a
segment of log curves used to correlate. The continuous cal-
The above analysis demonstrates that the physical
culation of the correlation coefficient, R , is similar to what
dependence of density and neutron logs facilitates their
we do in correlating any two correlation curves of dipmeter
combined application. The analysis of the correlation char-
measurements (no “search length” is involved, correlating
acteristics of the log responses also shows that a correlation
the points or segment of log curves with same depth). The
approach could be used to identify gas formations, perhaps
sign of R is termed the correlation polarity.
more easily than with the conventional overlay technique.
To understand the correlation characteristic of density
In a shaly hydrocarbon-bearing sandstone, the response
and neutron logs calculated by equation (1 1 ) and to extend it
equations of the density and neutron logs are respectively
to gas identification, the following formations with various
given by lithological and fluid features are qualitatively discussed.
p h I I ‘ , +@(l-Sh)p~t+ ~ P , / (7)
( ~ - @ - ~ I ) ~ I+@S/jph I I
Pmr - P =
It
1 (9)
Q N = @-@s/lL
= @.
For crude oil with normal gravity, those two logs respond
principally to the porosity of the formation in respect that
@ + @S/,bh + K/I
\/I
+ K/I &I
tion will become complicated in that the quantity might be resistivity log responses, the interval from A to E can be
small because of deep invasion, low porosity or low gas interpreted as hydrocarbon-bearing formation, and the
saturation. The correlation polarity method still works as oiVwater contact is considered at the level E. The gadoil
long as the response to residual gas exists whether it is great interface is not clear.
or small. In contrast the overlay technique only works when The correlation coefficient, R, calculated by equation
the response of gas is significant enough to cause a distinct (1 1) in which K = 7 is plotted in track 5 in Figure 1 (to
difference between density porosity and neutron porosity. emphasize the results of reservoir, R is artificially set to zero
In other words, the correlation polarity is more sensitive in shale intervals). 2 is the ratio of density porosity over
than the overlay technique to a small gas response. This rea-
soning also applies to the shaly formation discussed below.
@N=$-$shcsh +Khcsh = $ + K h c s h .
For dispersed or structural claylshale, Psh is often less
thanp,,. The values of @D and @Ndependupon whether the
effect of gas is greater than that of shale or not. If the effect
of shale predominates (i.e., KhdsIl > $Shdl, and V,/,&,, >
$ s / l c h ) and consequentlys
h is low, the two logs will respond
as if to a shale interval. The correlation method is still valid
but the overlay technique fails when the response to gas is
larger than that to shale. However, the superiority is not as
significant as when the overlay of the two logs exhibits a
distinct offset. This situation is frequently found in shaly
gas reservoirs because the response to gas might be counter-
acted by the response to shale and further reduced by a
lower gas saturation (caused by deep invasion, higher shale
content or low porosity).
Subscripts:
nu7 rock matrix
Sh shale
W formation water
h hydrocarbon
D density log
N neutron log
REFERENCES
Gayniard, R., and Poupon, A,, 1968, Response of neutron and for-
mation density logs in hydrocarbon bearing formations: The
Log A i d y s t , vol. 9, no. 5, p. 3-13.
Hepler, L. G. et al, 1992, AOSTRA Technical Handbook of Oil
FIG.2 The fluid pressure profile from the RFT measurements Sands. Bitumens and Heavy Oils, Chinese Version (translated
of the well. by Liang et al.), Press of University of Petroleum, Dongying.
Schlumberger Ltd., 1976, Log Interpretation, Volume I: Princi- ABOUT THE AUTHOR
ples.
Segesman, F., and Liu, O., 1971, The excavation effect, paper N, in Zhi-Qiang Mao is an associate professor of applied geophys-
12th Annual Logging Symposium Transactions: Society of ics with the University of Petroleum Beijing, China. He received
Professional Well Log Analysts. his doctors degree in applied geophysics from the postgraduate
Serra, O., 1984, Fundamentals of well-log interpretation, 1: The school of the Research Institute of Petroleum Exploration and
Acquisition of Logging Data, Elsevier. Development of Beijing. He has worked for China National Petro-
Tan, T. D., 1991, Log interpretation of deep gas zone by nuclear leum Cooperation (CNPC) in Tarim Petroleum E & D Headquar-
logging porosity difference method" (in Chinese): Petroleum ters as a log analyst from 1993 to 1997. As a visiting scholar, he
Exploration and Development, vo1.18, no. 6. had studied and worked shortly in MIT, USA, Institut Francais du
Petrole (IFP), and Gaz de France (GDF), France. He is in charge of
two key research projects of CNPC. His research interests include
log interpretation, petrophysics, and formation evaluation.