PETROLEUM EXPLORATION AND DEVELOPMENT

Volume 43, Issue 2, April 2016

Online English edition of the Chinese language journal

Cite this article as: PETROL. EXPLOR. DEVELOP., 2016, 43(2): 268–276. RESEARCH PAPER

Water spectrum method of NMR logging for identifying

fluids
HU Falong1, 2,*, ZHOU Cancan1, 2, LI Chaoliu1, 2, XU Hongjun1, 2, ZHOU Fengming3, SI Zhaowei3
1. PetroChina Research Institute of Petroleum Exploration & Development, Beijing 100083, China;
2. PetroChina Well Logging Key Lab, Beijing 100083, China;
3. PetroChina Jidong Oilfield Company, Tangshan 063004, China

Abstract: A new fluid identification method by constructing water spectrum based on NMR logging was put forward after the limita-
tions of existing nuclear magnetic resonance (NMR) fluid identification methods were analyzed. At present, differential spectrum method
(DSM) and shifted spectrum method (SSM) of NMR logging are commonly used fluid identification methods. Due to the effects of fluid
properties and pore structures, however, their coincidence rates of fluid identification are lower. A new fluid identification method named
water spectrum construction method was developed in this study. Based on the existing acquisition mode of NMR logging, T2 (transverse
relaxation time) spectrum of long waiting time and long echo spacing in completely watered conditions was constructed from the T2 spec-
trum which was measured in the mode of long waiting time and short echo spacing. And then, the types of fluids in reservoirs were iden-
tified by comparing the measured T2 spectrum with the constructed water spectrum. This new method was applied in Nanpu sag, Bohai
Bay Basin for identifying oil layers, oil-water layers, water layers, gas layers and low-resistivity oil layers. It is demonstrated that based
on the water spectrum construction method, the coincidence rate of fluid identification caused by pore structures is increased and fluid
identification capacity of NMR logging is improved. Water spectrum construction method is prospective for fluid identification and eval-
uation of complex reservoirs.

Key words: NMR logging; water spectrum method; fluid identification technique; complex reservoirs; fluid identification coinci-
dence rate

Introduction processing software and methods of explanation for identify-

Nuclear magnetic resonance (NMR) logging, which pro-
vides significant information of reservoirs and fluids, such as
porosity, permeability, irreducible water saturation, is an im-
portant non-electric method to identify fluid property[1]. Dif-
ferential Spectrum Method (DSM) and Shifted Spectrum
Method (SSM) of NMR logging are common methods to
identify fluids, but they have certain limitation because of
effect produced by both pore structure and fluid type, result-
ing in that NMR logging has low success rate in fluid identi-
fication[12]. The latest two-dimensional (2-D) NMR log-
ging can obtain transverse relaxation time (T2), diffusion coef-
ficient of fluid (D) and longitudinal relaxation time (T1). (T2,
D) and (T2, T1) 2-D NMR logging can counteract the effects of
pore structure on fluid identification, thus improving the ca-
pability to identify fluids. However, many conditions needed
to realize the 2-D NMR logging, such as designing new pat-
terns of NMR logging, researching 2-D inversion algorithm,
ing fluids. There are a lot of difficulties in its realization and
application. In the paper, a new method is proposed with cur-
rent tools and no need to design a new pulse sequence. The T2
spectrum (water spectrum) of long wait time (TW) and long
echo spacing (TE) has been constructed by the data of long TW
and short TE obtained by current acquisition mode of NMR
logging. By contrasting the logging T2 spectrum and water
spectrum of long echo spacing, fluid properties can be identi-
fied. The method can’t be affected by complex pore structure,
and strength the identifying ability of formation fluids. Water
Spectrum Method is an important method for fluid identifica-
tion and evaluation of complex reservoirs.
1. Limitations of current methods to identify
fluids in NMR logging
NMR logging makes use of the CPMG (Carr, Purcell,
Meiboom, Gill) pulse sequences to acquire the transverse
relaxation time (T2) of pore fluids. The NMR transverse re-

Received date: 20 Jan. 2015; Revised date: 10 Jan. 2016.

* Corresponding author. E-mail: hufalong@petrochina.com.cn
Foundation item: Supported by the China National Science and Technology Major Project (2011ZX05020-008); CNPC Basic Logging Research Project
(2014A-3910).
Copyright © 2016, Research Institute of Petroleum Exploration and Development, PetroChina. Published by Elsevier BV. All rights reserved.
 HU Falong et al. / Petroleum Exploration and Development, 2016, 43(2): 268–276

laxation time (T2) of pore fluid is mainly affected by the bulk Under the condition of long wait time and long echo spac-
relaxation, surface relaxation and diffusion relaxation[1], it can ing, the transverse relaxation time (T2,wl) can be written as:
be expressed as: 1 1 S D ( GTEl ) 2
  2  w (6)
1 1 S D( GTE ) 2 T2 ,wl T2B ,w V 12
  2  (1)
T2 T2B V 12 As for water, gas, light oil, and the medium viscosity oil,
According to the equation mentioned above, T2 is affected the bulk relaxation time is longer than the surface relaxation
by the pore structure (S/V) and fluid property (D). This effect time that is usually a few milliseconds, and its value is about a
not only results in inaccurate evaluation of pore structure, but few seconds and can be usually neglected. For the common
also decreases the ability to identify fluids. For the common MRIL-Prime tool, if the diffusion coefficient of water (Dw) is
DSM, when there exists big pores in water layer, water can’t (2.56.0)×108 cm2/ms and the short echo spacing (TEs) is 0.9
be completely magnetized, which results in the wrong expla- 12
ms, the diffusion relaxation time ( ) is 1 193.8
nation of oil due to the differential spectrums. DSM can only Dw ( GTEs )
be used for light oil identification. SSM identifies fluid types 2 865.1 ms; if the long echo spacing(TEl) is 3.6 ms, the diffu-
by D and mainly used for crude oil of medium viscosity. Tak- sion relaxation time is 74.6179.1 ms. If the echo spacing is
ing the DIFAN module as an example in Haliburton, it evalu- short, the contribution of diffusion relaxation time to the total
ates oil saturation by means of cross-plot. The module needs relaxation time can be neglected compared with the surface
abundant parameters of reservoirs and fluids, such as viscosity relaxation time. The diffusion relaxation time of gas, light oil
of hydrocarbon and water, temperature, stress. Lizhi Xiao[12] and medium viscosity oil (usually a few seconds) is relatively
ever pointed that the algorithm of the module didn’t make long if the echo spacing is short, which is similar to the water.
sense theoretically. The 2-D NMR logging can separate signals It can be neglected compared with the surface relaxation time
from oil, gas and water. The process of the 2-D NMR logging is of sandstone.
relatively complicated, in which a new acquisition mode, for Under the condition of short echo spacing, the transverse
example, the Diffusion Editing pulse sequence, algorithms for relaxation time of water can be shown below:
inversion, method of fluid identification should be designed. 1 1 S D ( GTEs ) 2 S
  2  w  2 (7)
2. Principle of water spectrum method in NMR T2,ws T2B,w V 12 V
logging If the echo spacing is long, it can be expressed as:
1 1 S D ( GTEl ) 2
Eliminating the effect of pore structure plays a pivotal role   2  w 
in fluid identification. The S/V is determined by the reservoir T2 ,wl T2B ,w V 12
and the D has a relationship with fluid properties in pores. The S Dw ( GTEl ) 2
 2 (8)
D is obtained more easily than the S/V and acquired accurately V 12
in laboratory. For oil, its diffusion coefficient (Do) bears on Under the short echo spacing, transverse relaxation time of
viscosity and temperature[1, 37], which can be written as: gas can be written as:
C (T  273.15) 1 1 S D ( GTEs )
2
S
Do  (2)   2  g  2 (9)
298 T2,gs T2B,g V 12 V
The water diffusion coefficient (Dw) is only related to tem- If the echo spacing is long, it is given by:
perature and can be given by:
S D ( GTEl )
2
1 1
Dw  1.041 3  0.039 28 T  0.000 403 18 T 2 (3)   2  g 
T2,gl T2B,g V 12
The gas diffusion coefficient (Dg) has something to do with
S Dg ( GTEl )
2
density (  ) and temperature (T) and can be expressed as: 2
 (10)
V 12
(T  273.15)0.9
Dg  0.085 (4) Under the condition of short echo spacing, transverse re-

laxation time of light oil and medium viscosity oil can be ob-
The relationship between three diffusion coefficients tained by:
is Dg  Dw  Do , and they distribute in different or- 1 1 S D ( GTEs ) 2 S
ders of magnitude. The composition of water molecules is   2  o  2 (11)
T2,os T2B,o V 12 V
relatively single, so its diffusion coefficient can be obtained
If the echo spacing is long, it is written as:
precisely by core analysis instrument in laboratory.
1 1 S D ( GTEl ) 2 S D ( GTEl ) 2
When porous medium is filled with water completely, and   2  o  2  o (12)
in the mode of long wait time (for example, TWl is 12.988 s) T2,ol T2B,o V 12 V 12
and short echo spacing (for example, TEs is 0.9 ms), the trans- The amplitudes of the decaying spin echoes under the long
verse relaxation time (T2,ws) can be expressed as follows: wait time and long echo spacing is given by:
 S D ( GTEl )2 
1 1 S D ( GTEs ) 2 t  2 i  
  2  w   M ie  Vi 12 
(5) MT (13)
T2 ,ws T2B ,w V 12 El ,TWl
i

 269 
 HU Falong et al. / Petroleum Exploration and Development, 2016, 43(2): 268–276

 S D ( GTEl )2 
Contrasting these equations from Eqs.7 to Eqs.12, only the t  2 i  o 
M t   M i e  Vi 12 
diffusion relaxation changes under the condition of long wait  t 
i
time (for example, TWl is 12.988s) and long echo spacing (for
 S D ( GTEl )2 
 t  2 i  w 
example, TEl is 3.6 ms), compared with the long wait time and
M e  Vi 12 
i (18)
short echo spacing (for example, TEs is 0.9 ms). Hence the T2 i
spectrum of long wait time and short echo spacing can be Now the amplitude of first echo train (∆Ml) among the echo
used for simulating the echoes of long wait time and echo trains (∆Mt) is more than zero and the standard deviation of
spacing. Integration the result of long wait time and short ∆Mt (σ（∆Mt）) is more than δ. Inverting ∆Mt can get the
echo spacing into Eqs.13 yields echoes of constructed water differential spectrum result, the 100% oil-saturated T2 spec-
spectrum as follows: trum (∆T2,o) which is called oil spectrum for short below. If
 1 D ( GTEl ) 2 
t   w  the reservoir is saturated with heavy oil, bulk relaxation can
  M ie  T2,ws,i 12 
MT ,TWl
 be neglected. Manipulating the measured echo trains and data
El
i

 S D ( GTEl ) 2  of water spectrum by DSM can obtain:

 t  2 i  w   S D ( GTEl ) 2 1 
M e  Vi 12   t  2  o  
(14)
M t   M i e  V
i 12 T2 B ,o 
i
i

i
Manipulating the echo trains of long wait time and long  S D ( GTEl ) 2 
 t  2  w 
echo spacing and data of water spectrum by DSM can get:  t   M ie
 V 12  i
(19)
 S D ( GTEl )2 
 t  2 i   i

M t   M i e  Vi 12 
 t  For heavy oil, if using the D9TWE3 mode for light oil, wa-
i
 S D ( GTEl )2 
ter spectrum and measured spectrum have little differences,
 t  2 i  w 
which is difficult to identify fluids. The longer echo spacing
M e  Vi 12 
i (15)
i should be used to acquire the logging data which cooperate
The ∆M is related to the properties of fluids in reservoirs with resistivity to increase the success rate for fluid identifica-
and inverted into the T2 spectrum (∆T2) by multiple index tion.
inversion algorithm, which is exactly the T2 spectrum of the Inverting the differences of echo trains above into T2 spec-
fluid to be recognized. trum by NMR data inversion algorithm. Oil spectrum and gas
If the reservoir is the water layer, manipulating the meas- spectrum can be confirmed by means of the size of difference.
ured echo trains and data of water spectrum by DSM can To calculate and evaluate the fluid properties quantitatively in
get: reservoirs, oil saturation and gas saturation can be calculated
 S D ( GTEl ) 2 
t  2 i  w 
using the following formulas:
M t   M i e  Vi
So   T2 ,i / I H,o NMR
12 
 t  (20)
i i
 S D ( GTEl ) 2 
t  2 i  w  Sg   T2 ,i / I H,g NMR (21)
M e  Vi 12 
i  t (16) i
i
In the practical application of logging data, bringing the
The different spectrum signal by multiple index inversion
hydrogen index of oil and gas measured in lab into the Eqs.20
algorithm can be expressed as: ∆T2,w=0.
and Eqs.21 so that the calculated saturation is closer to the
If it is a gas layer, because the gas diffusion coefficient is
true value. Oil saturation by water spectrum method is appar-
much greater than that of water, the measured amplitudes of
ent saturation and is less than the true oil saturation in reser-
echo trains is less than the that of water spectrum. Manipulat-
voir. In the division of oil layers into levels by oil saturation,
ing the measured echo trains and data of water spectrum by
combining the well test data can get more accurate outcome.
DSM can get:
Meanwhile the apparent oil saturations by water spectrum of
 S Dg ( GTEl ) 2 
 t  2 i  
different echo spacing are diverse.
ΔM t   M i e
 Vi 12 
 
 t 
i
3. Applications
 S D ( GTEl )2 
 t  2 i  w 

M e  Vi Rivers, delta, subaqueous fans and many other sedimentary

12 
i (17)
i facies are developed in Nanpu Sag of Bohai Bay Basin (Fig.
Inverting the ∆Mt into the 100% gas-saturated T2 spectrum 1). Sedimentary facies belts are overlaid and the reservoirs in
(∆T2,g) which is called gas spectrum for short below by this area are characterized by the dramatic change of the li-
Multi-Exponentials inversion algorithm. thology and physical property. And due to the existence of a
If the fluid in reservoir is light oil or medium viscosity oil, large number of oil reservoirs with low electrical resistivity, it
for commonly used D9TWE3 mode to measure light oil, ma- is pretty difficult to utilize the conventional well logging
nipulating the measured echo trains and data of water spec- methods for fluid identification[8]. On the basis of the theory
trum by DSM can get: research mentioned above, the authors apply MRIL-Prime to
 270 
 HU Falong et al. / Petroleum Exploration and Development, 2016, 43(2): 268–276
Fig. 1. the geologic structure and location in studied area.
verify the water spectrum fluid identification method in this
sag. Under the condition of the reservoir physical property
condition, the crude oil is identified as light oil, its average
viscosity is 4 mPa·s. According to the design requirements,
the mode of the D9TWE3 well logging is performed, which
can collect five groups of the data from the two wait times
mode, two echo spacings mode and clay irreducible water
mode, and identify the fluid using two wait times difference
spectrum method and two echo spacings shifted spectrum.
Specific parameters are given as follows, Group A:
TW1=12.988 s, TEs=0.9 ms, echo number is 500; Group B:
TWs=1 s, TEs=0.9 ms, echo number is 500; Group C: TWc=0.02s,
Fig. 2. Identification of water layer by water spectrum method in the depth of 2 7682 787 m in Well A of Nanpu Sag.
 271 
TEc=0.6 ms, echo number is 20; Group D: TW1=12.988 s,
TE1=3.6 ms, echo number is 125; Group E: TWs=1 s, TEl=3.6
ms, echo number is 125.
3.1. Water zone identification
The average porosity of Well A in the depth of 2 769.6
2 785.3 m in Nanpu Sag is 20.5%, and it’s proven to be water
by the oil production test and the water production in a day is
7.4 m3. According to the principle of the water spectrum es-
tablishment, the differential spectrum signal cannot appear in
the water zone, which means the measured spectrum basically
has the same distribution range and amplitude with estab-
lished spectrum. By the contrast of the 7th measured spectrum
(T2 spectrum of the long echo interval) and the 8th established
water spectrum (Fig. 2), it’s found that the difference signal
cannot be generated by the measured spectrum and the estab-
lished water spectrum, which means T  0 and demon-
strates the validity of the established water spectrum method
and establish a solid theoretical foundation of its application
in oil and gas reservoirs.
3.2. Oil layer identification
The average porosity of Well B in the depth of 2 955.7
2 971.4 m in Nanpu Sag is 24.1%, and with comparison of the
7th measured spectrum (T2 spectrum of the long echo interval)
and the 8th established water spectrum (Fig. 3), there exists a
obvious difference between the established water spectrum and
the measured spectrum, and the measured spectrum has a small-
er shift to the left than the established water spectrum, which
illustrates that the diffusion coefficient of the fluid in this
 HU Falong et al. / Petroleum Exploration and Development, 2016, 43(2): 268–276
Fig. 3. Identification of oil layer by water spectrum method in the depth of 2 952.52 975.0 m in Well B of Nanpu Sag.
reservoir is smaller than water. The difference spectrum of the
established water spectrum and the measured spectrum is
considered as the oil spectrum, and the apparent oil saturation
of this reservoir is 19.2% by calculation. By the criterion of
oil classification in this area of the established water spectrum
method ( So >15% is classified as oil), it’s identified as oil and
the oil production test indicates that the crude oil production
reaches 99 t in a day in the initial stage, which agrees well
with the identification results of the established water spec-
trum method.
3.3. Oil-water zone identification
Fig. 4 shows the application of the established water spec-
trum method in the oil-water reservoir and the average poros-
ity of Well C in the depth of 2 628.12 652.5 m in Nanpu Sag
is 24.6%. Compared with the 7th measured spectrum (T2 spec-
trum of the long echo spacing) and the 8th established water
spectrum, in the upper depth (2 628.12 638.1 m), there is an
obvious difference between the measured spectrum and the
established water spectrum, the shift in the measured spec-
trum is smaller than the established water spectrum, which
means the diffusion coefficient of the fluid in this reservoir is
smaller than water, therefore, the difference spectrum is con-
sidered as oil spectrum. In the bottom of the reservoir
(2 638.12 652.5 m), the established water spectrum is in
accordance with the measured spectrum, and the fluid is iden-
 272 
tified as water. After the analysis on the established water
spectrum method in the upper and bottom depth, it can be
confirmed as oil-water reservoir, and compared with the elec-
trical resistivity curve, the oil-water interface is displayed
more clearly by the established water spectrum method. The
oil production test shows that, the oil and water production in
this reservoir are 7.5 t and 55 m3 respectively in a day, which
is in agreement with the fluid recognition results by the estab-
lished water spectrum method.
3.4. Gas zone identification
Fig. 5 shows the application of the established water spec-
trum method in the gas reservoir and the average porosity of
Well D in the depth of 2 382.82 385.4 m in Nanpu Sag is
13.7%. Compared with the 7th measured spectrum (T2 spec-
trum of the long echo interval) and the 8th established water
spectrum, there is a obvious difference between the measured
spectrum and the established water spectrum, the shift in the
measured spectrum is obviously bigger than the established
water spectrum, and σ(∆Mt）is larger than δ, which means the
diffusion coefficient of the fluid in this reservoir is obviously
larger than water, therefore, the difference spectrum is con-
sidered as gas spectrum. And it’s proven by the oil production
test that, the gas production in this reservoir is 4.5×104 m3 in a
day, which is in agreement with the fluid recognition results
by the established water spectrum method. Notably, for gas
 HU Falong et al. / Petroleum Exploration and Development, 2016, 43(2): 268–276

Fig. 4. Identification of the oil-water layer by water spectrum method in the depth of 2 6252 655 m in Well C of Nanpu Sag.

recognition, when the gas saturation is relatively low, the es-
tablished water spectrum method cannot provide a satisfying
result, hence, data of other well logging methods is required to
improve the precision of the fluid recognition.
3.5. Identification of low- resistivity oil layer
Low-resistivity oil reservoir is defined as the oil reservoir
with the electrical resistivity ratio of the oil reservoir and wa-
ter reservoir, which is smaller than 2 in the same oil and water
system. The factors resulting in the low electrical resistivity in
the oil reservoir mainly include the high content of the irre-
ducible water, conductive clay, the great difference of the
mineralization degree and so on. It can be seen from Fig. 6,
the average electrical resistivity of Well E in the depth of
3 055.13 063.9 m in Nanpu Sag is 3 Ω·m, which greatly re-
stricts application of the fluid evaluation using Archie's law.
The calculated result of the 8th established water spectrum
indicates there exists strong oil signal and the calculated ap-
 273 
parent oil saturation in this reservoir is larger than 15.3%,
which can interpret this reservoir as oil reservoir. In this depth
of the reservoir, the irreducible water saturation is relatively
high, which is the main cause of the low electrical resistivity.
The oil production test has confirmed that, the crude oil pro-
duction in this reservoir is 57.5 t in a day and the validity of
the established water spectrum result by NMR well logging
method has also been verified successfully.
3.6. Comparison between Water Spectrum Method and
DSM
The well F in Nanpu Sag in the depth of 3 210.83 212.6 m
has relatively strong differential spectrum signal so that the
layer was judged as an oil layer. However the test confirms
that it is a water layer and its water production is 12.2 m3 at
the beginning. If using the water spectrum method, the water
spectrum can be got under the condition that the echo spacing
is 3.6 ms. As shown in the 10th trace of Fig. 7, the water
 HU Falong et al. / Petroleum Exploration and Development, 2016, 43(2): 268–276

Fig. 5. Identification of gas layer by the water spectrum method in the depth of 2 3702 387 m in Well D of Nanpu Sag.

Fig. 6. Identification of the low-resistivity oil layer by water spectrum method in the depth of 3 0463 066 m in Well E of Nanpu Sag.

 274 
 HU Falong et al. / Petroleum Exploration and Development, 2016, 43(2): 268–276

Fig. 7. Comparison of the application result with water spectrum and differential spectrum methods in the depth of 3 2043 216 m in
Well F of Nanpu Sag.

spectrum is almost consistent with the measured spectrum (the
9th trace). The fluid in the reservoir has the same diffusion
coefficient with water and is judged as water, which agrees
well with the oil test data. When long and short wait time are
applied to NMR logging for the water layer with big pores,
how much the fluid is polarized is the basis of fluid identifica-
tion by DSM. The water is not fully polarized for short wait
time measurement and that is why the DSM is not effective. It
is not by the wait time to identify the fluid for the water spec-
trum but by the diffusion coefficient of different fluids which
obviously increases the success rate to fluid identification.
4. Conclusions
Water spectrum method in NMR logging is a new kind of
technology to identify fluids. It combined diffusion coefficient
with measured T2 spectrum to establish the water spectrum,
which can eliminate the influence of pore structure on the
identification of fluid properties. By contrasting the difference
between measured spectrum and constructed water spectrum,
the identification of oil, gas and water can be conducted ac-
curately. Its can improve the ability of NMR logging for fluid
identification that has been applied to 120 wells in Bohai Bay
Basin. It advantages include that it can eliminate the effect of
 275 
pore structure while DSM can’t when recognizing light oil
reservoir[910]; and increase success rate of identification fluid;
provide accurate apparent fluid saturation; does not need input
the characteristic values of T2 spectrum (such as peak value,
the half of peak value or the geometric mean value) as the
SSM, which is difficult to recognize the fluid properties; it can
identify fluid properties by multiple components of T2 spec-
trum quantificationally; it can be realized more easily com-
pared with 2-D NMR logging, without needing the change of
well logging mode; it adopts water diffusion coefficient in
free state and works effectively to identify fluids. However, in
tight reservoirs, the method has relatively low success rate due
to effect of diffusion factors, and the diffusion coefficient of
formation water that obtained from rock petrophysics experi-
ment must be utilized to identify the fluids. So, it has signifi-
cant advantages to identify the complex reservoirs such as
low-resistivity oil layer formations, and increases accuracy in
identifying fluids in complex reservoirs combined with resistiv-
ity logging.
Nomenclature
C—constant, usually 1.4, dimensionless;
 HU Falong et al. / Petroleum Exploration and Development, 2016, 43(2): 268–276

D—fluid diffusion coefficient, cm2/ms; μ—viscosity, mPa·s;

Dg—diffusion coefficient of gas, 108 cm2/ms; ∆Mt—echo signal by DSM, %;
Do—diffusion coefficient of oil, cm2/ms; ∆T2,g—echo signal of gas by DSM, ms;
Dw—diffusion coefficient of water, cm2/ms; ∆T2,o—echo signal of oil by DSM, ms;
G—magnetic field gradient, Gs/cm; ∆T2,w—echo signal of water by DSM, ms;
I—formation resistivity index, dimensionless; σ—standard deviation, dimensionless.
IH,g—hydrogen index of gas, dimensionless;
IH,o—hydrogen index of oil, dimensionless; References
i—T2 component, dimensionless;
Mi—porosity of ith porosity components, %; [1] XIAO Lizhi, Nuclear magnetic resonance logging and core
MTEl, TWl—echo of long wait time and echo spacing, dimensionless; magnetic resonance imaging and its application. Beijing: Sci-
S—pore superficial area, cm2; ence Press, 1998.
Si—pore superficial area of ith pore components, cm2; [2] XIAO Lizhi. Some important issues for NMR logging
So—apparent oil saturation, %; applications in China. Well Logging Techology, 2007, 31(5):
Sg—apparent gas saturation, %; 401–407.
t—time of echo rains, dimensionless;
[3] HE Zongbin, NI Jing, WU Dong, et al. Hydrocarbon satura-
T—temperature, C;
tion determined by dual TE logging. Lithologic Reservoirs,
T2—transverse relaxation time, ms;
2007, 19(3): 89–92.
T2B—bulk relaxation time, ms;
[4] XIE Ranhong, XIAO Lizhi, DENG Kejun, et al.
T2B,g—bulk relaxation time of gas, ms;
Two-dimensional NMR logging. Well Logging Techology,
T2B,o—bulk relaxation time of oil, ms;
2005, 29(5): 430–434.
T2B,w—bulk relaxation time of water, ms;
[5] TAN Maojin, ZOU Youlong, LIU Bingkai, et al. Inversion
T2,gl—gas T2 of long echo spacing, ms;
simulation of (T2, D) 2-D NMR logging and analysis of
T2,gs—gas T2 of short echo spacing, ms;
observation parameters effects in gas-water model. Well
T2,ol—oil T2 of long echo spacing, ms;
Logging Techology, 2011, 35(2): 130–136.
T2,os—oil T2 of short echo spacing, ms;
[6] HU Falong, ZHOU Cancan, LI Chaoliu, et al. Fluid identifica-
T2,wl—water T2 of long echo spacing, s;
tion method based on 2-D diffusion-relaxation nuclear mag-
T2,ws—water T2 of short echo spacing, ms;
T2,ws,i—T2 of water in ith pore component, ms; netic resonance (NMR). Petroleum Exploration and Develop-
TE—echo spacing, ms; ment, 2012, 39(5): 552–558.
TEc—echo spacing of clay bound water, ms; [7] OUYANG Jian, MAO Zhiqiang, XIU Lijun, et al. Formation
TEl—long echo spacing, ms; mechanism and evaluation method of low contrast reservoir in
TEs—short echo spacing, ms; well logging. Beijing: Petroleum Industry Press. 2009.
TWc—wait time of clay bound water, s; [8] ZHOU Fengming, SI Zhaowei, MA Yuejiao, et al. Integrated
TWl—long wait time, s; identification method for low-resistivity hydrocarbon layers in
TWs—short wait time, s; Nanpu Sag. Petroleum Exploration and Development, 2008,
V—pore volume, cm3; 35(6): 680–684.
Vi—volume of ith pore component, cm3; [9] WANG Zhongdong, WANG Hao, Xiang Tiande. Integration of
ρ—density, g/cm3; NMR DTW Logs and DTE Logs to improve the oilbed inter-
ρ2—transverse surface diffusion coefficient, cm/ms; pretation accuracy. Well Logging Technology, 2001, 25(5):
t—random noise, dimensionless; 365–368.
—variance of random noise, dimensionless; [10] HU Falong, XIAO Lizhi. A method for determining NMR log
γ—gyromagnetic ratio of H, (ms·Gs)1; signal intensity. Nuclear Electronics & Detection Technology,
NMR—porosity in NMR, zero dimension; 2006, 26(5): 557–560.

 276 