Journal of Petroleum Science and Engineering 108 (2013) 10–13

Contents lists available at SciVerse ScienceDirect

Journal of Petroleum Science and Engineering

journal homepage: www.elsevier.com/locate/petrol

Porosity sensitivity study of the compensated neutron logging tool

Wensheng Wu a,n, Maosong Tong b, Lizhi Xiao a, Jianbo Wang a
a
China University of Petroleum–Beijing, Beijing, China
b
Daqing Drilling Corporation, PetroChina, Daqing, China

art ic l e i nf o a b s t r a c t

Article history: Porosity sensitivity is an important index for determining the spacing of the neutron porosity tool, and
Received 10 September 2012 optimized spacing can improve formation porosity resolution. Based on the neutron flux formula
Accepted 26 April 2013 obtained through a double-group approach, we study the variations of fast-neutron slowing-down
Available online 24 May 2013
length and thermal-neutron diffusion length and their derivatives with formation porosities; calculate
Keywords: porosity sensitivity for different porosity values; and derive relationships for how the distance between
neutron logging the source and the near detector or the distance between the two detectors depends on porosity
porosity sensitivity sensitivity. Both the theoretical analysis and the calculations show that the relationship between porosity
theoretical analysis sensitivity and the short spacing is not monotonic. For a fixed spacing interval, sensitivity increases with
Monte Carlo simulation
increasing spacing in lower porosity formations, whereas the opposite occurs in relatively higher
porosity formations. The spacing has little influence on the sensitivity. For a fixed short spacing, the
sensitivity is a monotonically increasing function of the spacing interval in the whole range of formation
porosity.
& 2013 The Authors. Published by Elsevier B.V. Open access under CC BY-NC-ND license.

1. Introduction
The neutron logging tool is used to obtain an estimate of
formation porosity. The measurement principle is based on the
fact that hydrogen is the element that most strongly slows down
fast neutrons in the formation. Because hydrogen in the formation
is usually in the form of hydrocarbons or water and tends to occur
in pore spaces, the correlation with formation porosity can easily
be made. In practice, the double-detector neutron tool is used to
obtain formation porosity. The detector closer to the source is used
to provide compensation for borehole effects and rock capture
effects on the farther detector by the ratio of near to far count rates
(Tittle, 1961; Kreft, 1972; Kreft et al., 1974; Ellis and Singer, 1987;
Ellis et al., 2003; Xu, 2009).
Porosity sensitivity (S) is one of the main indexes used to
determine the resolution of formation porosity (Huang, 1985;
Oraby et al., 1990; Xia et al., 2003; Zhang et al., 2006). It is a
function of counting ratio (R) and formation porosity (ϕ) and is
defined by
1 ∂R
S¼ : ð1Þ
R ∂ϕ
The larger the porosity sensitivity, the higher the resolution to
porosity.
Though considerable work on neutron porosity logging espe-
cially neutron slowing-down length has been performed by
specialists in the use of neutron–neutron devices, the variations
of porosity sensitivity with the distance between the source and
the near detector or the distance between the two detectors have
not been included. Based on the neutron flux formula obtained
through the double-group approach, in this paper we study the
influences of the distance between the source and the near
detector and the distance between the two detectors on porosity
sensitivity.
2. Porosity sensitivity
In a simple geometric structure, for a point source in an infinite
formation, the thermal-neutron flux of an arbitrary point in the
formation is given by
1 L2 t
ϕt ðrÞ ¼ ðe−r=Le −e−r=Lt Þ; ð2Þ
4πDt r Le 2 −Lt 2

where Dt is the thermal diffusion coefficient, Le is the slowing-

n
Corresponding author. Tel.: +86 10 89733310; fax: +86 10 89733305.
E-mail address: wwsheng@cup.edu.cn (W. Wu).
down length of fast neutrons, Lt is the diffusion length of thermal
neutrons, and r is the distance from the source to the point. In
actual logging, r is the distance from the neutron source to a
detector and is called spacing. It can be seen from Eq. (2) that the
thermal-neutron flux distribution depends on not only formation

0920-4105 & 2013 The Authors. Published by Elsevier B.V. Open access under CC BY-NC-ND license.
http://dx.doi.org/10.1016/j.petrol.2013.04.020
 W. Wu et al. / Journal of Petroleum Science and Engineering 108 (2013) 10–13 11

28
Borehole

Far detector 24 sandstone

Slowing-down length (cm)

limestone
dolomite
dr
20
Near detector

r 16

Neutron source
12
Formation

8
Fig. 1. Schematic diagram of compensated neutron logging. 0 10 20 30 40 50
Porosity (%)
fast-neutron slowing-down length but also formation thermal-
neutron diffusion and absorption. Fig. 2. Relationship between Le and ϕ (from Wiley et al., 1989).

In this paper, we suppose that the detectors are point detectors,

as shown in Fig. 1. In the figure r represents the distance from the 18

source to the near detector, and Δr, which is called the interval,
represents the distance from the near detector to the far detector.
15 sandstone
If Nt ðrÞ is used to represent the thermal-neutron count rate at
limestone
the location of a detector, N t ðrÞ is proportional to the thermal-

Diffusion length (cm)

dolomite
neutron flux ϕt ðrÞ, that is, N t ðrÞ ¼ Kϕt ðrÞ. The count rates of the
12
near detector and the far detector are, respectively,
K Lt 2
N t ðrÞ ¼ ðe−r=Le −e−r=Lt Þ; ð3Þ
4πDt r Le −Lt 2
2
9

K Lt 2
N t ðr þ ΔrÞ ¼ ðe−ðrþΔrÞ=Le −e−ðrþΔrÞ=Lt Þ: ð4Þ
4πDt ðr þ ΔrÞ Le −Lt 2
2 6

The ratio R of N t ðrÞ to N t ðr þ ΔrÞ is introduced to compensate

for the influence of formation absorption and the borehole. The 0 10 20 30 40 50
ratio R is Porosity (%)

N t ðrÞ r þ Δr e−r=Le −e−r=Lt Fig. 3. Relationship between Lt and ϕ (from Wiley et al., 1989).
R¼ ¼ : ð5Þ
N t ðr þ ΔrÞ r e−ðrþΔrÞ=Le −e−ðrþΔrÞ=Lt
In Eq. (5), Le and Lt are functions of porosity (ϕ). fitting the data in Figs. 2 and 3, the result of which gives
The partial derivative of porosity is used for Eq. (5); if we let
A ¼ ð1=Le Þ−ð1=Lt Þ, then Le ¼ −3:03lnðϕ=1000Þ;


∂R r þ Δr ΔreΔr=Le ð1−eAr Þ∂ð1=Le Þ=∂ϕ−reðΔr=Le ÞþAr ∂A=∂ϕ Lt ¼
1
: ð8Þ
¼
∂ϕ r 1−eAðrþΔrÞ 0:0026ϕ þ 0:0602
#
Δr=Le ArþðΔr=Le Þ
ðe þe Þðr þ ΔrÞeAðrþΔrÞ ∂A=∂ϕ The relationship between Le and ϕ can also be found in Lysne
þ : ð6Þ
ð1−eAðrþΔrÞ Þ2 (1989).
Next we analyze the variations of ð∂=∂ϕÞð1=Le Þ and ð∂=∂ϕÞð1=Lt Þ
Plugging Eq. (6) into Eq. (1) gives the porosity sensitivity:
  with ϕ.
∂ð1=Le Þ ðr þ ΔrÞeAðrþΔrÞ reAr ∂A According to Eq. (8), 1=Le and 1=Lt as functions of ϕ can be
S ¼ Δr þ − : ð7Þ
∂ϕ 1−eAðrþΔrÞ 1−eAr ∂ϕ obtained, as shown in Fig. 4. Fig. 4 shows that both 1=Le and 1=Lt
increase with increasing ϕ; the values of 1=Le are all less than
those of 1=Lt in the range of formation porosity; the increasing rate
3. Positive and negative properties of ∂A=∂ϕ of change of 1=Le with ϕ is greater than that of 1=Lt with ϕ in the
low-porosity range, whereas in the high-porosity range, the rate of
Because ∂A=∂ϕ ¼ ∂=∂ϕð1=Le Þ−ð1=Lt Þ ¼ ∂=∂ϕð1=Le Þ−∂=∂ϕð1=Lt Þ, the change of 1=Le is less than that of 1=Lt , and the difference between
following steps can be used to analyze the positive and negative the two rates will increase with increasing porosity.
properties of ∂A=∂ϕ. If partial derivatives of porosity are used for the rules in Fig. 4,
First we analyze the relationship between the slowing-down the variations of ∂=∂ϕð1=Le Þ and ∂=∂ϕð1=Lt Þ as functions of ϕ can be
length Le or the diffusion length Lt and porosity ϕ. It can be seen obtained. The relationships are shown in Fig. 5. As seen from the
from Wiley and Patchett (1989), Yu et al. (2010), and Smith (1986) figure, when ϕ is less than about 4.4%, the value of ð∂=∂ϕÞð1=Le Þ is
that both the slowing-down length Le and the diffusion length Lt greater than the corresponding value of ∂=∂ϕð1=Lt Þ: ∂=∂ϕ ¼
decrease with increasing porosity ϕ, as shown in Figs. 2 and 3. ∂=∂ϕð1=Le −1=Lt Þ 4 0. In this case, ∂A=∂ϕ is positive. When ϕ is
Figs. 2 and 3 show that Le or Lt exhibit different degrees of greater than about 4.4%, the value of ∂=∂ϕð1=Le Þ is less than the
reduction with increasing ϕ in different lithologic formations. corresponding value of ð∂=∂ϕÞð1=Lt Þ, and ∂A=∂ϕ is negative; when φ
Because compensated neutron logging is mainly used in sand- is about 4.4%, ∂A=∂ϕ¼0, and the second term in Eq. (7) is equal
stone, here the equations between Le or Lt and ϕ are obtained by to zero.
12 W. Wu et al. / Journal of Petroleum Science and Engineering 108 (2013) 10–13

0.18 the choice of the empirical equations (8) used to describe the Le and
Lt parameters. If we use other equations (for example, two logarith-
1/Lt mic or two polynomial equations), the two curves do not necessarily
0.15
1/Le cross; nevertheless, the meaning and the behavior of the porosity
sensitivity S remain identical to those described in the paper.
0.12
1/Le & 1/Lt

4. Variation of porosity sensitivity with the spacing r

0.09
To obtain the changing rule of porosity sensitivity (S) with the
spacing (r) porosity，the partial derivative of r is used for S in Eq. (7)
0.06 to obtain
" #
∂S eAr ðeAr −Ar−1Þ eAðrþΔrÞ ðeAðrþΔrÞ −Aðr þ ΔrÞ−1Þ ∂A
¼ − ; ð9Þ
0.03 ∂r ð1−eAr Þ2 ð1−eAðrþΔrÞ Þ2 ∂ϕ
0 5 10 15 20 25 30 35
Porosity (%) where the two terms in the brackets are two different forms of the
same function f ðyÞ ¼ ey ðey −y−1Þ=ð1−ey Þ2 , with the difference between
Fig. 4. Changing rules of 1/Le and 1/Lt with ϕ.
the two items being equivalent to the difference between the two
functions, respectively, with independent variables y1 and y2 [y1 ¼ Ar,
0.006 y2 ¼ Aðr þ ΔrÞ]. Because A ¼ ð1=Le Þ−ð1=Lt Þ o 0, y2 ¼ Aðr þ Δr Þ o y1 ¼ Ar o 0.
Consequently, as soon as the monotonicity of the function f ðyÞ ¼
1st derivative of 1/ Le ey ðey −y−1Þ=ð1−ey Þ2 can be determined, the sensitivity as a function of
1st derivative of 1/ Lt the spacing r can be obtained.
d(1/Le )/d Φ & d(1/Lt )/d Φ

0.004 It is difficult to determine the monotonicity of the function

f ðyÞ ¼ ey ðey −y−1Þ=ð1−ey Þ2 with partial derivatives, but a drafting
method is a better choice. Because y2 ¼ Aðr þ ΔrÞ o y1 ¼ Ar o 0; let
y in the function take negative values, and the curve of the
function can be drawn as shown in Fig. 7. It can clearly be seen
0.002
that the functionf ðyÞ ¼ ey ðey −y−1Þ=ð1−ey Þ2 is a monotonically
increasing function. For y2 o y1 o 0, f ðy2 Þ o f ðy1 Þ can be obtained.
Thus, the expression in brackets in Eq. (9) will be greater than
φ=4.4% zero. Because the positive and negative properties of ∂A=∂ϕ are
0.000
also known, when ϕ is less than about 4.4%, the term on the left-
0 5 10 15 20 25 30 35
hand side ð∂S=∂rÞ in Eq. (9) is greater than zero, which indicates
Porosity (%)
that neutron porosity sensitivity (S) increases with the increasing
Fig. 5. Changing rules of ð∂=∂ϕÞð1=Le Þ and ð∂=∂ϕÞð1=Lt Þ with ϕ. spacing (r); when ϕ is greater than about 4.4%, ∂S=∂r is less than
zero, which indicates that the sensitivity (S) decreases with
increasing spacing (r). Therefore, when designing a neutron
1.0 logging tool the measurement precision of the porosity can be
improved by selecting the correct relationship between spacing
Normalized porosity sensitivity

and formation porosity.

0.8
Fig. 8 shows the behavior of the derivatives ∂S=∂r as functions
of r for several values of the porosity. For low values of the porosity
0.6 (1% or 3%), the values of ∂S=∂rare positive. For higher values of the
porosity (≥5%), the values of ∂S=∂r are negative. For a fixed r, the
0.4 lower the porosity, the greater is the value of∂S=∂r: For all

0.2
1

e
y
(e y
− y − 1)
0.0 f ( y) = 2
1E-3
(1 − e y )
0 5 10 15 20 25 30 35 40
Porosity (%)

Fig. 6. Curve of the function f ðyÞ ¼ ey ðey −y−1Þ=ð1−ey Þ2 . 1E-6

f(y)

Fig. 6 shows the porosity sensitivity as a function of porosity for 1E-9

AmBe source. These porosity sensitivities are normalized for compar-
ison. It can be seen that the porosity sensitivities are very sensitive
1E-12
only to low values of porosity (up to 5%), corresponding to high
migration lengths to the epithermal and thermal neutrons, and
decrease very quickly. For higher values of porosity, these sensitivities
-35 -30 -25 -20 -15 -10 -5 0
are then almost constant and less affected by formation porosity.
y
Although the peculiar value of 4.4% for the porosity plays a key
role in all the next steps of the discussion, this value results only from Fig. 7. Response curves of the sensitivity with two spacings to formation porosity.
 W. Wu et al. / Journal of Petroleum Science and Engineering 108 (2013) 10–13 13

0.004 0.016
1% 1% Porosity
3% 3%
0.014
0.003 5% 5%
10% 10%
0.012
20% 20%
0.002
30% 30%
0.010

dS/d(Δr)
dS/dr

0.001
0.008

0.000 0.006

0.004
-0.001

0.002
-0.002
0.000
35 36 37 38 39 40 41 20 22 24 26 28 30 32
r (cm) Δr (cm)

Fig. 8. Response curves of the sensitivity with two spacing intervals to formation Fig. 9. Plot showing the behavior of the derivatives ∂S=∂ðΔrÞ vs r for several values
porosity. of the porosity.

monotonic. When the spacing interval is fixed, the sensitivity

measured porosities, the values of ∂S=∂rare very low, which means increases with increasing spacing in formations of low porosity,
that the spacing r has little effect on porosity sensitivity S. whereas the opposite is observed in formations of relatively high
porosity. However, the influence of the short spacing on the
sensitivity is very minor. When the spacing is fixed, the sensitivity
5. Variation of porosity sensitivity with the interval Δr has a monotonically increasing relationship with the spacing
interval in the whole range of formation porosity.
If we use the partial derivative of Δr for S in Eq. (7) we can
obtain the variation of porosity sensitivity (S) with the interval (r):
∂S ∂ð1=Le Þ eAðrþΔrÞ ð−eAðrþΔrÞ þ Aðr þ ΔrÞ þ 1Þ ∂A
¼ þ : ð10Þ Acknowledgments
∂Δr ∂ϕ ð1−eAðrþΔrÞ Þ2 ∂ϕ
Because the slowing-down length (Le ) decreases with forma-
This work was supported by the China Natural Science Fund
tion increasing porosity (ϕ), 1=Le increases with increasing ϕ, and
(no. 41074101) and the Science Foundation of China University of
the first term on the right-hand side of Eq. (10) is greater than
Petroleum, Beijing, China.
zero, as shown in Fig. 5. The following transformation can be
performed for the second term on the right-hand side of Eq. (10):

eAðrþΔrÞ ð−eAðrþΔrÞ þ Aðr þ ΔrÞ þ 1Þ ∂A −eAðrþΔrÞ ðeAðrþΔrÞ −Aðr þ ΔrÞ−1Þ ∂A References

¼
ð1−e AðrþΔrÞ Þ2 ∂ϕ ð1−eAðrþΔrÞ Þ2 ∂ϕ
AðrþΔrÞ AðrþΔr Þ
Fig. 7 shows that e ð−e þ Aðr þ ΔrÞ þ 1Þ=ð1−e Þ is
less than zero, so the positive and negative properties of the second
term depend on ∂A=∂ϕ. When sandstone porosity is less than about
4.4%, ∂A=∂ϕ 4 0, and the second term on the right-hand side of
Eq. (10) is less than 0, but the absolute value of ð∂=∂ϕÞð1=Le Þis large,
and the first term is much larger than the second one on the right-
hand side of Eq. (10), so the partial derivatives of Δr for the sensitivity
(ð∂S=∂ΔrÞ) are greater than 0. When the porosity is greater than about
4.4%, ∂A=∂ϕ o 0, the first and the second terms are both positive, and
∂S=∂Δris greater than 0. Therefore, ∂S=∂Δr is always greater than 0 for
a constant formation porosity, which indicates that the porosity
sensitivity of the neutron tool (S) increases with increasing spacing
interval ðΔrÞ.
Fig. 9 shows the behavior of the derivatives ∂S=∂ðΔr Þ as functions of
r for several values of the porosity. For all porosity values, the values of
∂S=∂ðΔr Þ are positive, which indicates that the sensitivity S increases
with increasing porosity. For a fixed r, higher porosity brings lower
∂S=∂ðΔr Þ; that is, the sensitivity S in a higher porosity formation is less
affected by the interval Δr than that in a lower porosity formation.
6. Conclusions
Theoretical analysis and calculations clearly demonstrate that
the relationship between porosity sensitivity and spacing is not
Ellis, D.V., Case, C.R., Chiaramonte, J.M., 2003. Tutorial: porosity from neutron logs I
AðrþΔr Þ 2 —Measurement. Petrophysics 44 (6), 383–395.
Ellis, D.V., Singer, J.M., 1987. Well Loggings for Earth Scientists. Elsevier, Dordrecht
250 pp.
Zhang, Jianming, Xia, Lingzhi, Qiu, Yixiang, 2006. Sensitivity of compensated
neutron oil logging tool. At. Energy Sci. Technol. 40 (1), 126–128.
Kreft, A.,, 1972. Generalization of the Multigroup Approach for Calculating the
Neutron Slowing Down Length. Inst. of Nuclear Physics and Techniques
(Cracow) Report No. ITJ-32, 1972, Poland.
Kreft, A., Lenda, A., Turek, B., et al., 1974. Determination of effective porosities by the
two-well pulse method. In: Isotope Techniques in Ground water Hydrology, vol.
II, International Atomic Energy Agency, Vienna, pp. 295–312.
Xu Libai, Schultz, W., Huiszoon, C., 2009. A comprehensive investigation of source
effects on neutron porosity response for logging-while-drilling measurements.
In: SPWLA 50th Annual Logging Symposium,June 21–24, 2009.
Huang Longji, 1985. The principle of nuclear logging. Petroleum Industry Press,
Beijing, pp. 108–148.
Lysne, P., 1989. Investigation of neutron-porosity log uncertainties: Ocean Drilling
Program Hole 642E. In: Proceedings of the Ocean Drilling Program, Scientific
Results, 1989, vol. 104, pp. 973–977.
Oraby, M., Verghese, K., Gardner, R.P., 1990. Investigation of an improved-
sensitivity neutron-porosity oil-well logging tool. Nucl. Instrum. Methods Phys.
Res. 299 (1–3), 674–681.
Yu Qingdong, Luan Shumei, Sun Huaxue, et al., 2010. The responses of compensated
neutron logging. Foreign Oilfield Eng. 26 (7), 36–40.
Wiley, R., Patchett, J.G., 1989. CNL Neutron porosity modeling, a step forward. In:
SPWLA 30th Annual Logging Symposium, June 11–14, 1989, Paper OO.
Smith, M.P., 1986. Calibration, checking and physical corrections for a new dual-
spaced neutron porosity tool. In: SPWLA 20th Annual Logging Symposium, June
9–13, 1986, Paper YY.
Tittle, C.W., 1961. Theory of neutron logging I. Geophysics 26 (1), 27–39.
Xia, L., Zhang, J., Yang, Z., 2003. Application of Monte Carlo method in compensated
neutron log tool design. Well Logging Technol. 27 (6), 477–480.