PETROLEUM EXPLORATION AND DEVELOPMENT

Volume 43, Issue 4, August 2016

Online English edition of the Chinese language journal

Cite this article as: PETROL. EXPLOR. DEVELOP., 2016, 43(4): 611615. RESEARCH PAPER

Physical properties modeling of reservoirs in Mansuri oil

field, Zagros region, Iran
ALI Dashti1, *, EBRAHIM Sfidari1,2
1. School of Geology, College of Science, University of Tehran, Tehran, Iran;
2. Petroleum Geology Research Group, Research Institute of Applied Sciences, ACECR, Iran

Abstract: The porosity and permeability distribution in four layers of the Cretaceous Ilam Formation was simulated using optimized
artificial intelligent algorithms based on conventional logging data of 50 wells in Mansuri oil field in Iran. First, the neutron porosity,
interval transit time and density wireline logs in five key wells with core data were used as input parameters to calculate porosity and
permeability of the reservoirs using backpropagation artificial neural network (BP neural network) and Support Vector Regression
methods, and based on the correlation between the calculated results and the core tested results, BP neural network method was taken to
do the physical property calculation. Then, the porosity and permeability distribution of the four layers were modeled using kriging
geostatistical algorithms. The results show that Layers 2.1 and 2.2 are high in porosity, Layers 1, 2.1 and 2.2 are high in permeability,
while Layer 3 is not reservoir; and the porosity and permeability are higher in the north and lower in the south on the whole.

Key words: reservoir; physical property modeling; BP neural network; Support Vector Regression; Mansuri oil field; Zagros region

Introduction of Upper Cretaceous Santonian - Campanian carbonate rocks

Reservoir property prediction is one of the most
complicated and significant affairs for oilfield exploration and
development[12]. Artificial neural network, especially back-
propagation artificial neural network (BP neural network) is
widely used in predicting porosity and permeability, whereas
support vector regression is another reservoir porosity and
permeability estimation method suggested recently[3]. In this
paper, based on conventional log data, BP neural network and
support vector regression methods were used to predict the
porosity and permeability of Cretaceous Ilam Formation in the
Mansuri Oilfield of the Zagros Region, Iran; the algorithm
with higher accuracy was adopted to calculate the reservoir
porosity and permeability of all wells; and then, the Kriging
algorithm was used to build a reservoir attribute model.
1. Geological setting
Based on tectonic style and sedimentary history, the Zagros
Region in Iran can be divided into such zones as Dezful
Embayment, Lurestan and Fars[47], among which, the Dezful
Embayment zone is controlled by three major faults, i.e.,
Kzerun fault, mountain front fault and Balarud fault
respectively. The Mansuri Oilfield is located in the south of
the Dezful Embayment zone, with structural trend the same as
the Zagros fold-thrust belt (Fig. 1); Ilam Formation composed
is studied in the paper, which is considered as the third
important reservoir in the southwest of Iran[810], and it is also
a high yield reservoir in the oilfield.
2. Porosity and permeability distribution
modeling
The conventional log data of 50 wells in the Mansuri
Oilfield were used for modeling, five of them with core
analysis derived physical property (porosity and permeability)
data. BP neural network and support vector regression
methods were used to build the correlation between log
parameters and physical property parameters (porosity and
permeability) of cores of cored wells. Firstly, for the purpose
of using the BP neural network to estimate outputs, the
network was trained and tested, and the accuracy of
structuring network was evaluated. Secondly, a support vector
regression model was built to estimate the reservoir property
characteristics of the study area. Thirdly, the calculation
accuracy of the two methods was correlated and analyzed, so
as to select more accurate method and apply its results in
physical property modeling. And fourthly, the Kriging
algorithm was used to build a 3D attribute (porosity and
permeability) model of reservoir to realize visualization and
separate the reservoir interval from the nonreservoir interval.
Because the 50 wells selected in the paper distribute evenly

Received date: 14 Dec. 2015; Revised date: 31 May 2016.

* Corresponding author. E-mail: Alidashti@ut.ac.ir
Copyright 2016, Research Institute of Petroleum Exploration and Development, PetroChina. Published by Elsevier BV. All rights reserved.
 ALI Dashti et al. / Petroleum Exploration and Development, 2016, 43(4): 611615

Fig. 1. Position and stratigraphic column of Mansuri Oilfield (modified from references[47]).

over the whole Mansuri Oilfield, the smooth effect of the Table 1. Correlation coefficient between conventional log pa-
Kriging algorithm has been obviously reduced, and the rameters and porosity and permeability
modeling results are reliable. Finally, the attribute model Conventional logs W/porosity W/permeability
obtained from modeling was used to analyze the reservoir Sonic differential time 0.73 0.60
property of four zones in the Ilam Formation of Mansuri
Density wireline log 0.69 0.58
Oilfield.
Neutron porosity 0.65 0.63
2.1. Neural network modeling Gamma Ray 0.54 0.45

Firstly, the core porosity and permeability data of four wells Photoelectric 0.50 0.48
were used to train the artificial neural network, and the Conductivity temperature 0.43 0.39
physical property data of cores taken from the fifth well were
Table 2. Statistics of data used in artificial neural network
used to test the network, so as to guarantee the accuracy. At
Number of Number of
the time of training the network, the conventional log Data set
porosity data permeability data
parameters having the best correlativity to porosity and
Training set 712 356
permeability were screened out. The correlation coefficient of
Validation set 152 76
some conventional log parameters and physical property data
Testing set 152 76
is listed as in Table 1, showing that the neutron porosity, sonic
All data 1 016 508
differential time and density log value have the highest
correlativity to porosity and permeability, therefore, the placed in training set to ensure the trained network to be
neutron porosity, sonic differential time and density log are reliable.
selected as the input parameters in the paper. All the data used
2.2. Support vector regression modeling
in artificial nueral network training and the statistics of them
are listed as in Table 2, showing that about 70% input data are The input data set of support vector regression modeling is
612
 ALI Dashti et al. / Petroleum Exploration and Development, 2016, 43(4): 611615
the same as that of BP neural network method, including
neutron porosity, sonic differential time and density log. A
forward selection and backward elimination mechanism was
run to select these logs as the input data set. In the forward
selection process, the model was designed without any
variable in the start point; in the next steps, all of variables
were input in the model and their effects were monitored by a
selected model comparision standard; iterative computation
was ceaselessly conducted and did not stop until the input and
output data of the model have the highest correlativity. The
function of the model comparision standard is to recognize the
best correlation coefficient between input variables and
outputs. In the backward elimination step, all variables were
input at the start point and would be eliminated one by one in
the next stages. The model comparision standard would record
every variable elimination effect on accuracy and stop the
process at the most reliable state.
2.3. Method correlation and optimization
The correlation between the porosity and permeability
predicted from the BP neural network and support vector
regression methods and anzlyzed from the cores are shown as
in Fig. 2. As shown in Fig. 2a and 2b, the correlation
coefficient between porosity predicted by two methods and
porosity analyzed from cores is 0.912 and 0.899 respectively,
ensuring the the porosity prediction accuracy of the trained
Fig. 2. Crossplot of physical property values analyzed from cores and predicted by BP neural network and support vector regression
methods (Rcorrelation coefficient).
613
network. Because the permeability has more influential
factors than the porosity, the permeability prediction accuracy
is not as good as porosity prediction accuracy. The correlation
coefficient in Table 1 also shows that the correlativity between
conventional log parameters and permeability is not as good
as that between conventional log parameters and porosity. As
shown in Fig. 2c and 2d, the correlation coefficient between
permeability precited by two methods and permeability
analyzed from cores is 0.825 and 0.814 respectively. Ob-
viously, when BP neural network and support vector regres-
sion methods are used to model the porosity and permeability
respectively, the correlativity of BP neural network method is
superior to that of support vector regression method, possibly
due to the repetitive network training process adopted in the
neural network method. As listed in Table 1 and shown in Fig.
2, the correlativity between the porosity and permeability
figured out by BP neural network method and that measured
from cores is obviously better than the correlativity between
log parameters and core porosity and permeability. Therefore,
the BP neural network method is adopted in the paper to
calculate the porosity and permeability of target zones in 50
wells and build attribute model.
2.4. Physical property distribution modeling
The planar size of geologic model of the Mansuri Oilfield
is 40 km 6 km, and the planar grid size is 100 m100 m; the
 ALI Dashti et al. / Petroleum Exploration and Development, 2016, 43(4): 611615

geologic model is subdivided into 283 layers vertically, with
vertical grid size of 1 m; as a result, the model is composed of
6 792 000 grids. The Kriging algorithm was used to model the
physical property parameters of four zones (zone 1, zone 2.1,
zone 2.2 and zone 3) in Ilam Formation, with the modeled
porosity and permeability distribution of the four target zones
as shown in Figs. 3 and 4 respectively. As shown in Fig. 3,
zone 2.1 has the highest porosity value, and it can be regarded
as the main reservoir of Ilam Formation; the porosity
distribution exhibits an increase trend from the south to the
north. Zone 2.2 has the secondary high porosity value in Ilam
Formation; zone 1 has lower porosity value, almost exbithing
a nonporous rock; whereas zone 3 has the lowest porosity
value, and it is classified into nonreservoir.

Fig. 3. Porosity distribution in Ilam zones based on modeling.

3
Fig. 4. Permeability distribution in Ilam zones based on modeling (Kpermeability, 10 m2).

614
 ALI Dashti et al. / Petroleum Exploration and Development, 2016, 43(4): 611615
The permeability distribution of four zones in Ilam For-
mation is shown as in Fig. 4. Because logarithm processing is
performed to permeability in the course of training the neural
network, the permeability obtained through attribute modeling
is also a logarithmic value. As shown in Fig. 4, zones 1, 2.1
and 2.2 have high permeability values, and they are regarded
as permeable zones. In zone 1, the permeability is higher in
local area, which possibly attributes to the development of
local fractures, and this zone is regarded as local fracture
development zone, having the same distribution feature as
porosity in Fig. 3a. Zones 2.1 and 2.2 have a relatively even
and stable distribution of high permeability values, and they
are judged as moderate - high permeable zones. Zone 3 has
the smallest predicted permeability value, and its storage
performance is negligible, which is also proved in porosity
distribution as shown in Fig. 3d. Observed from the field scale
permeability distribution (Fig. 4), the permeability is higher in
the north and central part of the oilfield, which is also the
same as the porosity distribution as shown in Fig. 3.
3. Conclusions
As to the Cretaceous Ilam reservoir in the Mansuri Oilfield
of the Zagros Region, Iran, the correlativity between the
porosity and permeability figured out using the BP neural
network method and that measured from cores is superior to
the calculation results of the support vector regression method.
Based on the calculation results of the artificial neural
network method, the Kriging geostatistical algorithm is used
to model the porosity and permeability in four zones of Ilam
Formation of the Mansuri Oilfield, showing that zones 2.1 and
2.2 are high porosity zones, zones 1, 2.1 and 2.2 are
permeable zones, the physical property of zone 1 is
complicated due to the development of local fractures, and
zone 3 is a nonreservoir, whose porosity and permeability
value are all relatively low. The reservoir porosity and
permeability distribution exhibits a characteristic of high in
the north and low in the south on the whole.
615
References
[1] LIM J S. Reservoir properties determination using fuzzy logic
and neural networks from well data in offshore Korea. Jour-
nal of Petroleum Science and Engineering, 2005, 49(3):
182192.
[2] MOHAGHEGH S, AREFI R, AMERI S, et al. Petroleum res-
ervoir characterization with the aid of artificial neural net-
works. Journal of Petroleum Science and Engineering, 1996,
16(4): 263274.
[3] VERMA A, SINGH T N, MAHESHWAR S. Comparative
study of intelligent prediction models for pressure wave ve-
locity. Journal of Geosciences and Geomatics, 2014, 2(3):
130138.
[4] FALCON N L. South Iran: Zagros Mountains. In: SPENCER
A M. MesozoicCenozoic orogenic belts: Data for orogenic
studies. London: Geological Society of London, Special Pub-
lication, 1974: 199211.
[5] SEPEHR M, COSGROVE J W. The role of the Kazerun fault
zones in the formation and deformation of the Zagros
fold-thrust belt, Iran. Tectonics, 2005, 24(5): 289311.
[6] ALAVI M. Regional stratigraphy of the Zagros fold-thrust belt
of Iran and its proforeland evolution. American Journal of
Science, 2004, 304(1): 120.
[7] SEPEHR M, COSGROVE J W. Structural framework of the
Zagros Fold-Thrust belt, Iran. Marine and Petroleum Geology,
2004, 21(7): 829843.
[8] JAMES G A, WYND J G. Stratigraphic nomenclature of Ira-
nian oil consortium agreement area. AAPG Bulletin, 1965,
49(12): 21822245.
[9] MOTIEI H. Geology of Iran, the stratigraphy of Zagros. Te-
hran: Geological Survey of Iran Publication, 1993.
[10] ADABI M H, ASADI-MEHMANDOSTI E. Microfacies and
geochemistry of the Ilam Formation in the Tang-E Rashid
area, Izeh, SW Iran. Journal of Asian Earth Sciences, 2008,
33(3/4): 267277.