Vous êtes sur la page 1sur 15

SPE-176101-MS 1

All final manuscripts will be sent through an XML markup


process that will alter the LAYOUT. This will NOT alter the
content in any way.

SPE-176101-MS

Adapting Fracture Corridors and Diffuse Fractures in Single Porosity Model


of Ujung Pangkah Carbonate Reservoir
Rini Saputra, Azalea Hidayat, Wahyu Agung Rahmanto, Saka Indonesia Pangkah Limited.

Copyright 2015, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE/IATMI Asia Pacific Oil & Gas Conference and Exhibition held in Nusa Dua, Bali, Indonesia, 20–22 October 2015.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract

A practical method to adapt fractures both micro and macro as flow enhancing properties in a single
porosity model is introduced to simulate Ujung Pangkah fractured carbonate reservoir. This approach is
taken because dual porosity modeling attempt fails to explain the behavior of many wells which
experience early water breakthrough and/or excessive water production in Ujung Pangkah field.

The enhancement factor term is used to define the degree of permeability enhancement by diffuse or micro
fractures. At well location, the enhancement factor can be determined by the ratio of production test to the
production of matrix-only model. The enhancement factor 3D distribution is derived from well data and
seismic minimum curvature attributes as trend. Fracture corridors correspond to macro fractures in the
order of meters extending vertically and/or laterally. As normally scattered spatially, fracture corridors
cannot be modeled in a discrete fracture network model which is the integral part of dual porosity model.
Some wells show behavior anomalies, such as rapid, early water breakthrough with excessive water
production in unlikely location. It is observed that the location of these anomalies coincide with the
fracture lineaments derived from the seismic incoherency attribute.

As the fractures are well characterized, diffuse fractures as permeability enhancement and fracture
corridors as high permeability streaks, the further improvement of history match is then easily achieved
by calibrating two other key parameters; relative permeability curves and aquifer strength. The relative
permeability is calibrated to the shape of fracture relative permeability. Oil rate match is greatly improved.
Water rate match is achieved by placing adequate aquifer strength.

Reservoir dynamic of Ujung Pangkah carbonate fractured reservoir can be simulated as a single porosity
model with permeability enhancement adapted from two types of fracture distribution. Diffuse fractures
enhance the overall permeability and fracture corridors dominantly influence flow dynamic in certain local
area. Compared to dual porosity model, adapting fractures as permeability enhancement in single porosity
2 SPE-176101-MS

model is more practical, more efficient in simulation run time – computational cost.

Introduction

Ujung Pangkah field as part of the Oligo-Miocene Kujung Formation is a fractured carbonate reservoir.
While fractures are identified in core and image logs from exploration and appraisal wells, no drilling loss
was reported as the exploration well vertical trajectories missed to encounter fractures. Seismic data at
this phase was very limited, only 2D seismic and not suitable for recognizing fractures.

Only during development well drilling, many fractures were penetrated by the high angle or horizontal
wells. Fracture system in this field starts to appear significant, indicated by drilling losses and suspected
to adversely contribute to the current production performance by allowing early water breakthrough.
Integration of subsurface data; drilling fluid loss, image logs, well logs, core data and seismic incoherency
attributes confirms the existence of fractures.

The conventional route to model fracture reservoir is to create a fracture model on top of the matrix model
and simulate it using a dual porosity method. Utilizing the wealth of fracture information in the field from
well and seismic attributes of fractures in Ujung Pangkah led to a poorly simulated dual porosity model,
no better than the single porosity matrix model. Well test analysis and historical production suggest that
Ujung Pangkah Kujung carbonate reservoir is a typical type III fracture reservoir, where fractures assist
permeability in an already producible reservoir where rock matrix has good porosity and permeability
(Nelson, 2001). With this in mind, single porosity model is selected and adapting fractures into the model
is tested.

The practice of defining adapting fractures as permeability enhancement has been applied in several fields
in Middle East (Charfeddine et al, 2002, Bahar et al, 2003 and Bockel-Rabelle et al, 2004). In Ujung
Pangkah similar practice is applied with different approach between diffuse fractures (micro) and fracture
corridors (macro).

Single Porosity Model vs. Dual Porosity Model Selection

Pre-calibration runs are exercised to compare the performance between single porosity model and dual
porosity model. Both model are initiated utilizing original static model without any engineering
modification and simulation was run with gas rate as control rate. The comparison of the simulation results
(field gas rate, oil rate, water rate and field pressure) between single porosity and dual porosity model is
presented in Fig. 1 and Table 1.

It is observed in dual porosity model, the simulation does not even match the control rate (gas rate). In
general, the simulation results from dual porosity model are very poor in comparison to single porosity
model. The single porosity model took six hour run time while dual porosity model took six days. Dual
porosity model needs longer run time due to complex solver equation to accommodate both matrix and
fracture properties. In conclusion, single porosity model gives better match and reasonable run time
compared to dual porosity model. It is decided to use single porosity model and then adapt fracture
properties to get a proper representation of Ujung Pangkah reservoir model.
SPE-176101-MS 3

Reservoir Dynamic Attributed to Fractures

Ujung Pangkah field has undergone several stages of deformation, both extensional as well as
compressional. The fractures observed are attributes to these differing deformation styles, which leads to
different scales of fractures present in the area (Easley, 2010).

Pre-simulation result of single porosity model which is matrix only could not match the actual flow
capacity. Permeability enhancement is expected to be provided by fractures. There are two fracture types
in play in Ujung Pangkah shown in Fig.2, adopting from the definition from Bockel-Rabelle, et.al (2004):

1. Micro fractures are also known as diffuse fractures or background fractures. Diffuse fractures
provide global permeability enhancement which gives additional transmissibility for horizontal
direction. Diffuse fractures increase fluid flow capability explaining the reason why the reservoir
could perform better than the flow capability from the matrix itself.

2. Macro fractures or fracture corridors provide local permeability enhancement in irregular pattern of
forming early water breakthrough in unlikely locations, such as up-dip structure (southern part along
the shelf edge of Ujung Pangkah field), despite the well completion is located 60-80 feet higher from
the oil-water contact. Fracture corridors give an extra vertical transmissibility allowing fluids from
below easily moving in a vertical direction.

Given the contributions of these two types of fractures to Ujung Pangkah reservoir and well production
behaviors, both fractures are adapted to generate realistic dynamic modeling result. The model selection
scheme is shown in Fig. 3.

Adapting Fractures for Single Porosity Model

Fractures adaptation for a single porosity model is very data driven. Diffuse fractures are normally sub-
seismic and identified commonly from borehole image logs. These fractures are adapted in the model as
global permeability enhancement factor using well data guided by seismic curvature attribute.

Fracture corridors detected from seismic attributes are echelons of fractures or faults, adapted from seismic
incoherency attribute validated by well historical production data.

Adapting Diffuse Fractures as Global Permeability Enhancement

The term enhancement factor is used in this study to define the degree of permeability enhancement by
diffuse fractures. The enhancement factor is defined as multiplier value of khwelltest ratio to khmodel (Bahar,
2003 and Bockel-Rabelle, 2004). In the case of Ujung Pangkah field where the all producing wells are
completed as horizontal wells and pressure transient analysis data are very limited, the multiplier value is
simplified and defined by rate ratio between Qprod-test and Qmatrix. The Qmatrix is based on the pre-calibration
simulation run of single porosity model.

To generate permeability enhancement factor property, the multiplier values (well rate ratio) as data points
are distributed laterally as 2D map and then converted into 3D property grid. The data points is given
geospatial information, based on well production section depth interval seen in Fig. 4. Data analysis of
4 SPE-176101-MS

major and minor direction and ranges of variogram is conducted using the krigging with trend algorithm
which is minimum curvature seismic attribute.

It is decided to utilize seismic minimum curvature attribute (Fig. 5) as it relates well to fracture intensity
in Ujung Pangkah to guide the distribution of enhancement factor as local varying mean. Result of the 2D
enhancement map can be seen in Fig. 6. 3D property of enhancement factor is generated by converting
the 2D map using geometrical modeling method of “constant or surface in segments and zones”. The 3D
enhancement factor visualized in Fig. 7 are then utilized as multiplier to the original Permeability IJ and
K. Comparison visualization and histogram permeability-IJ prior and after enhancement can be viewed
from Fig. 8.

Adapting Macro Fractures as High Permeability Corridors

Fracture corridors are extracted from the seismic incoherency attribute as fault patches and incorporated
in the reservoir model as high permeable streaks. Ant-Track seismic attribute is selected as the incoherency
attribute to generate fracture corridors in this study.

The fracture corridor input are generated by realizing a specific area of interest of the Ant-Track cube
(Fig. 9) to test the best parameter for fault/fracture patch extraction. After conducting several iterations,
the final setting which generates the optimal result of fracture patch in terms of orientation and size is by
customization of predefined aggressive setting.

By integrating the parameter to the overall 3D seismic variance cube the resulted fracture patches is then
subjected to a round of quality check process by filtering the data set by confidence level, size and
orientation. It is found some fractures are extracted as multiple parts of one object, therefor manual
merging and editing are required to generate the best representation of fractures. Final result of the fracture
corridor patches is shown in Fig. 10.

To generate the fracture corridors properties, the patches are converted to polylines as reference data in
geometrical modeling using “distance to object” method. Fracture corridors polyline data is set to define
cells up to 50 m from the lines as fracture corridors. Fracture corridor IJ direction is set to have
permeability of 500 mD while fracture corridor K direction equals to 1000 mD. The fracture corridor
permeability can be modified and it is subject to history matching.

Final permeability IJ and K properties enhanced by both diffuse fracture and fracture corridors can be
seen in Fig. 11. This properties will be used as input in dynamic simulation model.

History matching in single porosity model with fracture adaptation in Ujung Pangkah

As the fractures are well characterized, diffuse fractures as permeability enhancement and fracture
corridors as high permeability streaks, the further improvement of history match is then easily achieved
by calibrating two other key parameters; relative permeability curves and aquifer strength. The relative
permeability is calibrated to the shape of fracture relative permeability. Oil rate match is greatly improved.
Water rate match is achieved by placing adequate aquifer strength.

Fig. 12 shows the history matching improvement by introducing:


• Step 1 - permeability enhancement by diffuse fractures gives an overall global improvement
SPE-176101-MS 5

towards reservoir pressure history match


• Step 2 - permeability enhancement by fracture corridors gives improvement in overall water
production history match, especially in wells experiencing early water breakthroughs for example
well (X10ST3 and Y1ST1) shown on Fig. 13.
• Step 3 - relative permeability modification gives improvement in overall oil production history
match especially in the highly fractured platform B area shown in Fig. 14.
• Step 4 - Adequate aquifer strength placement

Run times of all cases in the scenarios above are ranging from 5,500 – 30,000 seconds, which are
significantly faster than the 498,000 seconds of dual porosity run.

Conclusions

1. Reservoir dynamic of Ujung Pangkah carbonate fractured reservoir can be simulated as a single
porosity model with permeability enhancement adapted from fracture distribution. Diffuse fractures
enhance the overall permeability and fracture corridors dominantly influence flow dynamic in certain
local area.

2. Compared to dual porosity model, adapting fractures as permeability enhancement in single porosity
model is more practical, more efficient in simulation run time – computational cost.

Acknowledgments

The authors would like to thank SIPL management for permission to publish this paper.

References

Bahar, A., Ates, H., Al-Deeb, M. H., Salem, S. E., Badaam, H., & Kelkar, M. (2003, January 1). Practical Approach in Modeling
Naturally Fractured Reservoir: A Field Case Study. Society of Petroleum Engineers. doi:10.2118/84078-MS

Bockel-Rebelle, M.-O., Hassall, J. K., Silva, F. P., Lozano, J. A., Al Deeb, M., Salem, S. E. A., Vesseron, M., Al Mehsin, K.
(2004, January 1). Faults, fracture corridors and diffuse fracturing: ranking the main structural heterogeneities within onshore
Abu Dhabi fields. Society of Petroleum Engineers. doi:10.2118/88676-MS

Charfeddine, M., Al-Deeb, M., El-Abd, S., Bahar, A., Ates, H., Soeriawinata, T., & Kelkar, M. (2002, January 1). Reconciling
Core Derived Permeabilities and Well Test Using A Fracture Network: A Field Case Example. Society of Petroleum Engineers.
doi:10.2118/78499-MS

Easley, D., Yustiana, F., and Hidayat, A.: “Seismic Lineament Analysis of a Fractured Limestone Reservoir in The Ujung
Pangkah Field” SEG-2010-1347 presented at the 2010 SEG Annual Meeting, Denver, CO (Oct. 17-22, 2010).

El-Abd, S., Al-Deeb, M., Abdou, M., Linthorst, S., Bahar, A., & Kelkar, M. (2004, January 1). Practical Flow Simulation Method
for a Naturally Fractured Reservoir: A Field Study. Society of Petroleum Engineers. doi:10.2118/88761-MS

Nelson, R. A.: Geologic Analysis of Naturally Fractured Reservoirs, Guld Professional Publishing Co., (2nd Ed. 2001)
6 SPE-176101-MS

Tables

Variable Single Porosity Model Dual PorosityModel


Gas Rate (control rate) Very good Poor
Oil Rate Fair – follow trend Poor
Water Cut Fair – follow trend Poor
Reservoir Pressure Fair – follow trend Fair – follow trend
Run time 6 hours 6 days

Table 1 - Comparison simulation results of single porosity model vs. dual porosity model.

Figures

Fig. 1 - Comparison single porosity and dual porosity at pre-calibration stage. The dotted points are observed data, the blue are the
dual porosity model and red lines are single porosity model
SPE-176101-MS 7

Fig. 2 Two fracture categories; diffuse fractures (micro fractures) and fracture corridors (macro fractures)

Fig. 3 Ujung Pangkah reservoir model selection process diagram


8 SPE-176101-MS

Fig. 4. Well enhancement factor point data input

Fig. 5. Seismic minimum-curvature attribute extracted as secondary trend for multiplier value map krigging
SPE-176101-MS 9

Fig. 6. Final result of enhancement factor map gridded using Krigging with trend method

Fig. 7. Enhancement factor property


10 SPE-176101-MS

Fig. 8 A. Original Permeability-IJ (Matrix), B. Permeability-IJ after EF property application, and C. Histogram difference of value
range in Permeability-IJ before and after implementation of EF property
SPE-176101-MS 11

Fig. 9. Seismic variance (Ant-track) attribute as input for fracture corridor patch extraction.

Fig. 10. Fault patches extracted from seismic variance (ant-track) attribute
12 SPE-176101-MS

Fig. 11. Fracture corridor result shown in red merged with predefined enhanced permeability property
SPE-176101-MS 13

Fig. 12 Simulation results with permeability enhancement attributed to fractures


14 SPE-176101-MS

Fig. 13 Simulation results (water rate) on two wells with early water breakthroughs with permeability enhancement attributed to
fractures
SPE-176101-MS 15

Fig. 14 The highly fractured platform B area where relative permeability modification applied and gives improvement in overall
history match