FRACTURED BASEMENT DELINEATION USING SEISMIC MULTI-ATTRIBUTES; THE MLB FIELD, JAMBI, INDONESIA

Report No.: GPM 12/09

By Oki Irawan Agung Sugiri Msc. (Geophysics) Curtin

This report is presented as part of the requirement for the units Geophysics Project 634 Part A & B, units totalling 50 credit points in the Master of Science (Geophysics) from Curtin University of Technology. The work is the result of supervised research; however, the report has been prepared by the student who is solely responsible for its contents.

DEPARTMENT OF EXPLORATION GEOPHYSICS Curtin University of Technology February 2010

Abstract

The basement area within Muara Bulian (MLB) Field, Jambi, Indonesia is predicted to have new hydrocarbon deposits. Some hydrocarbons have been located in fractured zones. Given that basement fracture hosted hydrocarbons do exist, it is important to investigate how seismic data can be processed to assist in delineating these areas.

The primary data sets used in this project were wire line logs from two wells that penetrate the basement and a 3D post stack seismic volume. The basement consists of variably fractured metasediments such as phyllite and schist. Fluid replacement modelling was completed to examine the effect of fluid type on the seismic signature. It was shown that a change from gas to brine water in the fractured interval would likely cause a significant decrease in seismic amplitude. Analysis of the synthetic traces from the fluid replacement modelling indicated that energy attribute could be effective in delineating fracture zones.

The instantaneous frequency attribute shows a low frequency zone in the upper basement around well BL-1. I predict that this low frequency response is associated with the known gas deposits in the fractured interval. Based on this observation, I used the instantaneous frequency to assist in imaging and predicting the spatial extend of gas filled fractures.

An assessment of coherence and most positive curvature attribute was completed. The analysis was intended to establish how effectively these attributes map the faults that are predicted to be associated with fractured zones. An important part of assessing these more structurally oriented attributes was the creation and assessment of dip steering cubes and fault enhancement filters (i.e. median and diffusion filters). It was demonstrated that imaging of structurally oriented attributes was markedly improved where dip steered was applied. That is, structural attributes generate a clearer and sharper image with dip steering than where dip steering has not been applied. Fault enhancement filter application appeared to reduce both random noise and noise related to the acquisition footprint in resulting images. In summary, the use of a steering cube and fault enhancement filters can improve the imaging of structurally oriented attributes such as coherence and most positive curvature.

Multi-attribute analysis in the MBL field has resulted in new insight into the location of possible gas filled fracture zones. These areas have high energy and low coherence value. The high energy values are predicted to be associated with low velocity related to gas filled fracture zones in the upper basement. Low coherence values are predicted to be associated with major and minor faults that relate to the fractures. Zones of low coherence and high energy at basement high proximal to the centre of the study area are identified as being prospective for gas accumulation in fractures.

ii

Other attributes (i.e. instantaneous frequency and most positive curvature) are not as effective as the combination of the energy and coherence for direct delineation of gas filled fractures. Instantaneous frequency appears to be sensitive to both gas and brine filled fracture zones. In this sense the instantaneous frequency is not effective for direct indication of gas in fractures zones. Most positive curvature seems to overestimate the size of faults zones. However the most positive curvature attribute appears to have been effective in showing fault orientation in the South Eastern part of the study area where the coherence failed to unambiguously show faults orientation.

Multi-Attributes analysis was shown to be effective enhanced imaging and analysis of fractured basement meta-sediments at the MLB field. Follow up work is required to verify this prediction. The multi-attributes analysis can assist the interpreter in making predictions and focusing exploration and possibly production activities.

iii

ACKNOWLEDGEMENTS

All praise to the almighty Allah SWT who has devoted his favour without stopping, so the writer was able to complete this thesis.

I would like to extend my gratitude to PT. PERTAMINA who has graciously provided the data for this study and approving its publication. Thank you to Curtin University of Technology, Department of Exploration Geophysics for providing the computer, software, and administrative facilities. Sincere thanks to my university supervisors, Dr. Brett Harris and Associate Professor Milovan Urosevic for their guidance, support, and time.

I would like to say thank you to my supervisor from PT. PERTAMINA, Mrs. Artini Soekotjo.

A special acknowledgment goes to my family, who are always giving me the spirit, support and inspiration every day.

iv

Table of Contents

Abstract .......................................................................................................................... i ACKNOWLEDGEMENTS ......................................................................................... iv Table of Contents .......................................................................................................... v Table of Figures .......................................................................................................... vii Chapter 1 Introduction .................................................................................................. 1 1.1 The Issue ............................................................................................................. 2 1.2 Research objectives ............................................................................................. 3 1.3 Study Area .......................................................................................................... 3 1.4 Work Program ..................................................................................................... 3 1.5 Thesis Layout ...................................................................................................... 4 Chapter 2 Regional Geology ......................................................................................... 6 2.1 Tectonic Setting .................................................................................................. 6 2.2 Stratigraphy ......................................................................................................... 9 Chapter 3 Methods ...................................................................................................... 13 3.1 Seismic Signatures of Fractures ........................................................................ 13 3.2 Other Geological Features Related to Fractures ............................................... 16 3.3 Seismic Attributes ............................................................................................. 16

3.3.1 Energy Attribute......................................................................................... 17 3.3.2 Instantaneous Frequency ............................................................................ 18 3.3.3 Coherence .................................................................................................. 19 3.3.4 Most Positive Curvature ............................................................................ 19 Chapter 4 The Application of Multi-Attributes Analysis ........................................... 21 4.1 Data ................................................................................................................... 21 4.2 Fluid Replacement Modelling........................................................................... 22 4.3 Seismic Pre-Processing ..................................................................................... 26 4.3.1 Steering Cube ............................................................................................. 26 4.3.2 Faults Enhancement Filter ......................................................................... 29 4.4 Attributes Analysis ............................................................................................ 32 4.4.1 Energy ........................................................................................................ 32 4.4.2 Instantaneous Frequency ............................................................................ 35 4.4.3 Coherence .................................................................................................. 38 4.4.4 Most Positive Curvature ............................................................................ 41 4.4.5 Combination of Energy and Coherence Attributes .................................... 44 Chapter 5 Conclusions ................................................................................................ 47 Chapter 6 Recommendations ...................................................................................... 50 Reference .................................................................................................................... 51

vi

Table of Figures

Figure 1, South Sumatera, structural network............................................................... 8 Figure 2, Stratigraphic column for Jambi sub-basin ................................................... 12 Figure 3, Illustration of possible pore-type effects on P-wave velocity. .................... 14 Figure 4, Amplitude spectra of incident and received waveforms (P-wave) for model that consists of four vertically aligned fractures (left) and for model that consists of seven vertically aligned fractures (right). ................................................................... 15 Figure 5, An illustrated definition of 2D curvature. ................................................... 20 Figure 6, Schematic work flow of the project............................................................. 21 Figure 7, Basement time structure map. ..................................................................... 24 Figure 8, Fluid replacement modelling reference model derived from wire-line logs (i.e. gas filled fractures). ............................................................................................. 25 Figure 9, Fluid replacement model for brine water. ................................................... 26 Figure 10, The upper image show the similarity attribute at the basement horizon without use of a dip steering cube. The lower images is the similarity attribute where a dip steering cube has be designed and used. ............................................................ 28 Figure 11, Seismic section at cross-line 5340 before fault enhancement filtering. .... 30 Figure 12, Seismic sections at cross-line 5340 after fault enhancement filter. ........... 31 Figure 13, Seismic section at line A-B with the rainbow colour indicating high energy anomalies. ................................................................................................................... 33

vii

Figure 14, Energy attribute image on the top basement horizon ................................ 34 Figure 15, Seismic section at line A-B, with the rainbow colour indicating the low instantaneous frequency anomaly ............................................................................... 36 Figure 16, Averaged instantaneous frequency attribute image on the top basement horizon. ....................................................................................................................... 37 Figure 17, Seismic section at line A-B, with the dark colour shows the low coherence value.. .......................................................................................................................... 39 Figure 18, Coherence attribute image on the top basement horizon........................... 40 Figure 19, Seismic section at line A-B with the dark-yellow colour indicating high most positive curvature.. ............................................................................................. 42 Figure 20, Most positive curvature attribute image on the top of basement horizon . 43 Figure 21, Coherence and energy attribute image on top of basement ....................... 45 Figure 22, Energy attribute image on the top basement horizon and a 3D cube of similarity attribute.. ..................................................................................................... 46

viii

Chapter 1 Introduction

The demand of fuel energy in the world is increasing. This is because of increasing global population and increasing demand for fuel in developing nation. In order to meet the demand, discovery of new hydrocarbon reserves are required.

Seismic is the main geophysical method in the oil and gas industry. This has provided the basis for rapid development of seismic acquisition, processing and interpretation methods. One of the newer seismic processing/interpretation technique is the application of seismic attributes.

Seismic attribute analysis is now commonly used at the interpretation stage. Seismic attributes are defined by Brown (2000) as a derivative of the basic seismic measurement.

The basement area within MLB Field is predicted to have potential new hydrocarbon deposits. Some hydrocarbons have being located in fractured zones. Given that basement fracture hosted hydrocarbons do exist, it is important to investigate how seismic data can be processed to illuminate these potentially important areas. If post stack processing can be tailored for illuminating fractured zones in basement, then results can be combined with wire-line logging and geological interpretation to assess potential for economic hydrocarbon accumulations. 1

Finding the optimum seismic attributes is the key to delineate fractured zones, because fractures can have many seismic attributes. The optimum attributes that can be used to delineate fractures will vary from site to site depends on the rock type that host these fractures, fracture geometry, fluid type that fills the fracture, and acquisition geometries.

In this project, I analyse several attributes that I think are related to the fractures in this field. The attributes are energy attribute, instantaneous frequency, coherence and most positive curvature. The energy and instantaneous attribute anomaly can show the gas accumulation within the fractured zones, while coherence and most positive curvature can show the location of faults zone that is predicted to be related to the fractures.

1.1 The Issue

This project will investigate the application of post stack seismic attributes for delineating gas filled fracture zones in Metasediments. The data used in this project is 3D post stack seismic data and wire-line log data from the wells; ML-1 and BL-1. Both wells penetrate the basement. The project sponsor is PERTAMINA.

1.2 Research objectives

The objective of this project is to delineate fractured basement area in MLB field, Jambi using seismic attributes with 3D stacked seismic reflection data as the input. The information obtained from this project can be used to help PERTAMINA in finding new hydrocarbon reserves in the MLB field.

1.3 Study Area

Study area of this project is the MLB field, located in Jambi Province, Indonesia. MLB field is within Jambi sub-basin which is a part of South Sumatera basin.

1.4 Work Program

This project work program consisted of the following: Literature review on the use of seismic attributes in fractured rock setting and geology of the site (i.e. includes all geology and geophysics completed at the site). Learn to use all the software required for project completion (e.g. HRS and OpendTect). Data analysis and pre-processing.

Identify sub-volumes for attribute testing and analysis. Qualitative seismic attributes analysis on sub-volumes Analyse the combined multiple attributes (multi attributes) analysis for the sub-volumes.

Semi-Quantitative analysis of attribute sub-volumes (i.e. compare with wire line logging).

Apply selected attributes and or combined attributes to complete volume and analyse results.

1.5 Thesis Layout

This thesis consists of 6 chapters. These chapters are briefly described below.

Chapter 1 Introduction This chapter discusses the background information of the project. It consists of literature review, project objectives, and work program.

Chapter 2 Regional Geology This chapter consists of geology information of the study area, such as: tectonic setting, stratigraphy and depositional history.

Chapter 3 Methods This chapter discusses the methods that were used in the project.

Chapter 4 Multi-Attributes Analysis This chapter contains the analysis of the result of the multi-attributes method application.

Chapter 5 Conclusions

Chapter 6 Recommendations

Chapter 2 Regional Geology

Jambi sub-basin is a part of South Sumatera Basin, Located in the South Eastern part of Sumatera Island. South Sumatera basin is a tertiary back arc basin. This area is limited by Sunda shelfs depositional area in the North East, Lampung high in the South and South Eastern edge, and Central Sumatera basin in the North West.

2.1 Tectonic Setting

Pulunggono et. al. (1992) stated that the South Sumatra basin was formed by three major tectonic phases:

1. Eocene Oligocene Rift phase with extensional force formed N-S trending grabens that controlled Pre-Talang Akar deposition.

2. Early Miocene Mid Miocene South Sumatra Basin changed its status from rift basin to back arc basin. This phenomenon was showed by the occurrence of basin subsidence which followed by the sedimentation of Talang Akar, Gumai, Air Benakat, and Muara Enim Formations. Tiga Puluh Mountains lifting at this tectonic phase

caused rifting to small grabens that were formed in the previous tectonic phase. NW - SE trending Folds also start to formed in this tectonic phase.

3. Pliocene Pleistocene Pliocene Pleistocene was a compression phase, which formed folds and thrusts with NE-SW trends. This compression also reactivated many normal faults and some of these normal faults even reversed.

In Jambi sub-basin, half grabens structure is controlled by NE-SW trending faults with the major faults general dip direction is to the SE.

Figure 1, South Sumatera, structural network. Jambi Sub-Basin is showed by the red box (PERTAMINA).

2.2 Stratigraphy

Stratigraphy in Jambi sub-basin based on Pulunggono et. al. (1992) consists of:

Basement (pre-Tertiary), consist of Paleozoic metamorphic and carbonate rocks, and also Mesozoic igneous, metamorphic and carbonate rocks. Paleozoic and Mesozoic metamorphic and sedimentary rocks have undergone folding, faulting and intruded by igneous rocks during Mesozoic volcanism.

Lahat Formation was the first sedimentary unit deposited in this basin. Lahat Formation only deposited in the deepest part of the basin. This formation deposited during Oligocene and consists of coarse fragmented sharp angled volcanic material such as breccias, andesite, and agglomerate at the lower part, and softer grain, such as shale with tuff, silt and coal intercalation at the upper part. This formation is predicted to be deposited in the continental fresh-brackish water environment.

Talang Akar Formation deposited unconformably above Lahat Formation and basement. This Formation deposition is assumed to have happened in Late Oligocene Early Miocene. The lower part of Talang Akar Formation, called

Gritsand Member (GRM), deposited only in the paleo-deep area. GRM is dominated by thick coarse grain feldspartic arenite sandstones, and thin shale, silt, and coal. GRM is deposited in fluvio-deltaic environment. The upper part of Talang Akar Formation, called Transition Member (TRM), consists of sandstone, shale, and thin bed coal. TRM is deposited in paralic-litoral to distal marine environment. Only TRM member that can be found in the Jambi sub-basin.

Baturaja Formation consists of reef limestone and limestone shale bedding. This Formation, which deposited during Miocene, only deposited in the Eastern part of Jambi sub-basin. This fact shows that in general, the Eastern part of Jambi sub-basin is shallower than the Western part of the area.

Gumai Formation, which deposited conformably above Baturaja and Talang Akar Formation during Early Miocene-, consists of brownish-black marine shale and marl with many foraminifera fossils. This formation was deposited in deep marine environment.

Air Benakat Formation was deposited conformably above Gumai and Baturaja Formation. This Formation consists of claystone and shale. Air

10

Benakat Formation was deposited in neritic - littoral environment during Mid Miocene Late Miocene.

Muara Enim Formation was deposited conformably above Air Benakat Formation. This formation consists of coal and sandstone bedding, with thin shale, claystone, and clay. This Formation was deposited in shallow marine paralic during Late Miocene Early Pliocene.

Kasai Formation was deposited during Late Pliocene Early Pliocene. This Formation consists of pumice tuff, tuff sandstones, claystone and gravel tuff. This formation was deposited during Early Pliocene Early Quaternary. Only found in very limited areas in Jambi sub-basin.

11

Figure 2, Stratigraphic column for Jambi sub-basin (PERTAMINA).

12

Chapter 3 Methods

3.1 Seismic Signatures of Fractures

The objective of this project is to delineate the fractured zone in the basement area for the MLB field.

Fractures can cause low P and S-wave velocities (Vp and Vs), anomalous reflectivity (seismic impedance), low Q (high seismic attenuation), low instantaneous frequency, anomalous AVO, and azimuthal variation in velocity. Because of this, fractures can have many seismic attributes.

The optimum attributes that can be used to delineate fractures will vary from site to site. This site to site variation is caused by differences in the rock type that host these fractures, fracture geometry, fluid type/distribution, and acquisition

geometries/parameters.

Fractures can lower seismic velocity and changes the Vp/Vs ratio. Fractures lower seismic P and S-wave velocity due to the rocks secondary porosity that was caused by the fractures. As the fracture density increases, P-wave velocity decreases. A plausible explanation for this is that, as fracture density increases, the overall velocity

13

becomes increasingly more sensitive to the fracture properties and less sensitive to the intact rock properties (Boadu, 1997) (see figure 3). But if the fractures is filled with minerals (not fluid), the changes in Vp and Vs is depend on the elastic properties of the minerals which filled the fractures. Compressional wave velocity (Vp) for hydrocarbon (HC) filled fractures is lower than brine filled fractures. While Vs of HC filled fractures will be slightly higher than brine filled fractures (Hilterman, 1977).

Figure 3, Illustration of possible pore-type effects on P-wave velocity. The dashed curves above the limestone reference curve indicate increased fractions of stiff pores, and those below indicate increased fractions of cracks/fractures, (Xu and Payne, 1999).

14

Fractures also lower seismic frequency and quality factor (Q). The fractures cause a time delay in the propagating waveform and act as filter by attenuating the high frequency components in the spectrum of the waveform (Boadu, 1997). Figure 4 show that as number of fractures increase, the seismic amplitude is significantly reduced and there is also a shift in the peak amplitude toward lower frequencies. The presence of fractures in a rock mass constitutes one of the major causes of scattering of body and surface seismic waves (Lerche and Petroy, 1996, opcite, Boadu, 1997). This scattering process causes effective attenuation, usually expressed by Q, of the propagating waves (see Figure 4 below).

Figure 4, Amplitude spectra of incident and received waveforms (P-wave) for model that consists of four vertically aligned fractures (left) and for model that consists of seven vertically aligned fractures (right) (Boadu, 1997). As the number of fractures increase, the amplitude of the received waveforms is significantly reduced and there is also a shift in the peak amplitude toward lower frequency.

15

3.2 Other Geological Features Related to Fractures

There are some geological features that are typically related to fractures. Sometimes it is easier to detect these geological features than the fractures. These geological features are lithofacies and structural geology features such as large scale faults and folds.

Fractures often prefer certain lithofacies such as rocks with high Vp and Impedance such as limestone. Fractures that related to faults and fold is classified as tectonic fractures. These tectonic fractures formed in networks with specific spatial relationship to folds and faults.

3.3 Seismic Attributes

Seismic attributes have become an integral part of seismic interpretation. Since their introduction in 1970s, various methodologies have been developed for seismic attributes application to help seismic interpreter to make the best decision in hydrocarbon exploration and field development project. Complex trace attributes. Seismic inversion, response attributes, coherence, spectral decomposition, and the use of neural network are several examples of seismic attribute developments. Now,

16

Seismic attributes are being used widely in oil industry for reservoir lithological and petrophysical prediction.

Brown (2000) defined seismic attributes as a derivative of basic seismic measurement. All of the attributes are not independent one of another. The difference is only in the way each of the attributes analyse the basic information of the involved seismic wave and their appearance. Seismic attributes are classified based on the basic information of seismic used in the calculation of the attributes. The basic information of seismic are time, amplitude, frequency and attenuation.

In this project, I use several attributes that can detect fractured zones in the basement. These attributes are energy, instantaneous frequency, coherence, and most positive curvature. All of these attributes are computed using OpendTect software.

3.3.1 Energy Attribute

Energy is a seismic attribute that returns the energy of a trace segment (dGB Earth Sciences, 2009). This attribute calculates the squared sum of the sample values in the specified time-gate divided by the number of samples in the gate. The Energy is a measure of reflectivity in the specified time-gate. The higher the Energy, the higher the Amplitude. This attribute enhances, among others, lateral variations within

17

seismic events and is, therefore, useful for regional correlations of major reflectors, exposing the reflectors difference due to significant change in depositional environment and early indication DHI bright spot or dim spot phenomena which indicated by a significant decrease or increase of acoustic impedance.

3.3.2 Instantaneous Frequency

Instantaneous frequency attribute outputs the instantaneous frequency at the sample location (dGB Earth Sciences, 2009). The instantaneous frequency attribute responds to both wave propagation effects and depositional characteristics, hence it is a physical attribute and can be used as an effective discriminator. Instantaneous frequency is useful for hydrocarbon indicator, which is indicated by the low instantaneous value, fracture zone indicator, since fractures may appear as lower frequency zones, and bed thickness indicator, higher frequencies indicate sharp interfaces such as exhibited by thinly laminated shales, while lower frequencies are indicative of more massive bedding geometries, such as sand-prone lithologies.

Energy and instantaneous frequency are complex seismic attributes. Tanner and Sheriff, 1997 provide detailed information about complex seismic attributes such as energy, instantaneous phase etc.

18

3.3.3 Coherence

Coherence is a multi-trace attribute that emphasizes areas of highest discontinuity (dGB Earth Sciences, 2009). Coherence or Coherency is a term used for a group of algorithms establishing how much adjacent traces are alike. OpendTect implemented Coherency type 1, the simplest. This type shifts one of the traces up or down to find the maximum cross-correlation. This attribute thus enhances variations in seismic data related to changes in continuity among neighbouring traces. This attribute can be used to enhance structures, such as faults and fractures, but it can also be used to delineate specific lithologies like salt and basement, which both have chaotic seismic character.

3.3.4 Most Positive Curvature

Curvature can be defined as the reciprocal of the radius of a circle that is tangent to the given curve at a point (Chopra et.al, 2006). Mathematically, curvature may simply be defined as a second order derivative of the curve (Chopra et.al, 2006). In the OpendTect software, the most positive curvature attribute is computed using the dip steering cube as the input. The advantage of using dip steering cube is minimizes wrong curvature value due to miss interpretation of the horizon. Most positive

19

curvature is derived by searching the normal curvatures for the most positive values. Most positive curvature helps in small fault interpretation.

Figure 5, An illustrated definition of 2D curvature. Anticlinal features have positive curvature, synclinal features have negative curvature and planar features (horizontal or dipping) have zero curvature.

20

Chapter 4 The Application of Multi-Attribute Analysis

The work flow for this project is showed in the Figure 6.

Figure 6, Schematic work flow of the project.

4.1 Data

The data used in this project are a 3D seismic volume that consists of 885 inlines and 457 crosslines, and wire-line log data from 2 wells (well BL-1 and MB-1) that penetrate the basement. First, I completed the seismic to well tie. Then the basement

21

top horizon was picked troughout the volume. Figure 7 shows the time structure map of the picked top of basement horizon. Next I analysed the time structure map of the top of the basement. I decided to reduce the seismic volume to include only target area, which is around the highest part of the basement structure. This target area consists of 600 inlines and 280 crosslines.

4.2 Fluid Replacement Modelling

A simple fluid replacement modelling (FRM) was completed for the wire-line log data in the well BL-1. The BL-1 velocity log provided in Figures 8 and 9 show several very low velocity intervals below the top of basement that were interpreted as gas filled fractures (note some indication of gas was recovered drill stem test (DST)) This FRM was completed to examine any change in the seismic signal in the fractured interval if the fractured interval was filled with brine water. The basement lithology is described as metasediment such as schist & phyllite. Such rocks may consist of both quartz and/or platy minerals (i.e. like muscovite, chlorite and biotite).

The results of the FRM as shown in Figure 7 and 8 indicate significant amplitude change if gas is replaced by brine. Based on this result, it is reasonable to conclude that there may be a high amplitude anomaly if the fractures are filled with gas.

22

Figure 7, Basement time structure map. Line A-B is marked in Blue. The line has been selected to intersect wells MB-1 and BL-1. Line A-B is used for evaluation of attributes as in Figures 13, 15, 17 and 19

23

Figure 8, Fluid replacement modelling reference model derived from wire-line logs (i.e. gas filled fractures). Figure shows P-wave and computed acoustic impedance log plus synthetic and real seismic traces at well BL-1 before FRM. The Red circle shows the high amplitude at the top basement interface (i.e. high acoustic impedance contrast). 24

Figure 9, Fluid replacement model for brine water. Figure shows P-wave and computed acoustic impedance log along with synthetic and real seismic traces at well BL-1 after FRM (brine case). Red circle show the low amplitude of the top basement interface. Notice the significant change in the seismic amplitude due to change in fluid type. 25

4.3 Seismic Pre-Processing

Before the seismic data is used to generate the attributes, there are two pre-processing steps that should be completed. These are:

1. Creating steering cube and, 2. Faults enhancement filter process.

The impact of both pre-processing steps are examined.

4.3.1 Steering Cube

A steering Cube is a cube that contains the dip of the seismic events in inline and crossline direction at every sample point (dGB Earth Sciences, 2009). These dip attributes are guided along a three-dimensional surface on which the seismic phase is approximately constant. The Steering cube is made to generate dip attributes (the dip in the steering cube itself is an attribute), to correct other attributes for structure, and to be used for structurally oriented filtering.

Several steps are required to generate the steering cube. First, a raw steering cube is generated as an input to the median filter process. Then I complete the median filter

26

process. The median filter process consists of 3 steps i) detailed steering, ii) intermediate steering, and iii) background steering (see www.opendtect.org) . The final result is used to generate attributes for structures, such as most positive curvature. Figure 10 shows the similarity attribute using the steering cube. It shows clearer and sharper images of the faults compared to the un-steered similarity cube.

27

Figure 7, The upper image show the similarity attribute at the basement horizon without use of a dip steering cube (upper). The lower images of the similarity attribute where a dip steering cube has been used. Application of dip steering resulted in a clearer and sharper image of the faults.

28

4.3.2 Faults Enhancement Filter

Fault enhancement filtering is a process in OpendTect that is designed to sharpen the faults and suppress non-faults discontinuities, specifically non-geological

discontinuities such as random noise and acquisition footprint (dGB Earth Sciences, 2009). This filter actually consists of two filters; (i) diffusion type filter and (ii) median dip filter. The diffusion type filtering migrates the amplitude information on both sides of the fault toward the fault. This filter is used to sharpen the contrast at the fault. While median dip filtering removes spiky noises by applying an N-point median operator. The spiky noise is removed while trends are preserved.

A cut off value is established from the similarity attribute or another fault attribute. The fault enhancement filter applies the diffusion type filter to the areas which have similarity value below the established cut off value. While in the areas which have similarity values above the established cut off, the median dip filter is applied.

Figure 12 shows that after the fault enhancement filter application the seismic image shows less noise than before (Figure 11). The fault (red circle area) looks sharper and clearer after fault enhancement filter. This will make faults interpretation easier and the results of attribute structures clearer and sharper.

29

Figure 8, Seismic section at cross-line 5340 before fault enhancement filtering. Red circle shows one of the major faults in the study area.

30

Figure 9, Seismic sections at cross-line 5340 after fault enhancement filter. Figure 12 shows less noise than Figure 11. The fault (red circle) looks sharper and clearer after fault enhancement filter.

31

4.4 Attributes Analysis

4.4.1 Energy

Energy attributes are obtained using the basic 3D seismic volume as input. The time gate selected was 0 25ms qualitatively, because it provided the best image.

Figure 13 shows high energy values mapped over the basic seismic image. The high energy zones are marked out by a black ellipse proximal to the top of basement near well BL-1. This high energy anomaly is related to elevated acoustic impedance contrast associated with low velocity zone immediately below the top of the basement. This low velocity interval is likely to be associated with gas filled fractures (see velocity logs in Figure 8 and 9 and FRM in section 4.2).

From Figure 14 we can see that these high energy areas are localized in the higher part of the meta-sedimentary basement. There is a high energy zone close to well BL1. This zone is where the well BL-1 intersected a gas filled fractured zones in basement. It would seem reasonable that this high energy zones is related to gas filled fractures in the top of the basement around the well BL-1. More work is needed to confirm this hypothesis.

32

Figure 10, Seismic section at line A-B with the rainbow colour indicating high energy anomalies. The blue line shows the top basement horizon. The logs on the left side of both wells show the resistivity wire-line logs and the right log of well BL-1 shows P-wave logs. High resistivity values are indicated by yellow to purple on the left. Low velocity is indicated by yellow purple colour on the right of BL-1. The Black ellipse marks out the high energy anomaly at the basement interface near well BL-1.

33

Figure 11, Energy attribute image at the top basement horizon (i.e. the attribute are shown on time structure surface for top of basement).

34

4.4.2 Instantaneous Frequency

The instantaneous frequency attribute is generated using the standard 3D seismic volume as input. The instantaneous frequency is averaged within 100 ms time windows. The relatively long window is selected to capture the instantaneous

frequency attribute below the top of basement interface. That is, the low instantaneous frequency anomaly is expected to occur within the upper basement zone but not at the interface itself.

It can be observed in Figure 15 that there is a low value for instantaneous frequency immediately below the basement horizon as marked out by the black ellipse. This low value of instantaneous frequency appears to be located in the expected gas filled fractures zones.

Figure 16 indicates that the low frequency anomaly is not effective when compared to energy attribute (i.e. see Figure 14) in predicting the location of gas filled fractures. The instantaneous frequency attribute shows no clear pattern. The weakness in the instantaneous attribute may be that it is sensitive to both gas and brine filled fracture zones.

35

Figure 12, Seismic section at line A-B, with the rainbow colour indicating the low instantaneous frequency anomaly. Blue line shows the top basement horizon. The logs on the left side of both wells show the resistivity logs value. High resistivity values are indicated by yellow to purple on the left. Red ellipse marks out a low frequency anomaly below the basement interface.

36

Figure 13, Averaged instantaneous frequency attribute image at the top basement horizon (i.e. the attribute are shown on time structure surface for top of basement).

37

4.4.3 Coherence

The Coherence attribute is generated using a fault enhancement filtered volume as input. The time gate used for coherence was 0 25 msec.

Figure 17 shows that low coherence values delineate the major fault (see red ellipse). Low coherence also delineates minor faults (see red box). There are main low coherence features at the basement. This is likely the result of the older metasediments being more highly deformed.

From Figure 18 shows an image of coherence at top of basement (defined as two way travel time to top of basement). We see the major and minor faults (expressed as low coherence value) more clearly using coherence attribute than using the standard seismic. The faults generally have NE-SW or NW-SE trends. However we cannot clearly see the fault trend in the South Eastern part of the study area.

38

Figure 14, Seismic section at line A-B, with the dark colour shows the low coherence value. Blue line shows the top basement horizon. The logs on the left side of both wells show the resistivity logs value. High resistivity values are indicated by yellow to purple on the left. Red ellipse shows major fault near well MB-1. Inset (red box) shows a minor fault that crossing well BL-1.

39

Figure 15, Coherence attribute image at the top basement horizon (i.e. the attribute are shown on time structure surface for top of basement).

40

4.4.4 Most Positive Curvature

The most positive curvature is generated using the steering cube as input. Figure 19 shows that the most positive curvature also can delineate faults (red ellipses), even though the result is not as detailed as coherence. From Figure 20, we can see that most positive curvature seems to overestimate the fault zones. Despite this we can see the fault trends more clearly for the coherence attribute. This is especially true in the South Eastern part of the study area. The trend of the faults is still NE-SW or NWSE. The fault patterns obtained from most positive curvature attribute is similar to the structural trend of the Jambi sub-basin (e.g. as explained by Pulunggono, et. al. 1992).

41

Figure 16, Seismic section at line A-B with the dark-yellow colour indicating high most positive curvature. Blue line shows the top of basement horizon. The logs on the left side of both wells show the resistivity wire-line log. High resistivity values are indicated by yellow to purple on the left. Red ellipses mark out major and minor faults near well MB1 and BL-1.

42

Figure 17, Most positive curvature attribute image at the top of basement horizon (i.e. the attribute are shown on time structure surface for top of basement). Interpreted faults patterns showed by black arrows.

43

4.4.5 Combination of Energy and Coherence Attributes

The combination of energy and coherence attributes is completed to determine the potential fractured zone areas which contain gas accumulation. The areas which have high energy and low coherence are believed to have a higher potential to contain gas filled fractures. There are several areas that have high energy and low coherence. The most interesting of these is the area South West of well BL-1. This area is marked by a blue box in Figure 21. The energy values are high, which indicates possible gas accumulation and the coherence values are low, which indicates that there are some faults in the area. Also the areas depth is the shallowest, compared to other area that have the same anomaly. From Figure 22 we can see that there is at least one fault (red ellipse), associated with a low coherence value, which cuts through the basement, and seems to intersect the high energy value (blue ellipse) located around this fault.

44

Figure 18, Coherence and energy attribute image at top of basement (i.e. the attribute are shown on time structure surface for top of basement). White colour shows the low coherence value and rainbow colour shows high energy value.

45

Figure 19, Energy attribute image at the top basement horizon and a 3D cube of similarity attribute. Rainbow colour shows high energy value and dark green colour shows the low coherence value. Low coherence area (red ellipse) indicates a fault that cut through the basement. Black ellipse shows a high energy area around that fault.

46

Chapter 5 Conclusions

Detailed research into seismic attribute analysis for the delineation of gas filled fractures in meta-sedimentary rocks has been completed. There are a number of important outcomes from the work completed. They are provided below:

Wire-line logs from the study site show a strong and characteristic signature from interpreted gas filled fracture zones. Lower velocity in the gas filled fractures within the upper metasediments result in a large acoustic impendent contrast.

Fluid replacement modelling shows that a change from gas to brine water in the fractured interval would likely cause a significant decrease in seismic amplitude at the interface.

I demonstrate the use and design of steering cubes. Correctly designed steering cubes are shown to improve imaging with fault enhancement filters. Furthermore the use of dip steering cubes improves imaging of structurally oriented attributes such as most positive curvature.

47

Energy attribute can be used to delineate the gas filled fractures because gas will likely cause high energy in the top of the basement, as demonstrated by the fluid replacement modelling.

Coherence attribute can delineate the major and minor faults that are predicted to be related to the fractured zones.

Instantaneous frequency is not effective for direct indication of gas in fractures zones. This is because for this study area, instantaneous frequency appears to be sensitive to both gas and brine filled fracture zones in basement.

Most positive curvature can delineate major and minor faults. However the resulting images are not as detailed as those generated from the coherence attribute. Most positive curvature also appears to overestimate the fault zones.

The fault patterns obtained from most the most positive curvature attribute is consistent with the structural trend of the area.

48

At the centre of the MBL study area there are several zones with high energy and low coherence. These tend to be related to fractures around faults, which have the potential to contain gas.

This research clearly demonstrates the value of multi-attribute analysis. The results have the ability to direct the interpreter to early and more robust predictions for the location of gas filled fractures in basement in MLB field development plan. Further work is required to further develop the concepts studied in this work.

49

Chapter 6 Recommendations

Even though energy attribute can show the gas accumulation, they cannot show the saturation of the gas filled fractures. AVO attributes, obtained from pre-stack seismic data, might provide the gas saturation information. Analysis of azimuthal variation (anisotropy) in velocity and AVO may provide more detailed information about the fractures orientation and fluid content. More detailed rock physics analysis will also help predict the seismic response caused by gas and brine water filled fractures.

50

Reference

Aguilera, A., 1998, Geologic aspects of naturally fractured reservoirs. The Leading Edge, 17, 1667-1670. Boadu, F. K., Fractured rock mass characterization parameters and seismic properties: analytical studies. Journal of Applied Geophysics. Vol. 36, page 119. Brown, A.R., 2000, Interpretation of three-dimensional seismic data. AM. Assoc, Pet. Geol. Memoir 42. Chopra, S., Marfurt, K., and Alexeev, V., 1996, Practical aspects of curvature computations from seismic horizons. SEG Expanded Abstract. de Coster, G. L., 1974, The geology of the Central and South Sumatra Basins: Proceedings Indonesian Petroleum Association Third Annual Convention, June, 1974, p. 77-110. dGB Earth Sciences, 2009, OpendTect User Documentation version 4.0. Pulunggono, A., Haryo, S. Agus, and Kosuma, C. G., 1992, Pre-Tertiary and Tertiary fault systems as a framework of the South Sumatra Basin; a study of sar-maps: Proceedings Indonesian Petroleum Association Twenty First Annual Convention, October, 1992, page 339-360. Taner, M. T. and Sheriff, R. E., 1997, Application of amplitude, frequency and other attributes to stratigraphic and hydrocarbon determination, in Seismic Stratigraphy application to hydrocarbon exploration. AAPG Memoir. Vol. 26.

51