Reservoir delineation beneath a heterogeneous

SPECIAL TOPIC: MARINE SEISMIC

Reservoir delineation beneath a heterogeneous shallow gas overburden using ‘True-3D’ seismic imaging approaches A.R. Ghazali1*, N. ElKady1, M.F. Abd Rahim1, R.J.J. Hardy1, F.S. Dzulkefli1, S. Chandola1, S. Kumar1, S. Shukri1, S.M.T. Mohi Eldin1, S.S. Elkurdy1, N.L. Rafiuddin1, S. Maitra2, F.F. Basir 2, M.L. Ghazali2, M.H.F. Abdul Latib2 and S. Zainal2 describe a high-density experimental project to image the reservoir from both above and within existing boreholes. Introduction Imaging through a heterogeneous shallow gas-charged overburden, such as a gas cloud, presents several imaging challenges and is a demanding problem to solve. Our preferred technical solution for imaging beneath gas clouds is to utilize converted wave imaging (Radzi et al., 2015), but this is not always available or cost effective and velocity model building is still difficult. Many previous case studies have been produced from Malaysia which demonstrate subsurface imaging techniques and improvements for fields affected by gas clouds, e.g., Akalin et al. (2010); El Kady et al. (2012); Abd Rahim et al. (2013); Ghazali et al. (2016) and Gudipati et al. (2018). In this paper, we describe a new comprehensive high-density experimental project to readdress these ever-challenging seismic issues by imaging the reservoir from both above and within existing boreholes. The integration of multiple technologies has significantly improved the subsurface

images of the field including better-quality velocity models below gas clouds. The new data reveal a larger scale of near-surface heterogeneities than previously expected and future studies will selectively reprocess subsets of the acquired data in order to optimize the images; and, by extension to other similar fields, address a cost-effective imaging strategy. The Field B with an average water depth around 70 m located 45 km offshore Sarawak, Malaysia, has been in oil production since 1971. The structure is a gentle domal uplift with a collapsed crest superimposed on to a rollover anticline resulting from growth faulting and is located in part of the Upper Tertiary Baram Delta province. The legacy 3D seismic data acquired in 1991 is good to very good on the flanks of the Field B structure but deteriorates rapidly towards the crest where the Field B is located as shown in Figure 1a. The reasons for this are multi-fold: firstly, the drop in quality is caused by outdated

Figure 1 a) Conventional 3D Anisotropic PreSDM section of the Field B reprocessed in 2015 from a vintage 1991 surface seismic survey. Effects of gas clouds are strongly observed to hamper the subsurface images, especially in the centre of the section where the main reservoirs and many major faults are interpreted. Reacquisition of the field was completed in 2016 using the most modern marine seismic acquisition techniques available. b) Shallow gas cloud affected area on the field is illustrated by the dashed red outline polygon. The smaller inset shows the subsurface earth model where multiple depth layered grey polygons represent shallow gas bodies created from the extraction of near-surface high amplitudes.

Exploration Technology, Group Research and Technology, PETRONAS | 2CGG Services (Malaysia)

1

* Corresponding author: [email protected]

FIRST

BREAK

I

VOLUME

36

I

NOVEMBER

2018

1


acquisition restrictions and limitations; secondly, various velocity anomalies have a strong impact on the migrated results and the frequency bandwidth. At very shallow positions just beneath the sea bottom, very bright events can be observed which are likely to be mixtures of gas-filled channels and carbonate cemented sediments (hard grounds and shallow corals). Beneath these amplitude anomalies, outlined by the red polygon in Figure 1b, the seismic data quality deteriorates dramatically. Absorption, azimuthal anisotropy, dispersion, mode conversion of energy and time sags owing to slow gas velocity all contribute to an overall ‘gas chimney’ effect. Notable uncertainties arising from the poor seismic images are the concern of the reservoir management, especially in regards to reservoir continuity, unconsolidated reservoir sands, fault seal integrity, injection water management, sand production and unpenetrated fault blocks, which have been the biggest concerns (Vijapurapu et al., 2013; Jiang et al., 2014). The main hydrocarbon accumulations are located within the northern, up-thrown fault block. The Upper Cycle V/Lower Cycle VI (Upper Miocene) succession in Field B displays a vertical stacking of ca. 160 separate reservoirs with a hydrocarbon-bearing interval that is more than 2000 m thick. The reservoirs are deposited in coastal/deltaic environments where the coastal and inshore areas are influenced by both tidal and wave processes, which include distributary channels, mouth bars, tidal channels/ estuaries, tidal flats and coastal barrier/shoreface sand bodies. Shallow, major hydrocarbon-bearing reservoirs consist of unconsolidated sands and silt with minor thin shale beds. Figures 2a and 2b show the rock physics template for the field that describes the relationship of the velocity with porosity for both the Vp and Vs controlled by pore filling clay minerals, load-bearing clay minerals and diagenetic cementation. Figure 2c highlights very low Vp and Vs anomalies which contrast against the background velocity. As a result of the thick vertically stacked intervals, reservoir properties are both a function of depositional facies and depth. Elastic stiffness below average is mostly associated with load-bearing clay minerals. Reservoir quality is best developed in a laterally continuous shoreline – upper shore-face deposits, with

porosities ranging between 20 and 35% and permeability values between 1 and 3 Darcy. Approaches for acquiring marine ‘True-3D’ seismic data The vintage 3D seismic data was acquired in 1991 with a conventional 3 km short offset survey. Half of the field area was masked by shallow gas overburden or so called ‘gas-clouds’. Prior to the current study, in 2015, a Q-PreSDM was unsuccessfully applied to attempt to resolve large subsurface uncertainty. Subsequently, it was decided to reacquire the marine seismic data with the most modern techniques to significantly improve the subsurface images (see Chandola et al., 2014). The key objectives are to achieve better reservoir resolution (via broadband) and delineation (via wide-azimuth) so that the oil production and recovery factor can be optimized, and additional prospective resource volumes from deeper reservoir targets can be matured. A complex near surface velocity model was built from all of the available data for simulation of the elastic wavefield propagation to ensure amplitude illumination, and the acquisition designs were robust with respect to acquiring both PP and PS wavefields as demonstrated in Figure 3. Several comprehensive designs have been evaluated that generate the best subsurface images at considerable acquisition cost. A new high-density wide azimuth 3D 4C Ocean Bottom Node (OBN) survey was acquired in 2016. The seismic acquisition was comprised of high receiver density OBN rolling patches and a lower density static patch covering five surface platform facilities near the center of the field. There are over 10,075 m active receiver lines with a node spacing of 25 m and the distance between the receiver lines was 150 m. Source lines were shot orthogonal to the receiver lines as shown in Figure 4a. The safest method used to deploy and recover the OBN nodes placed nearby within 25 m to 50 m of the platform facilities was by using nodes on a cable system as shown in Figure 4b. This allowed the vessel to put extra nodes at positions nearer to the platforms to maximize the near offset coverage without risk of node recovery failures. This wide azimuth OBN survey has led

Figure 2 (a) and (b) Illustrate rock physics templates for the field describing the relationship between the velocities and porosities for the Vp and Vs that relate to the diagenetic cementation, sorting and load-bearing clay minerals associated with elastic stiffness. Both the Vp and Vs velocities describe the same diagenesis effects, which mostly follow soft-sand model trends. (c) Low Vp and Vs velocity anomalies are from the shallow overburden against the background as denoted by the red arrows. The sonic log shown in blue and red curves denotes the vintage processing velocity.

2

FIRST

BREAK

I

VOLUME

36

I

NOVEMBER

2018


Figure 3 OBN survey amplitude illumination study conducted prior to the acquisition. Attributes shown are simulated migration amplitude (SMA) from the shallowest reservoir map. (a) Vintage streamer acquisition SMA response. (b) and (c) SMA amplitude illumination of the PP and PS, respectively, from the resulting OBN hybrid rolling and static patch configuration. (d) Multi-depth layered gas cloud model (grey polygon) overlain on top of the shallowest reservoir (plan view).

Figure 4 The field is congested with several fixed installations in the centre of the survey area. (a) Marine OBN 4C multi-component data acquisition for Field B that is comprised of both rolling and static patches. The rolling patch receiver area is 97 km2 and the static patch area is 12 km2. (b) Details of the extra nodes and short cable which were deployed 25 m on either side of the main platform. (c) Rose diagram showing azimuthal information and traces of the fold distribution in the common offset – common azimuth direction. The black circle represents a 6 km common offset and the red circle illustrates a further 12 km offset. (d) Three highly deviated wells were used in the 3D VSP DAS fibre-optic acquisition. The survey shared the same airgun sources as the marine OBN survey. Illuminating rays were illustrated from the surface shot points to the fibre-optic DAS receivers in one well. In the background was the initial PP velocity model prior to the traveltime tomography inversion of the VSP direct arrivals.

to many advantages for illuminating the target reservoirs below the gas cloud region and produced products with higher signal to noise ratios. Even distribution of traces in most azimuths can be seen with a high fold of 6 to 9 km offsets, as illustrated in the common-offset-common azimuth rose diagram in Figure 4c. The vertical and azimuthal anisotropy effect was more clearly observed since this marine acquisition survey was rich azimuth that was densely sampled. There was also a strong probability correlation of shallow hydrocarbon filled unconsolidated producing reservoir sands with anisotropy effects seen on the seismic images. This effect produced large temporal and spatial velocity variations that coincided with the multi depth levels of the shallow gas bodies, regional principal stresses and major fault directions. The only way to solve this velocity issue is by using 3D orthorhombic FWI and more advanced imaging techniques.

Surface recorded multicomponent data have been used many times to attempt to successfully improve images through gas charged overburden, but they cannot provide a perfect image since any downgoing P wave is affected by the overburden before mode conversion at the target. Even with permanent installations, such experiments are expensive to repeat, especially for time lapse measurements. For this experiment, we decided to supplement the surface 3D seismic image by placing additional receivers in the subsurface in a ‘True-3D’ experiment. The most cost-effective method was to utilize Distributed Acoustic Sensing (DAS), in existing fiber optic cables from three highly deviated production wells to acquire the 3D VSP data recorded simultaneously with the OBN static patch, as illustrated by Figure 4d. The borehole 3D VSP data was acquired by sharing the surface OBN nodal survey airgun sources without shutting down the oil production. FIRST

BREAK

I

VOLUME

36

I

NOVEMBER

2018

3


Figure 5 Vertical seismic section extracted from the 3D Q-APreSDM volume overlain with their associated velocity models. (a) PP (b) PS images, respectively. Strong vertical and lateral velocity variation on this field is observed.

Figure 6 Far angle stack of 3D PP Q-APreSDM sections (at 35-degree mute shown by the yellow curve) and its CDP common offset-azimuth gathers denoted by the black dashed lines. (a) VTI APreSDM (b) Orthorombic APreSDM images, respectively. The smaller insets are showing a part of the common offset common azimuth gathers.

4

FIRST

BREAK

I

VOLUME

36

I

NOVEMBER

2018


A total of 55,568 OBN static patch airgun source locations, as part of the dense source grid, illuminated around these three adjacent fiber optic DAS equipped wells with the deepest fiber optic receiver channel placed at a 2854 m measured depth and effective 8 m receiver spacing. Synchronizing the surface seismic acquisition operation and downhole DAS VSP measurements required integrated multi-disciplinary teams and a focus execution plan. We demonstrated that a high-resolution seismic signal can be recorded and recovered beneath the gas charged overburden without the need to halt the in-situ field complex production environment which would have been required for the VSP via conventional geophones. This proves that the technology is potentially viable for future proactive and time lapse reservoir monitoring in the vicinity away from the borehole. It is certainly cost effective since we only need an airgun source boat for future surveys without the need to redeploy the VSP geophones or all surface seismic OBN nodes in the field. The permanently installed fiber optic cable downhole is also used for distributed temperature sensing and as an active sensor to monitor the flow dynamics in these wells, providing a better understanding of the well production and reservoir performance. Improved subsurface images in the gas cloudaffected region Velocity model building is still the most important task leading to any migrated image. The densely sampled wide azimuth survey allowed identification and characterization of both vertical and azimuthal anisotropic effects. There is a strong correlation of shallow hydrocarbon-filled unconsolidated producing reservoir sands with heterogeneity and the anisotropic effects seen on seismic images. Temporal and spatial velocity variations coincide with the multiple depth levels of shallow gas bodies, regional principal stresses and major fault directions. We used 3D orthorhombic FWI and imaging techniques to provide appropriate degrees of freedom to address all of these issues. Regional tectonic compressional stresses on the Field B create a velocity field with azimuthal anisotropy, and the velocity

trends between the PP and PS are significantly different. Figure 5 demonstrates PP and PS seismic sections overlain with their associated velocity models that show strong vertical and lateral velocity contrasts, but for brevity, the azimuthal anisotropic components are not shown. Figure 6b shows that the alignments and degree of the common image gather flatness and travel time jitters, which are reduced in all common azimuth and common offsets, illustrate that the orthorhombic anisotropy velocity model is more accurate in producing PP and PS images, which are then used for quantitative interpretations. Anisotropic velocity model building and imaging of these field data have been discussed in detail by Maitra et al. (2018) in achieving a high-resolution velocity model to address these strong lateral velocity variations across faults and gas bodies. Figure 7c shows the imaged fibre-optic DAS VSP results merged and integrated into the OBN PP seismic volumes to produce a higher-resolution combined seismic stack neighbouring to the well bore location. In order to image the upgoing VSP responses, it is cost effective to utilize reciprocity to apply a Reverse Time Migration (RTM) algorithm supplemented with the downgoing wavefield of the VSP data using the RTM imaging with Multiples (RTMM). The additional benefit of the 3D VSP RTMM is that it can extend the reflection illumination away from the borehole. The 2017 processed 3D PP image of the new OBN data has been found to be superior to the vintage 1991 data reprocessed in 2015 with the anisotropic Q-PreSDM. The comparisons among the legacy seismic images, PP and PS Q-APreSDM are shown in Figures 8a to 8c with their respective depth slices at 1000 m denoted by the red dash lines. Improvements include better resolution around the reservoir levels, recovered data at the platform undershoot areas, enhanced reflector continuity, reduced sagging and amplitude distortions, all contributing to higher confidence in the structural interpretation and reduced uncertainties. Fault shadows are significantly reduced, resulting in defined, sharp discontinuities with some lateral shifts of up to 100 m. Additional faults are illuminated, which are important in well

Figure 7 (a) Vertical profile of the fibre-optic DAS 3D VSP image using reverse time migration (RTM) and RTM imaging using multiples (RTMM) from one of the wells. Image below is the time slice at 800 m depth and the green dash line illustrates the well trajectory in plain view. (b) The same vertical profile as (a) but extracted from the 3D Q-APreSDM (PP) volume. (c) 3D VSP image from the borehole merged with the 3D Q-APreSDM (PP) using a global matching filter.

FIRST

BREAK

I

VOLUME

36

I

NOVEMBER

2018

5


Figure 8 Improved subsurface images comparing the 1991 vintage 3D data to the 2016 4C OBN marine data acquisition. (a) conventional 2005 Kirchhoff PreSDM, (b) 2017 3D Q-APreSDM PP volume showing enhanced fault plane definition, better imaging beneath gas clouds and improved reflector continuity. (c) 2017 3D Q-APreSDM PS volume. The respective depth slices denoted by the red dashed lines on the vertical section were taken at the same depth of the reservoir level. The red arrows show data improvements against the vintage PreSDM.

Figure 9 Better-quality seismic interpretation using the improved PP volume. (a) Vertical seismic section as shown by A to A’ in the map in (b). (b) Structural map of a reservoir. New fault interpretation has been added as highlighted by the arrow. The purple dots denote the well locations in the field.

placements for the next drilling campaign. Figure 9a shows a vertical profile from A to A’ as indicated in Figure 9b, showing an additional fault which has been interpreted and extended to the deeper targets. This fault was not seen in the legacy PreSDM image. The new fault explains a fluid difference at a deeper target level during a previous drilling campaign. Currently, full reservoir characterization efforts including seismic attribute analysis and seismic stratigraphy interpretation are being carried out and these inputs are expected to further improve our understanding of this field. Discussions Our other field experiences in acquiring and imaging 4C multi-component data show that the PP data is more affected by the amplitude attenuation and very low Q-values of the overburden gas clouds and that the PS data can be used to supplement the PP data in these regions. For this project, the resulting mode converted 3D PS image volume of the field data is still not satisfying. 6

FIRST

BREAK

I

VOLUME

36

I

NOVEMBER

2018

Figures 10a and 10b compare the image quality between the PP and PS Q-APreSDM. The PS images suffer from gas cloud effects more than the PP image, with more amplitude attenuation in the centre of the field. Possibly, this is because the PS data are more affected by the visco-elastic wavefield propagation through the heterogeneity of the overburden which is more complex than we realized in our pre-survey modelling, where gas bodies were represented as rather simplified zones. Granli et al. (1999) discussed the effect of processing assumptions on mode converted energy in gas-filled sediments, where they assumed that the compressional wave was attenuated when propagating through the gas chimney hence resulting in poor PS images. Most of the recorded acoustic field data in the Malaysian basin that is affected by gas clouds is found to show strong internal multiples owing to multiple scattering within a high contrast shallow gas overburden (Ghazali, 2011; Ghazali et al., 2016). In order to understand this phenomenon further, we modified the acoustic gas cloud model from Ghazali et al. (2008) to include elastic properties, and we


Figure 10 Vertical seismic section extracted from the centre of the field passing through the gas cloud affected area. (a) and (b) are Inline images of the PP and PS Q-Anisotropy PreSDM taken from the centre of Field B. The PS Q-APreSDM image is still affected by strong amplitude attenuation and this phenomenon is poorly understood. Typically, in other areas, the PP amplitude attenuation is more strongly affected than the PS. (c) Elastic wavefield propagation simulation using finite difference modelling Vz and Vx component time snapshots. Decomposing the wavefield into P-waves and S-waves aids in understanding the field reflection data. The background model illustrates the gas cloud model (modified from Ghazali et al., 2008 with elastic parameters) with the shallow lens and inclusions mimicking the gas cloud affected field data found in most Malaysian basins.

Figure 11 Comparison between the (a) numerical elastic wavefield simulations and (b) actual Field B 4C OBN field data. Travel time delays are observed in the Vz component of the wavefield as shown by the red arrows. Similar mode converted S-wave linear energy is seen on the Vx actual field data and numerical model as shown by the red ellipse.

decomposed the wavefield into P-waves and S-waves to further analyse the propagation on the wavefield and understand the reflection data as shown in Figure 10c. Figure 10c shows the time snapshots of the full elastic wavefield propagation in this model. It has been confirmed through these elastic wavefield simulations, that the P-wave

energy is trapped within the shallow gas bodies before continuing to propagate to deeper depths. This multiple scattered P-wave converts to S-wave energy within that gas cloud with complex multi-path propagation and trails of S-wave codas. Its wavefront propagation is more linear as seen in Figure 10c (S-wave plot) since the S-wave travels with a slower velocity compared to FIRST

BREAK

I

VOLUME

36

I

NOVEMBER

2018

7


the P-wave. Imaging PS data is non-trivial. Figure 11 compares field data and a simulated shot record of Vz and Vx components exactly in the gas cloud area. Travel time delays are clearly demonstrated by the red arrows in the Vz plots of both datasets owing to shallow gas. Illustrated by the red ellipse in the Vx plots, similar S-wave events are also observed and the reflection from the main reflector is shown by the red arrows. Propagation of an elastic wavefield in a complex gas cloud area is still poorly understood. This fundamental understanding is important in order to correlate the amplitude and reflectivity behaviour with the elastic properties derived from the well logs for the PP-PS joint inversions in the reservoir characterizations.

References Abd Rahim, M.F., Ulugergerli, E.U., Konuk, T. and Ghazali, A.R. [2013]. Characterization of Gas Clouds Using Non-linear Full Waveform Inversion. EAGE Seismic Attenuation Workshop, Extended Abstracts. Akalin M.F., L.C. Foo, S. Kumar, S. Kumar, T.W. Hoong, N. ElKady, M.A.M. Nahar, N.S. Abd Majid, R.K. Pratama, S. Stephen, A.W. M Yusoff, B. Viratno and N.S. Alias [2010]. 3D-streamer and 4C-OBC seismic imaging in gas cloud affected area offshore Malaysia. 6th Mediterranean Offshore and Exhibition, Expanded Abstracts. Chandola, S., Ghazali, A.R., Ghazali, F.A., Teck, L.C., ElKady, N., Sharma, S.K. and Abdul Rahman, M.R. [2014]. An Innovative Approach to 3D Survey Design to Meet Exploration and Development Objectives - An OBC Case History from Offshore Malaysia.

Conclusion We have demonstrated a ‘True-3D’ seismic approach to solve seismic imaging problems beneath a heterogeneous shallow gas overburden for reservoir delineation. Diligent planning of rich azimuth OBN surveys for the acquisition of PP and PS data as well as utilizing fibre optics DAS 3D VSP are found to be successful for this intent. Understanding elastic wavefield propagation for both PP and PS is important prior to undertaking any 4C OBN marine multi-component survey. Multiple scattering in the P-leg of PP wavefield in shallow gas bodies produces more complex scattering of the mode converted S-waves. This has an impact the amplitude attenuation or so called ‘absorption’ in this field. It is also observed from the ‘True-3D’ seismic approach that the PP and PS velocity models for the field have significantly different trends from each other, which requires more advanced estimation techniques, such as orthorhombic FWI and model building. The merged 3D DAS VSP produced a higher resolution and vital reservoir images especially beneath the gas cloud affected area. The new 2017 processed 3D 4C OBN marine data quality is superior to the 1991 3D seismic data (reflector continuity, recovering of undershoot area, fault imaging). This has led to a new interpretation that has significantly changed the overall Field B structural framework (depth surfaces and fault positioning) and led to the volumetric variation of the field. Full reservoir characterization efforts (multi-seismic attribute analysis, seismic stratigraphy interpretation) on the PP and PS data, including joint PP-PS inversion, are expected to improve reservoir lateral distribution. New structural and stratigraphic inputs to static and dynamic models for reservoir management and improved reserve estimation thus, also provide opportunities for the upside potential in this area. We have demonstrated that both the kinematics and the amplitude of gas cloud in Figure 1 can be solved satisfactorily by this ‘True-3D’ seismic imaging approach, utilizing both surface and subsurface recording technologies for the same surface shots.

EAGE Land and Ocean Bottom; Broadband Full Azimuth Seismic Surveys Workshop, Extended Abstracts. ElKady, N., Z. Mohd Dom, Y. Prasetyo, M. Bayly, G. Nyein, P. Vasilyev and N. Mat Don Ya, [2012]. Imaging Solutions for Geophysical Challenges in South East Asia. 35th Petroleum Geoscience Conference & Exhibition, Expanded Abstracts. Ghazali, A.R., Verschuur, D.J. and Gisolf, A. [2008]. Seismic imaging through gas clouds: A data‐driven imaging strategy. SEG Technical Program Expanded Abstracts, 2302-2306. doi: 10.1190/1.3059342. Ghazali, A.R. [2011]. True-Amplitude seismic imaging beneath gas clouds. PhD Thesis, Technical University of Delft, The Netherlands. Ghazali, A. R., Hardy, R.J.J., Konuk, T., Masiman, R.I., Xin, K, Mad Zahir, M.H. and Abd Rahim, M.F. [2016]. Velocity model building and imaging in the presence of shallow gas. First Break, 34 (10), 79-84. Granli, J.R., Arntsen, B., Sollid, A. and Hilde, E. [1999]. Imaging through gas-filled sediments using marine shear wave data. Geophysic,s 64 (3), 668-677. Gudipati, V., Jaynes, S., Lee, S., Reilly, J., Lazaratos, S., Neelamani, R., Martinez, A., Zulfitri, A., Onn, F., Shaw, C., Ghazali, A.R., Konuk, T., A. Khalil, A.A. and Nghi, N.H. [2018]. Improving old seismic using full-wavefield inversion and broadband processing: Imaging complex structures under shallow gas. SEG Technical Program, Expanded Abstracts. Jiang, L., Kittrell, C., Duncan, B., Gaffar, G.R., Tarabbia, P. and Whitney, P.A. [2014]. The Role of Petrophysics in Enhanced Oil Recovery in Brown Field Development. EAGE/FESM Joint Regional Conference Petrophysics Meets Geoscience, Extended Abstracts, Doi: 10.3997/2214-4609.20132131. Maitra, S., Basir, F.F., Ghazali, M.L., Ghazali, A.R., M. Sapiai, S., ElKady, N. and Konuk, T. [2018]. Orthorhombic full-waveform inversion and model building for azimuthal anisotropy in the presence of gas bodies. SEG Technical Program, Expanded Abstracts, 5123-5127. Radzi, N.A.M, Yusoff, Y.B.M., Khalil, A. and Amdan, A. [2015]. Structural Interpretation and Reservoir Characterization using 3D 4C OBC Seismic Dataset in S Field, Malay Basin. Annual Petroleum

Acknowledgements We would like to thank the management of Petronas for allowing us to pursue and publish this work and acknowledge the input of many colleagues who have contributed to the Petronas gas cloud research, including past and current Field B FDP teams.

8

FIRST

BREAK

I

VOLUME

36

I

NOVEMBER

2018

Geoscience Conference & Exhibition, Expanded Abstract. Vijapurapu, S., Ghosh, A., Kho, S.F., Leong, K.H., Reijnders, G.J. and Stone, G. [2013]. Using Integrated Modelling Approach to Evaluate the Use of IWAG EOR Process. Petroleum Geoscience Conference & Exhibition, Expanded Abstracts.