Third EAGE Integrated Reservoir Modelling Conference

In Malaysia, more than 70% reserves are held within carbonate reservoir with more than 13TcF of hydrocarbon gas locked in high CO2 fields. With that PETRONAS are embarking journey to unlock and monetize the high CO2 gas fields via CO2 sequestration project with the aim to separate CO2 and store CO2 underground thru pilot project in Field A. Field A is unique in several aspects whereby it has 70% CO2 and 30% of potential hydrocarbon gas sale which is equivalent to approximately 700 million tons of CO2 to be stored and would be the biggest CO2 sequestration project recorded in industrial. The geological storage site will also be within the same field which is located in deep seated aquifer below the producing reservoir. An appraisal well was successfully drilled in Field A with specific well objectives designed to address critical factors identified in CO2 sequestration project ranging from surface to subsurface. The well recorded several key highlights en route, among others, Malaysia’s first cap rock mini fracture test and first water injectivity test in aquifer interval, PETRONAS longest continuous carbonate coring and first cap rock shale core with 98% and 100% core recovery respectively. The cores data is crucial to de-risking initial GIIP and to update current geological model for further evaluation on A5 field as CO2 storage potential sequestration sites

Subsurface imaging can best be realized by Prestack time/depth migration PSTM/PSDM which is the most widely used tool. PSTM processing would be highly effective should we have an appropriate migration velocity field that is very crucial since a velocity model that is significantly different from the medium velocity, may result a misleading migrated section. Further, it may be noted that residual velocity/move-out analysis is the basic step in velocity field refinement. Further, migration velocity scan is important to arrive at an accurate velocity that is being used as the input to RMO correction. RMO correction is very obvious in wide azimuth (WAZ) than in narrow azimuth (NAZ). Generally, the effect of azimuth is more obvious in far offset than in near offset. Residual move out is applied on pre-stack migrated gathers as it is necessary in anisotropy model building. Due to anisotropy, there may be a deviation in velocity even if migration with correct velocity on almost flat gathers which may be offset by RMS velocity. Some of these aspects are studied with the 3D seismic data from Yibal Al Huwaisah field, Sultanate of Oman and the results are presented highlighting salient features.

Evaluating and commercializing low-resistivity, low-contrast (LRLC) reservoirs are increasingly at the forefront of the industry’s concern in diverse geological settings ranging from deep water laminated formations, high conductive formations and heterolithic and silty formations within coastal plain. Current scenario of plunging oil prices and cost optimization practice makes such reservoirs even more interesting as LRLC pay zones are examined via behind casing and no new wells needs to be drilled for production. Traditionally, low-resistivity formation is defined as having an apparent deep resistivity value of less than 5.0 ohm-m. Major challenges in these reservoirs include identification of the hydrocarbon interval, which is usually masked by the lack of resistivity contrast between the hydrocarbon and water zones and evaluation of their commercial-quality. To unlock the hidden potential of LRLC pay sands in the offshore Sarawak Malaysia, the effective integration of subsurface disciplines including petrophysics, geology and quantitative derivatives from seismic analysis is vital. This study covers geological perspective of low contrast reservoirs from an offshore oil field deposited in a lower coastal plain settings located within offshore Sarawak Malaysia. An improved understanding of the geological, petrophysical and geophysical parameters was achieved by adopting a holistic and multidisciplinary approach. This includes the integration of core, logs, rock physics modelled parameters, stratigraphic, depositional and lithofacies information along with stochastic inversion derivatives. The paper quantifies rock physics parameter uncertainties for LRLC pay zones and establishes a framework for LRLC reservoir characterization. Stochastic inversion derived P-Impedance and Vp/Vs ratio are used to predict fluid and facies probabilities for LRLC reservoirs, which then further integrated with stratigraphic information. The results offered an effective way of establishing analogues of producing and non-producing LRLC zones. Analysis of fluid and facies probabilities derivatives driven surface attributes is a way seismic can potentially contribute in indicating areas of relatively better or worse LRLC reservoir continuity. Identified LRLC reservoirs proved to be of commercial-quality and increased oil production to the extent of several hundred thousands of barrels over the years and currently producing.

With ever improving 3D seismic, the trend in static reservoir modelling is towards more direct constraining of 3D-model reservoir architecture by the seismic data. Whilst seismic constraining can result in more realistic, better quality reservoir architecture models, there are a number of key problems to overcome namely: 1) limited vertical resolution of the seismic; 2) limited ability to capture sedimentary body-shapes (e.g., channels and bars) by variograms; and 3) limited ability for geostatistics-based modelling algorithms to capture spatial facies associations in a geologically realistic manner. This paper discusses a number of different approaches to reservoir architecture-model constraining by 3D seismic based on some actual field examples, to distil best practice methods and guidelines. The emphasis in this paper is on clastic reservoir settings with complex “labyrinth-type” reservoir architecture.

The goal of this work is to study the effectiveness (robustness, accuracy and speed) of an objective function in the context of seismic history matching; this aspect is the key component for a successful update of the reservoir model. Two main characteristics of the objective function are in focus: which seismic attributes should be matched to, and how to measure the matching. The sensitive factors currently studied are, the attributes at the simulation domain (pressure, water and gas maps, and combinations of these) and the metrics (Kendall Tau, L2 norm, Minimum ration, Pearson correlation) used to compare the two maps, from the seismic and from the model. The optimisation method used to perform the seismic history matching is an auto-adaptive differential evolution algorithm (SHADE). This study has been carried out on a North Sea field. Based on the results and analysis of the seismic history matching experiments, we are able to draw some practical recommendations, on what kind of objective function should be established to update the simulation model with seismic data.

A reservoir in a field characterised by a turbidite stacked channel and levee-overbank system was presented. Capturing the proper channel geometries and understanding the reservoir connectivity is of utmost importance for optimum field development. It was therefore crucial to build a good reservoir model that is consistent with the all the available subsurface data. The synthetic seismic generated from the reference case static model did not properly match the seismic data. The difference between the original model and the seismic data was minimized by refining the channel distribution and reservoir properties in the model using probabilistic model-based seismic inversion. The post inversion model showed a better fit with the seismic data resulting in significantly different net-to-gross ratios; better porosity distribution; and more refined channel shapes. The updated net-to-gross model also highlighted areas where reservoirs are more poorly developed. The seismically updated intra reservoir shale revealed the possibility of a less laterally extensive shale that was previously thought to separate the two reservoir sequences. The updated static model provided an improved understanding of the geology that would enable the robust dynamic simulation and help better field development planning.

The reconciliation of seismic-derived properties and numerical simulation can be achieved through the application of a workflow collaboration that incorporates discipline-specific processes. This work highlights two methods for relating simulation results back to seismic properties to validate the subsurface description determined from earth modeling and flow simulation, in addition to classical forms of calibration. A primarily simulation-driven petro-elastic modeling (PEM) method based on a streamlined computation of the Gassmann equation in a reservoir simulator, following Ramsay and Yarus (2015 ), is achieved and juxtaposed with an empirically based Batzle and Wang (BW) counterpart ( Batzle and Wang, 1992 ) to assess the sensitivity and applicability of each in a fractured low permeability environment.

Reservoir models require accurate estimation of hydrocarbon saturation for correct initial inplace hydrocarbon volume estimation and for initialization of simulation model. There are two fundamental limitations with capillary pressure measurements i) It is unable to measure poor quality rocks (low RQI) therefore samples are inherently insufficient to represent all type of silicicalstic rocks present in subsurface ii) It is expensive and time consuming therefore not necessarily all the rocks and fields have this kind of measurement available for reservoir modelling. The cost of these limitations leads to errors in STOIIP and uncertainties in reservoir simulation. The common method used to populate hydrocarbon saturation in reservoir model is J function ( Leverett, 1941 ). This model is meant for clean rocks only. Therefore can not predict saturation in shaly sands. To overcome these issues we have designed a saturation height function which captures the saturation height behavior for all kind of siliciclastic rocks. Saturation estimated through this saturation heigt function matched very well with resistivity derived saturation. Therefore we propose to utilize this saturation height function which can overcome the measurement insufficiency and can also can be applied to the reservoir models where core measurements are not available at all.

Field S has undergone through few stages of static modelling in order to improve the quality of the reservoir model as more data became available to the team. The most recent static model was deemed necessary as to add value of 1. New seismic acquisition of 3D 4C OBC 2. New Environment of Depositional (EOD) concepts 3. A higher resolution approach to well log correlation 4. Updated inputs including petrophysics, FMI and well tests This paper aims to demonstrate the improvement in reservoir modelling through the recent acquisition of a 4-Component 3-Dimension Ocean Bottom Cable (4C 3D OBC) seismic survey, new sedimentological study, a higher resolution approach to well log correlation and updated petrophysical inputs including FMI and well test data. In essence all available subsurface data should be integrated to produce the most optimum reservoir model to ensure the best technical result possible. This effort requires collaborative and proactive team work that has proven to be successful in Filed S where the Delta Group reservoir has now been managed to have a reliable history match compared to previous reservoir models