The fractured granite basement in the Nam Con Son basin of southern offshore Vietnam has become a new exploration target in recent years. However, there are two major challenges in the early days of exploration. First of all, it is not always easy to locate a drilling target, for example, a faulted zone or a better fractured area on seismic due to the often degraded data quality within basement. Secondly, it can be difficult to drill the target with better well trajectory due to often poor understanding of the potential productive fracture systems. Those two challenges are also very common for most of the fractured basement reservoirs. In this study, available regional geology data including structural geology information and outcrops have been collected to help geoscientists to gain a basic understanding of the local geology related to the basement and fracture development. A full suite of well data from an existing wildcat well in the same block including conventional open hole logs, advanced well logs (i.e. borehole images and SonicScanner), mud logs as well as PLT (production logs) and DST data was carefully studied to extract all possible information on natural fracture and fault. This resulted in two main outputs - the fracture density log along the wellbore and the defined potential fracture/fault system with dip and strike. Then the fracture output from the existing well was used to fully calibrate the seismic attribute analysis for the whole block, which mapped out the fracture density in 3-D volume. A new exploration well with high deviation was designed on the basis of fracture density volume data as well as the understood potential productive fracture/fault attitude (dip & orientation) and drilled smoothly. A commercial hydrocarbon discovery was made in this new basin first time in the fractured basement. Introduction Hydrocarbon has been discovered in the Miocene and Oligocene clastics as well as carbonates in the Nam Con Son basin, southern offshore Vietnam since 1990's. However, drilling activity for the Pre-Tertiary granite basement has been still very limited without any commercial discovery until late 2010. This paper intends to summarize the fracture characterization for the studied field where the 1st commercial oil discovery in the basement was made in the Nam Con Son basin as well as the learned experience on how to define a drilling target based on the limited existing well data and seismic data in the exploration stage for this unconventional fractured tight reservoir.

No abstract

Sedimentary geometry on borehole images usually summarizes the arrangement of bed boundaries, erosive surfaces, cross bedding, sedimentary dip, and/or deformed beds. The interpretation, very often manual, requires a good level of expertise, is time consuming, can suffer from user bias, and become very challenging when dealing with highly deviated wells. Bedform geometry interpretation from crossbed data is rarely completed from a borehole image. The purpose of this study is to develop an automated method to interpret sedimentary structures, including the bedform geometry, from borehole images. Automation is achieved in this unique interpretation methodology using deep learning. The first task comprised the creation of a training dataset of 2D borehole images. This library of images was then used to train machine learning (ML) models. Testing different architectures of convolutional neural networks (CNN) showed the ResNet architecture to give the best performance for the classification of the different sedimentary structures. The validation accuracy was very high, in the range of 93–96%. To test the developed method, additional logs of synthetic data were created as sequences of different sedimentary structures (i.e., classes) associated with different well deviations, with addition of gaps. The model was able to predict the proper class and highlight the transitions accurately.

Central Luconia Province is one of the world's largest carbonate complexes, with over 200 known Miocene build-ups, created by coral reef structures growing on horst blocks. Most build-ups are gas reservoirs, but also include some oil fields. Traps are stratigraphic, as porous carbonates are encased in sealing shales. The main method used to model properties in Miocene carbonate patch reefs is a direct porosity modeling using Acoustic Impedance as a secondary variable. The lack of long cored sections in this block is the main reason for the absence of facies integration in the static model. Another challenge is the limited number of wells per patch reef, which makes the understanding of lateral facies variation very limited. The main challenge to model these reservoirs is to characterize the distribution of the multi-scale heterogeneities, where hydraulic properties are not necessarily directly related to porosity. Accurate rock typing and porosity determination are key components for establishing porosity – permeability relationships based on rock type. The study area is located in Central Luconia Province, offshore Sarawak. The discovery well proved several hundred meters of gas column down to the gas-water contact within a Cycle IV carbonate patch reef. A complete set of logs was acquired, although no cores were available for facies identification and calibration. This case study presents a methodology to populate facies and porosity in 3D through a holistic approach to data integration and using conceptual models from field analogues. Introduction There is not a single industry-wide agreed-upon scheme that defines how to group lithofacies and environmental settings for mapping purposes or for subsequent reef/platform modeling. Within the Central Luconia Province this may be due to complex interactions between diagenesis and depositional facies, particularly within the shallow-water part of carbonate build-ups. Ali and Abolins (1999) proposed six basic groupings of facies based on diagenetic features: chalkified limestone, moldic limestone, tight argillaceous limestone, sucrosic dolomite, moldic sucrosic dolomite, and "overdolomite". These terms are difficult to relate back to recognized lithofacies terminology and certainly do not relate well to depositional settings. Also, no significantly dolomitized carbonates have been observed in any of the cores from the fields in the study area, and the term "overdolomite" was not defined. Therefore this classification doesn't appear to be directly applicable for this case study. Based on field studies over a 5 year period Epting (1980) proposed that diagenetic processes affected the various depositional units selectively. He concluded that Central Luconia carbonate build-ups can be subdivided into four lithological groupings with discrete ranges of petrophysical properties. He referred to these groupings as "rock types". Based on core examination, correlation with logs suites, and an understanding of the biological and depositional zonation found on modern patch reefs, Epting (1980) suggested four environments that could provide useful lithofacies groupings within Central Luconia study area. His classification of carbonate settings includes protected, reefoid, shallow open marine off-reef and deeper open marine off reef. Other summaries have been offered for mature shelf-margin reefs (Longman, 1981), ribbon reefs from the Great Barrier Reef (Vernon and Hudson, 1978) and many others. Definitions for a well-defined, not overly-complex scheme that is well matched to patch reef settings in the Indonesia-Malaysia area by Jordan (1998) include the following settings: a. Middle Shelf, b. Inter-reef Lagoon, c. Reef-Sediment Slope (Base of Slope), d. Reef, e. Reef Flats, and f. Island. This study is based on this latest classification.

One of the main challenges in the fractured basement exploration and development is the identification of proper well targets with minimized drilling risk. The estimation of fracture porosity and permeability distribution in reservoirs with low matrix porosity is also very challenging as the production relies mostly on the complex fracture network. The proposed method aims at identifying the potential flow-contributing fractures, modeling accurately the fracture density in the interwell space, and finally, building a robust and well-calibrated discrete fracture network (DFN) that is used to compute porosity and permeability properties. This holistic approach integrates the latest technologies for interpreting borehole images, identifying open fractures, computing fracture aperture, and generating a DFN. It also includes at each step qualitative and quantitative analyses to optimize the choice of 3D parameters: seismic attributes and estimations of fracture length, orientation and aperture distribution.