The test study area is the Miocene reservoir of Nam Con Son Basin, offshore Vietnam. In the study we used unsupervised learning to automatically cluster hydraulic flow units (HU) based on flow zone indicators (FZI) in a core plug dataset. Then we applied supervised learning to predict HU by combining core and well log data. We tested several machine learning algorithms. In the first phase, we derived hydraulic flow unit clustering of porosity and permeability of core data using unsupervised machine learning methods such as Ward’s, K mean, Self-Organize Map (SOM) and Fuzzy C mean (FCM). Then we applied supervised machine learning methods including Artificial Neural Networks (ANN), Support Vector Machines (SVM), Boosted Tree (BT) and Random Forest (RF). We combined both core and log data to predict HU logs for the full well section of the wells without core data. We used four wells with six logs (GR, DT, NPHI, LLD, LSS and RHOB) and 578 cores from the Miocene reservoir to train, validate and test the data. Our goal was to show that the correct combination of cores and well logs data would provide reservoir engineers with a tool for HU classification and estimation of permeability in a continuous geological profile. Our research showed that machine learning effectively boosts the prediction of permeability, reduces uncertainty in reservoir modeling, and improves project economics.
7 August 2024 in Vietnam Journal of Marine Science and Technology
Machine learning (ML) is an artificial intelligence (AI) that enables computer systems to classify, cluster, identify, and analyze vast and complex sets of data while eliminating the need for explicit instructions and programming. For decades, machine learning has become helpful for complex reservoir characterization such as carbonate reservoirs. Permeability prediction from well logs is a significant challenge, especially when the core data is rarely available due to its high cost. In this study, we aimed to bridge this gap by demonstrating the practical application of integrating Hydraulic Flow Units (HFU) and machine learning methods. Our goal was to provide a reliable estimation of permeability using core and wireline logging data in the complex Middle Miocene carbonate reservoir of the CX gas field in the southern part of the Song Hong basin. In the first step, due to the reservoir’s heterogeneity, the core plug dataset was classified into 5 HFUs based on the flow zone indicators (FZI) concept from the modified Kozeny-Carman equation using unsupervised machine learning - K-means method. The porosity - permeability for each HFU was defined after HFU clustering. In the second step, we designed three different workflows to predict permeability and HFU using supervised machine learning from a combination of core and log data. These workflows were rigorously test and compared with the core data. The most accurate result was chosen as the base, providing a high confidence level in our predictions’ reliability.