Fault-Guided Seismic Stratigraphy Interpretation via Semi-Supervised Learning

Delineating seismic stratigraphic features and depositional facies is of importance to successful reservoir mapping and identification in the subsurface. Robust seismic stratigraphy interpretation is confronted with two major challenges. The first one is to maximally automate the process particularly with the increasing size of seismic data and complexity of target stratigraphies, while the second challenge is to efficiently incorporate available structures into stratigraphy model building. Machine learning, particularly convolutional neural network (CNN), has been introduced into assisting seismic stratigraphy interpretation through supervised learning. However, the small amount of available expert labels greatly restricts the performance of such supervised CNN. Moreover, most of the exiting CNN implementations are based on only amplitude, which fails to use necessary structural information such as faults for constraining the machine learning. To resolve both challenges, this paper presents a semi-supervised learning workflow for fault-guided seismic stratigraphy interpretation, which consists of two components. The first component is seismic feature engineering (SFE), which aims at learning the provided seismic and fault data through a unsupervised convolutional autoencoder (CAE), while the second one is stratigraphy model building (SMB), which aims at building an optimal mapping function between the features extracted from the SFE CAE and the target stratigraphic labels provided by an experienced interpreter through a supervised CNN. Both components are connected by embedding the encoder of the SFE CAE into the SMB CNN, which forces the SMB learning based on these features commonly existing in the entire study area instead of those only at the limited training data; correspondingly, the risk of overfitting is greatly eliminated. More innovatively, the fault constraint is introduced by customizing the SMB CNN of two output branches, with one to match the target stratigraphies and the other to reconstruct the input fault, so that the fault continues contributing to the process of SMB learning. The performance of such fault-guided seismic stratigraphy interpretation is validated by an application to a real seismic dataset, and the machine prediction not only matches the manual interpretation accurately but also clearly illustrates the depositional process in the study area.

PETROPHYSICAL ANALYSIS AND SEISMIC STRATIGRAPHY INTERPRETATION TO DETERMINE HYDROCARBON RESERVOIR IN TARAKAN BASIN, BUNYU ISLAND WATERS

Tarakan Basin area of Bunyu Island Waters is known to have hydrocarbon potential with complex geological structures. This study aims to determine reservoir characterization and to obtain prospect of hydrocarbon reservoir zones based on petrophysical and seismic stratigraphy analysis with reference to Well DDS-1 and 2D seismic Line S88. Petrophysical analysis results 3 zones that have potential as hydrocarbon reservoirs. Based on petrophysical quantitative analysis, Zone 1 has values of 52.25% for shale volume, 18.48% for effective porosity, 39.84% for water saturation and 13.03 mD for permeability. Zone 2 has values of 54.66% for shale volume, 10.27% for effective porosity, 40.9% for water saturation and 1.14 mD for permeability. Zone 3 has values of 49.22% for shale volume, 9.33% for effective porosity, 56.33% for water saturation and 0.22 mD for permeability. Out of these three reservoir zones in Well DDS- 1, Zone 1 has the prospect of hydrocarbons which is supported by the net pay value. Based on seismic stratigraphy interpretation, the reservoir zone is correlated to the Tabul Formation, which comprises calcareous clay and limestone.

Seismic reservoir characterization of gas-sand deposits in the Kutei basin, Indonesia

Kutei Basin has the second largest hydrocarbon reserve in Indonesia. In addition to the Miocene inversion related structural traps, slope-fan and channel stratigraphic traps are also important traps in this basin. To guide stratigraphic traps explorations in the basin, the seismic stratigraphy, attributes, and AI inversion methods are integrated to identify and map the reservoir seismic facies, porosity, and pore-fluid. Well data indicates that the studied reservoirs are filled by gas. Seismic data shows that there are two main gas-sand reservoirs corresponding to strong amplitude anomaly. Seismic stratigraphy analysis, guided by seismic attributes, shows that these gas-sand reservoirs were deposited in the channel and local fan facies. The AI inversion is applied to identify and map the porosity and pore-fluid of these two sand reservoirs. Future well locations are identified by integrating the facies, porosity, and pore-fluid maps.

Estratigrafía sísmica 3D aplicada a la estimación de litología en un sistema deltaico

Seismic stratigraphy becomes a useful tool when it comes to 3D lithology distribution, since it gives the interpreter insights of the facies most likely to be present in a certain sedimentary environment. On the other hand, it is also the main input information while modeling petrophysical properties like water saturation, effective porosity and permeability, which are critical in the process of evaluation of a hydrocarbon reservoir. In this context, techniques such as seismic inversion allows the geoscientists to get 3D models of P-impedance, S-impedance and density, which are used as the main input to estimate the reservoir petrophysical properties just mentioned and additionally useful parameters used as a lithology indicator. This paper proposes a workflow to achieve the goal of integrating seismic stratigraphy, seismic inversion and attributes to get a lithology 3D model. Now, to get a suitable correlation between the facies interpreted using well logs and core data with the elastic properties, rock physic templates (RPT’s) were made where proper elastic modulus was carefully chosen to define probability distribution functions (PDF’s) for each facies defined in the correlation wells. On the other hand, based on a set of stratigraphic surfaces created on a different study, 3D models of P-impedance, S-impedance and density were obtained from seismic inversion so that the RPT’s could be built. For this specific instance, only a set of the elastic properties and seismic attributes offered a suitable correlation with the facies defined in the calibration wells. Moreover, the probability distribution functions (PDF’s) already generated allowed the distribution in 3D and the definition of the ranges in which each facies previously stated varies for the elastic modulus estimated.