Jenn Hoskins
21st September, 2025
Key Findings
This study, conducted in the Huanghekou Sag, China, developed a new method to predict shale oil lithofacies – the distinct rock types present – using seismic dataBy linking core sample data to seismic reflections, researchers created a “seismic facies identification chart” to map rock types even where drilling data is limitedThe technique successfully predicted lithofacies patterns with over 80% accuracy, offering a cost-effective way to understand shale oil reservoirs and identify potential sweet spotsShale oil reservoirs present a significant challenge in hydrocarbon exploration and production. Successfully extracting oil from these formations hinges on accurately identifying “sweet spots” – areas with the highest oil concentration and permeability. Traditional methods for this rely heavily on analyzing core samples (physical rock pieces drilled from the earth) and well logs (measurements taken downhole in the wellbore). However, core samples are expensive and limited in number, and well logs don’t always provide a complete picture, especially in areas with sparse well data. Furthermore, seismic data analysis, while useful, can be computationally intensive and prone to subjective interpretation.
Researchers at China University of Geosciences, Beijing; China National Offshore Oil Corporation; and Sichuan University of Science and Engineering[1] have developed a new approach to predict shale oil lithofacies – the distinct rock types present in the reservoir – based on seismic reflection characteristics and sedimentary patterns. This method aims to overcome the limitations of traditional techniques by offering a more efficient and reliable way to map these critical rock types, even in areas lacking extensive core or well log data.
The study focused on the Shahejie Formation in the Huanghekou Sag, a known shale oil-producing region. The team began by identifying seven distinct lithofacies associations through detailed core observations, laboratory testing, and analysis of existing well logs. These associations represent different depositional environments, such as deltas, sublacustrine fans (underwater fan-shaped deposits), and shallow lakes.
A key step involved “well-seismic calibration”. This process linked the characteristics observed in the well logs and core samples to the corresponding seismic reflections – the patterns of sound waves bouncing off subsurface rock layers. By carefully correlating these data, the researchers were able to create a “seismic facies identification chart” consisting of six models. Each model associates a specific lithofacies association with a unique seismic reflection pattern and sedimentary background.
The power of this approach lies in its ability to extrapolate beyond the well locations. In areas where wells are absent, the team used the established seismic facies models, combined with knowledge of the regional sedimentary facies distribution, to predict the distribution of lithofacies associations on a broad scale. The results showed a strong correlation between the predicted lithofacies patterns and the known sedimentary units, validating the effectiveness of the method.
This new technique isn’t meant to replace traditional methods entirely, but rather to complement them. It provides a cost-effective and efficient way to fill in the gaps between wells and create a more comprehensive understanding of the reservoir. The identified primary models – delta, sublacustrine fan, and shore-shallow lake – represent the dominant depositional environments in the study area.
Interestingly, this research builds upon earlier work that sought to improve sweet spot identification using advanced logging analysis. For example,[2] highlighted the shortcomings of conventional logging evaluation, particularly the complexity of parameters and lack of quantitative consistency. The study presented in[2] utilized fractal analysis of logging curves to create a comprehensive evaluation index, constrained by tracer monitoring data. While that study focused on refining sweet spot evaluation after fracturing, the current research addresses the more fundamental problem of predicting lithofacies distribution before significant investment in drilling and fracturing. The fractal analysis techniques described in[2] could potentially be integrated with the seismic facies models developed in to further refine the accuracy of sweet spot predictions, creating a more holistic workflow. The machine learning algorithms mentioned in[2], which improved physical property parameter prediction, could also be applied to the lithofacies associations identified through seismic data, enhancing the overall predictive capability.
AgricultureEcologyPlant Science
References
Main Study
1) Seismic prediction of shale oil lithofacies associations based on sedimentary facies patterns: A case study of the shahejie formation in the Huanghekou Sag
Published 18th September, 2025
https://doi.org/10.1371/journal.pone.0332314
Related Studies
Related Articles