A new approach to evaluate hydrocarbon generation characteristics by pyrolysis Tmax in lacustrine shale oil plays

It has been revealed that significant disparities in both the organic matter source and hydrocarbon generation characteristics of lacustrine sedimentary environments, causing challenges in the assessment of continental shale oil prospects  . Lacustrine-sourced shale oil resources in China exhibits notable longitudinal and vertical heterogeneity, which poses a substantial challenge in objectively assessing geological resources and shale oil prospects, especially in a region characterized by overall low thermal evolution . Advanced pyrolysis or bulk kinetic experiments are invaluable tools to refine the understanding of petroleum generation timing . Nevertheless, such experiments are expensive and time-consuming and hence cannot be executed on extensive sets of samples to capture the overall lateral and vertical variability that a source formation may inherit  .In this study, we proposed a new method to rapidly evaluate the hydrocarbon generation characteristics of lacustrine source rocks utilizing anomalies in the Rock-Eval pyrolysis parameter Tmax across various lacustrine shales.

The workflow is depicted as flows: The analysis workflow starts with the selection of samples with TOC exceeding 1 wt.%, given the economic exploration potential of these shales. Subsequently, these samples are categorized into low and high maturity profiles based on the measured vitrinite reflectance (Ro). The two maturity profiles are further classified into low and high Tmax classes using machine learning data analysis. The kmeans clustering method in the Python library scikit-learn was utilized to classify different Tmax values to specific classes  . In certain instances, a third cluster or class may be necessary, depending on the data structure. Samples in the “low Tmax” class typically exhibit high Production Index (PI = S₁/(S₁+S₂)) while the Hydrogen Index (HI: = S₂/TOC*100) values decrease with increasing maturity. In contrast, the “high Tmax” class maintains consistently high HI and low PI at different maturity levels. This analysis workflow facilitates the identification of distinct hydrocarbon generation characteristics for source rocks at different maturity levels based on the Tmax values.

Overall, the “low Tmax” class shows characteristics of early hydrocarbon generation, low activation energy, and wide hydrocarbon generation windows, while the “high Tmax” class shows characteristics of late hydrocarbon generation, high activation energy, and narrow hydrocarbon generation windows. Notably, these diverse hydrocarbon generation characteristics are mainly related to the composition of the primary organic matter, a correlation that can be confirmed through organic petrographical observations.

This analysis workflow is validated with three examples. There are a great data pool of Tmax，and it is recommended to shift the focus towards source rocks that host organic matter favorable for early oil generation. This involves identifying rocks with low Tmax values and hence low activation energy, as they are indicative of conditions conducive to the initiation of oil generation. When it comes to in-situ heating, The exact prediction of hydrocarbon generation processes enables a more precise calculation of current geological recoverable resources. This study has important guiding significance for oil and gas exploration.

Fig. 1. Workflow for determining hydrocarbon generation characteristics of source rocks by a classification according to T_max variability.