Rapid geochemical screening of light hydrocarbons in pre-salt crude oils using a tile-based Fisher ratio approach and GC×GC-TOFMS

Understanding compositional changes in crude oils during Water Alternating Gas (WAG) injection is essential for time-lapse petroleum geochemistry, as production-induced variations can provide valuable insights into reservoir connectivity, fluid displacement, and mixing processes. Light hydrocarbons (C₅–C₁₅) are among the most sensitive molecular tracers of these changes, but extracting meaningful geochemical information from high-resolution chromatographic datasets remains challenging. Here, a tile-based Fisher ratio workflow coupled with GC×GC-TOFMS was employed to investigate geochemical changes in the light hydrocarbon range (C₅–C₁₅) of Brazilian pre-salt crude oils subjected to WAG injection. The dataset comprised 38 crude oil samples, including 25 collected during the pre-production stage (t₀) and 13 during the production stage (t₁). Sample preparation and GC×GC-TOFMS analyses were performed according to the methodology described by Coutinho et al. (2025). Data processing was carried out using ChromaTOF Tile (LECO, St. Joseph, MI, USA), which divides GC×GC chromatograms into tiles for statistical comparison using the Fisher ratio. Partial Least Squares-Discriminant Analysis (PLS-DA) combined with cross-validation generated a robust and highly predictive classification model (R² and Q² > 0.99), enabling the identification of the molecular geochemical markers that best discriminate the t₀ and t₁ crude oils. Variable Importance in Projection (VIP) scores and heatmaps identified branched hydrocarbons, including trimethylcyclohexane and dimethylcyclohexane isomers, as well as branched alkanes, as the main molecular geochemical markers distinguishing the production stages. In addition, crude oils collected during the t₁ phase exhibited a relative enrichment of light hydrocarbons compared with t₀ samples. These results demonstrate that the tile-based Fisher ratio approach provides a rapid and reliable workflow for petroleum geochemical investigations, enabling the efficient identification of 16 molecular markers associated with gas injection. Beyond marker discovery, the proposed workflow establishes a robust framework for untargeted interpretation of production-induced compositional changes in crude oils.