Abstract

Summary This study explores the presence of natural hydrogen (H2) in the Paris Basin, employing Optical Character Recognition (OCR) technology to sift through an extensive database of older drilling records that have not been fully utilized in the past. Our research focuses on the BEPH French database, which contains detailed information on drilling activities, but in a non-searchable PDF image format. OCR is a tool that turns images containing text, like scanned documents, into text files that we can easily search and analyze. Our findings indicate the presence of H2 in several wells across the French sedimentary basins. Particularly intriguing are the results from the Aquitaine Basin, where hydrogen occurrences align with the geological context, and the Paris Basin, which exhibits unexpected H2 occurrences not directly linked to anticipated geological factors known for H2 exploration. In the Paris Basin, the highest hydrogen concentration (52 vol%) was discovered in the Dogger formation. These wells are situated along the Bray fault and thrust, suggesting a geological influence on the distribution of H2. This research underscores the utility of OCR technology in re-evaluating historical drilling data for natural hydrogen exploration. It highlights the necessity for thorough exploration strategies in this nascent yet promising field.