University of Twente Student Theses
Detection, Mapping and Classification of Crude Oil in Drill Core using Hyperspectral Imagery
D'Silva, Shanan (2024) Detection, Mapping and Classification of Crude Oil in Drill Core using Hyperspectral Imagery.
|
PDF
8MB |
| Abstract: | Oil, a complex mixture of hydrocarbons, is a crucial energy resource and industrial commodity with a wide range of applications, from powering transportation to serving as the foundation for various chemical products. Understanding its spatial distribution within drill cores is essential for optimizing drilling strategies and targeted exploration efforts, such as horizontal drilling and hydraulic fracturing. Beyond exploration, mapping the oil concentration in drill cores provides valuable insights into the porosity and permeability of different lithologies, which is critical for enhancing fluid flow in geothermal reservoirs. Additionally, identifying impermeable caps in hydrocarbon reservoirs is vital for effective carbon storage. Previous studies on oil in drill cores have primarily focused on the Alberta region of Canada, known for bituminous oils. These studies use uncemented drill cores containing heavy, viscous oil and minerals such as smectite, quartz, and clays, which have non-overlapping absorption features with oil. This study, however, addresses the detection, mapping, and classification of oil and minerals in drill cores from Dutch hydrocarbon reservoirs, which contain lighter oils and minerals such as clays, carbonates, and evaporites, whose absorption features overlap with those of oil. Two datasets were utilized in this study. The first dataset comprised pure oil samples, which served as a training set to determine the spectral properties of oil in the VNIR, SWIR, and TIR ranges. The second data set consisted of two drill cores, differentiated based on their mineral composition and the type of oil they contain. The first drill core contains transparent oil with minerals like kaolinite, gypsum, calcite, quartz, and hematite, while the second drill core contains heavy oil along with minerals like kaolinite and quartz. The understanding obtained from measuring and comparing the spectra of the oil samples was then applied to detect, map, and classify oil and minerals within the drill core. Oil detection was achieved independently through the UV fluorescence property of oils, confirming the presence of oil in both the drill cores. Furthermore, the mapping of oil and minerals in a drill core was achieved using techniques like Hydrocarbon Index, Spectral Angle Mapper, and Minimum Wavelength Mapper. The existing Hydrocarbon Index was modified to include another significant oil feature, resulting in improved accuracy and reduced noise. The Spectral Angle Mapper and Minimum Wavelength Mapper were applied over specific wavelength ranges, differing from previous studies that applied SAM over the full range. The ranges of 1.65 µm to 1.85 µm and 2.3 µm to 2.36 µm were identified as optimal for mapping oil in the drill core. These specific ranges highlighted by this study helped to better map the spatial distribution of the oil and minerals. Additionally, the mapping techniques used also reflected the relative concentration of oil within the drill cores. This study also aimed to classify the oil in the drill cores based on the functional groups present in the oil. Hydrogen and carbon are the most abundant elements in oil, making their absorption features more prominent compared to those of nitrogen, oxygen, and sulphur. The functional groups formed by these elements are best differentiated in the TIR range, while the SWIR and VNIR range only show major absorption features due to C-H bonds and minor absorption features due to the O-H and N-H bonds. When present in a drill core, these minor absorption features of oil are masked by the other minerals or water present. Thus, the study reveals that classifying the oil in the drill core based on the functional groups is not possible from the VNIR and SWIR spectra. Lastly, this study provides a prototype workflow to detect and map the oil and minerals in a drill core. |
| Item Type: | Essay (Master) |
| Faculty: | ITC: Faculty of Geo-information Science and Earth Observation |
| Subject: | 38 earth sciences, 43 environmental science |
| Programme: | Geoinformation Science and Earth Observation MSc (75014) |
| Link to this item: | https://purl.utwente.nl/essays/101529 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page