University of Twente Student Theses

Login

LWIR spectroscopy on feldspars from rock plugs for the detection of permiable zones in geothermal systems.

Nyahucho, Doreen (2020) LWIR spectroscopy on feldspars from rock plugs for the detection of permiable zones in geothermal systems.

[img] PDF
3MB
Abstract:Hydrothermal minerals indicate inferred permeability in geothermal systems besides being indicative of formation temperature, fluid composition and host rock composition. Minerals, particularly hydrothermal feldspars (adularia and albite), are important minerals in permeability studies. Adularia has a direct link to highly permeable zones, whereas albite is indicative of inferred low permeability. The ability to recognize and distinguish adularia from other feldspar minerals is, therefore, crucial. Spectroscopy has been used as an alternative approach relative to convectional methods to study feldspars minerals. However, there are limited spectroscopy studies on adularia, this study explores LWIR spectroscopy on rock plugs from the Karangahake epithermal Ag-Au system to map adularia and other associated minerals. These samples are used in the development studies of geothermal systems because the epithermal systems are fossil equivalents of geothermal systems, particularly low sulphidation systems, which are mainly found in highly permeable zones. The objective is to (1) understand spectral characteristics of feldspars; adularia and other hydrothermal minerals associated with it, (2) assess the performance of LWIR in the mapping of feldspar minerals and, (3) establish mineral assemblage that can be used to identify adularia-rich areas for permeability studies in the future. A total of 50 hyperspectral images of 50 rock plugs of approximately 1inch diameter were acquired in the LWIR wavelength range by a hyperspectral imaging system. After processing of the acquired images for noise reduction and reflectance to emissivity conversion, wavelength mapping and Principal Component Analysis (PCA) algorithms were applied to facilitate endmembers collection. Once the endmembers were collected, they were compared to reference spectral libraries for mineral identification and naming. Then, the Spectral Angle Mapper (SAM) was used to classify and quantify the images based on the most dominant mineral per pixel. Also, Iterative Spectral Mixture Analysis (ISMA) algorithm was used to obtain fractional abundances of constituent mineral per pixel for easy comparison with TIMA data. Given unrealistic ISMA results, i.e., negative fractional abundance values, it was difficult to quantify the sample images for comparative analysis. Therefore, mineral quantification was done based on SAM classification results and ISMA provided an alternative approach to understand the spectral behavior of mixed pixels. Validation of the LWIR identified minerals was done using XRD measurements and TIMA data on the selected 35 samples. Results present various feldspar minerals characterized by broad spectral features within 8100-10000nm spectral range. Other identified minerals are silica, carbonate and clay minerals. Spectral differences in wavelength position and shape (defined by width, depth and symmetry) of the diagnostic features reflect variability in their mineralogical composition and can be used to distinguish the identified minerals. LWIR mineral estimates indicate that rock samples have been altered to a certain extent due to the presence of alteration minerals such as adularia, albite, pyrite and calcite. The Identified minerals were common in all datasets; LWIR, TIMA and XRD data. This is despite the shortcomings in mineral mapping, such as misrepresentation of minerals in LWIR maps. Generally, LWIR data were correlated to TIMA and XRD data by up to ~0.5 correlation coefficient. The relationship between albite and adularia maintains its inverse relationship, suggesting the possibility of using this relationship in permeability studies. Hydrothermal minerals that commonly occur with adularia are quartz, calcite, albite and pyrite; however, this is insufficient to establish mineral assemblage for the identification of adularia-rich areas.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:https://purl.utwente.nl/essays/83634
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page