International Journal of Applied Earth Observation and Geoinformation 59 (2017) 1–8 Contents lists available at ScienceDirect International Journal of Applied Earth Observation and Geoinformation jo ur nal home page: www.elsevier.com/locate/jag Predicting the abundance of clays and quartz in oil sands using hyperspectral measurements Iman Entezari a,* , Benoit Rivard a , Mirjavad Geramian b , Michael G. Lipsett c a Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, AB, Canada b Department of Chemical and Materials Engineering, University of Alberta, Edmonton, AB, Canada c Department of Mechanical Engineering, University of Alberta, Edmonton, AB, Canada a r t i c l e i n f o Article history: Received 17 January 2017 Accepted 21 February 2017 Keywords: Oil sands Clay minerals Hyperspectral data Infrared Spectroscopy a b s t r a c t Clay minerals play a crucial role in the processability of oil sands ores and in the management of tailings. An increase in fine content generally leads to a decrease in both bitumen recovery performance and tailings settling rate. It is thus important to identify clay types and their abundance in oil sands ores and tailings. This study made use of oil sands samples characterized for quantitative mineralogy by x-ray diffraction, to gain an understanding of changes in the reflectance spectra of oil sands. The sample suite included bitumen-removed oil sands ore samples and their different fine size fractions. Spectral metrics applicable to the prediction of quartz and clay contents in oil sands were then derived with a focus on metrics correlating with sample content in total 2:1 clays (total of illite and illite-smectite) and kaolinite. Metrics in the shortwave infrared (SWIR) and longwave infrared (LWIR) were found to correlate with mineral contents. The best predictions of clays and quartz were achieved using LWIR metrics (R 2 > 0.89). Results also demonstrated the applicability of LWIR metrics in the prediction of kaolinite and total 2:1 clays. © 2017 Elsevier B.V. All rights reserved. 1. Introduction Oil sands located in northern Alberta, Canada, contain a vast reserve of heavy oil. Alberta oil sands are natural mixtures comprised of 55–80 wt% minerals (primarily quartz and clays), 2–15 wt% water, and 4–18 wt% bitumen; a highly thick and viscous oil (Kasperski, 2001). Bitumen production through open-pit min- ing uses a water based process that results in production of large volumes of tailings; mixture of solids, water, and residual bitu- men. Among the factors affecting the bitumen recovery efficiency, ore mineralogy plays a critical role. Silicates and phyllosilicates, sulfides and sulphates, oxides and hydroxides, phosphates, and carbonates, are the mineral classes found in Alberta’s oil sands (Bichard, 1987; Hepler and Hsi, 1989). Clay minerals are partic- ularly important, as a relationship exists between increased clay content in oil sands ore and decreased bitumen recovery (Liu et al., 2004). Kaolinite and illite have been observed to comprise a major fraction of clays in oil sands (Kessick, 1979; Kotlyar et al., 1995; Mercier et al., 2008). However, evidence of the presence of minor amounts of chlorite, vermiculite, smectite, and mixed layer clays * Corresponding author. E-mail addresses: entezari@ualberta.ca, iman.entezari@gmail.com (I. Entezari). including illite-smectite and kaolinite-smectite has been reported in several studies (Yong and Sethi, 1978; Scott et al., 1985; Omotoso and Mikula, 2004; Omotoso et al., 2006; Kaminsky, 2008). It has been reported that active clays, mainly interstratified and ultra- fine illite and kaolinite, are responsible for poor bitumen recovery and slow settling of tailings (Kasongo et al., 2000; Wallace et al., 2004; Omotoso and Mikula, 2004). Consequently, characterization of the abundance and type of clay minerals is of importance to the oil sands industry due to their influence on ore processability and tailings operations. Several techniques are employed for the characterization of minerals in oil sands. As summarized by Kaminsky (2008), energy-dispersive x-ray spectroscopy (EDX) analysis and x-ray fluorescence spectroscopy (XRF) are used to investigate elemen- tal composition, electron diffraction (ED) and x-ray diffraction (XRD) are employed to determine the mineralogy, quantitative XRD (QXRD) methods are used to determine the quantitative content of minerals in the samples, and transmission and scanning optical and electron microscopy (TEM and SEM) techniques are used to study the morphology and interactions between minerals. These meth- ods are time-consuming, expensive, and require extensive sample preparation prior to data collection and analysis. Reflectance spectroscopy is a quick and reliable method widely used for mineral detection and quantification. Hyperspectral sen- http://dx.doi.org/10.1016/j.jag.2017.02.018 0303-2434/© 2017 Elsevier B.V. All rights reserved.