Doctor of Philosophy
School of Geosciences
Mandile, Andrew J., Application of the Rietveld method to oil shale sequences and related rocks, Doctor of Philosophy thesis, School of Geosciences, University of Wollongong, 1996. http://ro.uow.edu.au/theses/1964
The Rietveld method is a standardless technique that corrects for all X-ray diffraction (XRD) intensity aberrations. From crystal structure data and least-squares refinement of Rietveld parameters, a Rietveld analysis fits a calculated XRD pattern to an observed XRD pattern to calculate the weight percentage of minerals in a sample. In the Rietveld program SIROQUANT , used in this study, the crystal structure data are converted into (hkl) files that are used for calculated pattern modelling. SIROQUANT allows amorphicity and microabsorption (Brindley) corrections to be made in the Rietveld analysis.
In order to apply the Rietveld method to samples from Australian Tertiary oil shale sequences, a methodology was developed so that the Rietveld method could quantify clay minerals, organic matter could be quantified from XRD data, and the Rietveld method could be used as a rapid and routine procedure.
Calculated pattern fitting of the XRD patterns for illite, illite/smectite and montmorillonite was achieved by using observed (hkl) files which allowed for the imperfect crystal structure of these minerals. The observed (hkl) files were calculated by measuring (hkl) file data directly from standard clay mineral XRD patterns for each of the clay minerals. A two halfwidth function was tested on the Duaringa clay mineral patterns and although the calculated fits improved, there was only a small difference in the quantitative results. Intensity calibration curves were used to correct for intensity losses caused by beam overflow and the X-ray absorption properties of each rock type in the Duaringa sequence.
The absolute mineral abundances were calculated using the Rietveld method by including the organic content in the Rietveld analysis. The organic matter was quantified with a calibration curve which related the area of the organic (amorphous) hump to the organic content Calibration curves were calculated with the direct method, which involves quantifying the organic content in 'spiked' calibration mixtures using the Rietveld method, or the addition method, which involves preparing calibration mixtures with a standard organic matter sample. An iron-correction curve and formula were used to correct for microabsorption effects produced by iron-bearing minerals.
A user-friendly method consisting of a four-file system and Parameter Groups was developed to analyse a large number of samples using the Rietveld method. The four files stored and allowed easy manipulation of files used in a Rietveld analysis. From each Parameter Group, a representative parameter (PAR) file was used to prescale the Rietveld parameters in all the other PAR files in the same Parameter Group; this increased the speed of the Rietveld analysis by a factor of five to ten. The only Rietveld parameters that require refining for the Rietveld analysis of the non-representative PAR files are the more variable Rietveld parameters such as scale and preferred orientation.
One hundred and fifty samples from the Tertiary Duaringa oil shale sequence, consisting of oil shale, claystone and carbonate-rich rocks, were analysed using the Rietveld method. The quantitative mineral data of each sample was used to interpret lithological changes in the sequence. Illite/smectite, quartz and kaolinite are the dominant detrital minerals in the oil shales with minor albite, montmorillonite and anatase. The oil shales formed during relatively high water stands where minor siderite and pyrite formed in reducing conditions at the bottom of the lake. Siderite layers, which contain greater than 30 wt% siderite, formed in a zone which occurred in between an anoxic zone, which produced Fe++, and a methanogenesis zone, which produced C03-.The depth of water in the lake was lowered due to relatively low precipitation levels. Claystones formed at these times and are composed of the same detrital minerals that were deposited during the formation of the oil shale. However, the shallower waters produced higher energy conditions in the lake which in turn increased the deposition of the coarser-grained and slightly higher density minerals such as quartz, albite and anatase relative to illite/smectite and montmorillonite. The distribution of kaolinite, and to a lesser extent albite, was mainly influenced by their availability in the detrital provenance. During relatively high precipitation levels during the formation of claystones, sandy layers and sandstones were formed. The water level in the lake was low because of arid climates. These conditions produced saline-rich waters which formed analcime and dolomite. Ostrocodes lived on the sediment during low saline and highly oxic conditions and represent the main proportion of calcite in the sequence. Diatoms lived in the lake waters during the deposition of some oil shale intervals and were later diagenetically altered to opal-CT. Relatively low water levels during the formation of oil shale favoured the formation of carbonaceous shales in the lake and coals in peat swamps.
The Rietveld methodology was also applied to samples from the Triassic Leigh Creek coal sequence and reservoir rocks from the Macedon-Pyrenees gas and oil fields.
The Leigh Creek sequence comprises coal and oil shales with minor siderite-rich lithologies. The detrital minerals in the oil shale consist of co-dominant kaolinite, quartz and illite. The abundance of these minerals is fairly consistent throughout the sequence whereas the organic matter gradually increases from the lower oil section to the upper oil shale section. Siderite occurs as a minor mineral in the oil shales and in siderite layers which contain greater than 30 wt% siderite. Quartz and kaolinite are also co-dominant minerals in the coal but illite is absent and the kaolinite in the coal is authegenic. The kaolinite to quartz ratio is higher in the Main Series coal compared to the Lower Series coal. The difference in this ratio between the two coal zones and the higher organic matter content in the Main Series coal suggests that the peat swamp environment during the formation of the Main Series coal zone was more established than the peat swamp environment that existed during the formation of the Lower Series coal zone. The Rietveld data produced a better database for managing mining and combustion operations and assessing mine wall stabilities.
The Rietveld method was useful for quantifying matrix minerals in the reservoir rocks from the Macedon-Pyrenees field which cannot be quantified in thin section. Vertical variations of the matrix minerals, such as illite/smectite and kaolinite, are important for assessing the potential effects of formation damage. Glauconite was a minor but important mineral and it was necessary to use a 'semi-observed' glauconite (hkl) file which was modelled on an observed illite (hkl) file. Rietveld quantitative mineral data should be useful for calibrating geophysical logs and correlating mineralogical trends from one well to another.