Font Size: a A A

Statistical analysis of coal quality parameters for the Pennsylvanian system bituminous coals in eastern Kentucky, United States of America

Posted on:1994-12-21Degree:Ph.DType:Dissertation
University:The University of North Carolina at Chapel HillCandidate:Collins, Steven LeeFull Text:PDF
GTID:1471390014494832Subject:Geology
Abstract/Summary:
Multivariate statistical analyses were performed on 372 samples from 35 coal beds of Early and Middle Pennsylvanian (Namurian through Westphalian D) age. Most samples are full-bed channel samples. Compositional variables were transformed to centered logratios. Minimum and maximum, and 1st, 5th, 25th, 50th, 75th, 95th, and 99th percentiles are reported for variables before transformation. The medians of mean maximum vitrinite reflectance, dry ash, and dry sulfur are 0.78%, 9.7%, and 1.0%. Mean, standard deviation, skewness, kurtosis, Lilliefors tests of normality, and the 50 largest magnitude correlation coefficients are reported for transformed variables.;Principal components analysis of 84 variables yields three interpretable principal components. Spatial trends in component scores are quantified by linear regression. Mineral content increases toward S56;Basement features identified by previous workers from gravity anomalies affected coal sedimentation. Mineral and sulfur contents of coals are high in the topographically low Floyd County channel. Coals over the topographically high Perry County and Pike County uplifts have low mineral and pyrite contents. Hot fluids expelled during the Alleghany orogeny may have deposited minerals and organically bound elements, and increased vitrinite reflectance as they traveled N43;Principal components analysis of eight rank-related variables yields one rank principal component. Rank is anomalously low over the Rome trough and Floyd County channel where thicker sediments shielded coals from heat generated in the crystalline basement.;Statistical analyses of 22 potentially harmful trace elements show that they are primarily associated with aluminosilicate minerals and pyrite. Cleaning coal by density separation probably would concentrate trace element pollutants in mining waste and reduce them at the point of use. Relative Trace Element Concern indexes are defined to quantify the potential for coals to cause environmental harm. Maps of these indexes, which have non-nugget effect variograms, show that coals from different areas vary considerably in their pollution potential.
Keywords/Search Tags:Coal, Statistical
Related items