材料科学
光学显微镜
焦炭
色谱法
数学
生物系统
扫描电子显微镜
复合材料
化学
冶金
生物
作者
Kaustav Chaudhuri,Estrella Rogel,Thomas Rea,Danielle Cuspard,Harris Morazan,Joanna Rudenko,César Ovalles
出处
期刊:Carbon
[Elsevier]
日期:2023-09-01
卷期号:213: 118282-118282
标识
DOI:10.1016/j.carbon.2023.118282
摘要
Predicting the petroleum coke or petcoke morphology tendencies of Delayed Coking feeds is paramount to optimizing refinery operation and profitability. Herein, we report a new analytical methodology employing cross-polarized light optical microscopy (CPL-OM) of Microcarbon Residue Test (MCRT) cokes, followed by image segmentation and statistical analysis to determine several structural parameters, i.e., average area, average Feret maximum, and percentage of isochromatic domains with Feret maximum <8 μm (Feret is defined as the maximum distance between the two parallel planes of a given coke particle). Several advantages over previous CPL-OM methods include using more extensive data sets and machine-learning software to accurately segment images to yield better statistics, eliminating isochromatic domains on edge to reduce error, and stitching multiple images to account for larger domains. The new CPL-OM-image segmentation-statistical analysis procedure was effectively used to determine the influence and predict the morphology of MCRT cokes based on Delayed Coking feed properties (C7-asphaltenes percentage/MCRT percentage ratio and the asphaltene solubility parameter) and in blends of vacuum resids. The structural parameters Feret Max. Percentage <8 μm and average area showed excellent agreement within the errors of the technique. This technique represents a convenient analytical tool to qualitatively rank Delayed Coking feedstocks in terms of coke morphology tendencies.
科研通智能强力驱动
Strongly Powered by AbleSci AI