Microstructural Quantitative Analysis of Polymer amended Fluid Fine Tailing’s Using Digital Image Processing Techinques

Abstract

Understanding long-term dewatering behavior of oil sands tailings is significant to the success and optimization of tailings reclamation plans. Long-term dewatering of tailings is somewhat complex due to mechanical creep, structuration drove by electrochemical forces between clay particle and other potential mechanisms such as cementation. To modernize conceptual models of tailings dewatering, the fabric evolution of tailings over months has been studied using low vacuum SEM as well as optical microscopy methods.Morphological information is extracted from different images using a sequence of image processing techniques, designed by trial and error to produce repeatable information for the studied material, flocculated fluid fine tailings (FFT). The designed image processing method produces binary images defining pore-space and solid particles (solid grains or flocs). The binary images are then analyzed using a software, primarily "Fiji-Image J"for quantitative analysis of the image for tends in pore and particle size distributions. The samples that were imaged were obtained from amended FFT dewatering experiments, comprising many replicates of 0.10 m high columns that were sampled over months. The tailings were dosed with a high molecular anionic polymer at two doses (600 and 800 ppm polymer per dry tailings). Columns either had one way or two-way drainage. Repeatable trends were found for all samples for both image processing techniques. The most dramatic observation was the increase in floc size over 48 hours using optical microscopy.iii DedicationThe thesis is dedicated to my wonderful parents, Retd. Prof.

“Microstructure Analysis of Polymer amended Fluid Fine Tailings Using Fiji-ImageJ Image analysis software”

“…To the author's knowledge, the only previous work that assessed the microstructural level of tailings using optical microscope analysis in combination with Fiji-ImageJ was done by Khattak (2018).The software was previously highly dependent on the user's input making the process more time consuming and subjected to variability from one trial to another. Hence, this study is based on enhancing and building on the methodology used in Khattak (2018) by producing repeatable results, limiting user's input through the use of the machine learning algorithms in ImageJ software in a time efficient manner.…”

“…Image analysis of optical microscope images are used in this thesis. While the only previous work of using optical microscopy to understand morphology in tailings or even soft clays has been Khattak (2018), considerable work has been done on interpretation of SEM images using imaging processing software. Khattak's (2018) work is reviewed at the end of the chapter, here follows a review of image analysis of SEM images in clays.…”

“…The SEM is a type of electron microscope that produces images of the sample by using a beam of high energy electrons to scan the surface. The electrons interact with the sample creating signals that contain information about the sample composition, texture/ morphology and topography (Khattak, 2018). A scan electron microscope consists of an electron gun with an electron column (creates the beam of high energy electrons), a sample chamber and an electron detector, which records and monitors the signals produced.…”

“…The newer optical microscopes produce a high-resolution live image of the sample when connected to the computer. One of the major limitations to older optical microscope devices is the low resolution, usually only few hundred nanometers due to light diffraction limit (Khattak, 2018). High powered optical microscopes are designed to enable capturing the smallest features by using a system of lenses and backlighting.…”

“…Optical microscope technique allows the light to penetrate through the sample to obtain the images, however, after 48 hours the polymer amended FFT become denser, hence it is challenging to analyze. Khattak (2018) studied the use of optical microscope images for the quantification of the fFFT fabric and morphology. The raw images were then processed using image analysis software Fiji-ImageJ to enhance the images by treating the images with filters and adjusting the contrast and brightness to distinguish between the flocs and the voids.…”