Thomas, George
(2022)
Fluidity development of coking coal.
EngD thesis, University of Nottingham.
Abstract
In most of the previous analytical studies on coking coal fluidity, single coal behaviour is well-documented but few studies have addressed coal blends regarding interactions in blends leading to possible non-additive behaviour. Many different techniques have been used to study single coals and blends, but of the two most promising are high temperature rheometry and 1H NMR. However, there is little work reported to date on using these techniques to study coal blends and this is the primary focus of this investigation.
For rheometry, a novel cup geometry has been developed for its efficacy to complement the already established parallel plate (PP) geometry. A suite of 10 coals were studied and used throughout this project, initially to compare the PP and cup geometries. The PP suffered from sample loss, while the cup geometry displayed extended expansion due to the retaining walls and this can potentially provide a proxy for gas pressure. Complex viscosity profiles showed fair agreement between the two geometries for high fluidity coals and reasonable similarities for medium fluidity coals. One major benefit of the cup was found to be use of larger masses, compared to 1.5 g for the PP, with up to 3 g for medium and high fluidity coals, making it more favourable for investigating blends.
Using the cup geometry, a range of binary blends were studied and viscoelastic properties of the blends (minimum complex viscosity (η*min), complex viscosity (η*) and tanδ) were compared with the development of gas pressure to quantify any interactions between coals, such as adsorption of fluid species on the remaining non-softening matrix. Significant deviations were found from additive behaviour for η*min, where both negative and positive interactions were found. Correlations with elastic modulus (G’) and loss modulus (G”) were used to rationalise that G’ can provide information about interactions involving macromolecular changes and that G” can provide information about interactions involving mobile molecular entities. These were subsequently linked to the gas pressure, where it was found that volatile absorption was evident when blending non-coking coals with fluid exhibiting coals. Additionally, the use of the cup geometry has shown that blends exhibiting η*min of 2 – 3 x10^3 Pa s provide dangerous swelling. This was shown to be the limit of which sustained swelling occurred, or weak – semi cokes are formed due to volatile release from low η* suspension formation.
High temperature 1H NMR has been used to understand the mobile molecular behaviour of the 10 coals where %fluid hydrogen (H) and mobility of the fluid phase (T2L) have been studied to quantify changes in such moieties. %Fluid H and T2L at maximum %fluid H increased as a function of volatile matter content and bulk fluidity, apart from the high volatile coal 3 for the latter. %Fluid H exhibited similar generation profiles as a function temperature for all coals. For T2L, non- and medium fluid coals exhibited maximum mobility at or below 400 oC, below that of maximum %fluid H, followed by a decay with increasing temperature, whilst high volatile coals exhibited maximum mobility close to maximum %fluid H. For blends, additive and non-additive behaviour was exhibited, where the coal with the highest mean vitrinite reflectance (MVR) and anomalous behaving coal 3 featured heavily in additive deviating blends with respect to %fluid H. For T2L, fewer blends displayed additive behaviour, which was most prevalent for the majority of blends containing non-fluid coals. Consequently, it was evident that the apparent non-fluid coal 3 was distinct on possessing micro-fluidity detectable by 1H NMR.
Limited correlations were found for single coals between rheometry and 1H NMR, where G” and η* exhibited R2 values of between 0.8 – 0.9 with the %fluid H. However, good correlations were found between G” (0.90 – 0.99) and η* (0.89 – 0.98) with %fluid H for medium fluidity coal blends. Consequently, viscous behaviour was predominantly dictated by mobile molecular entities, whereas elastic behaviour was dictated by bulk phenomena. Ultimately, NMR exhibited fewer deviations from additive behaviour compared with η*min.
Macroporosity and total porosity for the semi-cokes obtained exhibited good correlations with maximum plate gap (R2 > 0.9), where the cup geometry could be used in future studies to predict porosity. This was due to good sample retention from the novel cup geometry. Regarding the microfluidity generated for coal 3, the semi-cokes had high microporosity, but low meso- and macroporosity, explaining why the fluid materials could not propagate through the bulk of the sample.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[img]](/style/images/fileicons/application_pdf.png)