Development of a Combustibility Predictor for Thermal Coal Utilisation in Pulverized Fuel Boilers

Coal is a heterogeneous material, consisting of a range of carbonaceous matter and a variety of minerals. Each particle is different from the next, and because of its diversity, each particle behaves differently during combustion. Coal in general is talked about in terms of averages or totals, e.g. total minerals or average volatile content. Using averages does not allow a good understanding when only parts of the coal contribute to a problem. For example, carbon in ash or early ignition are issues that result from portions of the coal.

This project aimed to assess if the pulverised coal density distribution could be used as an index for combustion, so that a washability type curve for combustion could be used to assess the proportion of the coal that could be correlated with issues such as carbon in ash.  

Four coal samples were separated into density splits using a reflux classifier. The average density of each split was measured before it was combusted in a drop tube furnace.  The burnout produced by the density splits from three of the coals was found to decrease with increasing density. A fourth coal produced the opposite trend with burnout increasing with increasing density. Vitrinite content of this coal decreased with increasing burnout and density, indicating another factor other than maceral reactivity is affecting its combustion behavior, though it is unclear at this time why the fourth coal behaved in this manner.

Density is a function of the phases present, and as minerals in coal have a high density compared to the carbonaceous material, they can dramatically affect the value of the density determined. Therefore, TIMA analysis was used to determine the density associated with the minerals in the splits, allowing the density of the carbonaceous material to be determined for selected samples. Burnout and carbonaceous density produced a parabolic trend, with a minimum around 1.25 g/cm³. This has not been observed before in other studies but aligns with previous work that shows a parabolic trend between density and carbon content.  It appears an over estimation of minerals by TIMA analysis is occurring, which results in lower carbonaceous densities for samples containing large amounts of mineral matter. While this is not expected to be the cause of the minimum observed, it may cause the value of the minimum to move when corrected. More work is required to ensure the observed minimum in burnout with density applies to other coal ranks, especially higher ranks that would be expected to have lower combustibility. Additionally, work to understand what causes the minimum observed is also required.

Plotting a cumulative curve of burnout with carbonaceous density allows coals of different rank and combustion behavior to be compared across the range of reactivities for each of the different portions of the coals. It also provides an understanding of the portion of material responsible for issues associated with poor or extreme burnout that would otherwise not be determined if only the average properties of the coal were available.