ACARP Project Number: C12057
Published: March 08
Philip Bennett, Adrian Reifenstein, Graham
O’Brien, Barry Jenkins
Extended
Abstract
Coke quality in the blast furnace is measured by
its resistance to degradation. The roles of coke as a fuel and as a
reducing agent are not limiting factors in terms of blast furnace
performance. However, the role of the coke as a permeable medium is
crucial in economic blast furnace operation. The degradation of the
coke varies with the position in the blast furnace but in all cases
involves the combination of reaction with CO2, H20 or 02 and the
abrasion of coke lumps against each other and other components of
the burden. In trying to relate coke quality to its ultimate use in
the blast furnace, one of the primary attributes of the coke
quality is the coke strength.
The Coke Reactivity test is a highly regarded
measure of the performance of coal. This test has two
components; the Coke Reactivity Index (CRI) and the Coke Strength
after Reaction (CSR). A coal which, when coked, achieves a
low CRI value and a high CSR value is highly regarded in the
market, primarily because this test has been related to blast
furnace performance, particularly fuel rate and permeability of the
burden.
Coal characteristics which have been shown to
influence coke reactivity include coal rank, ash %, various
components of the ash chemistry and full maceral reflectograms. In
most Australian coals these coal characteristics account for the
major proportion (~ 70%) of the variation seen in the Coke
Reactivity Index (CRI) and the Coke Strength after Reaction
(CSR).
It has proved difficult to quantify what controls
the remainder of the variation observed in the coke reactivity
test. It is known that ash percentage and ash chemistry
control variability to some extent, with some minerals or elements
catalysing the gasification reaction. However, the equations
developed using these elements can, at times, be poor predictors of
the CRI and CSR values.
This project examined three distinct
characteristics of the coke and their impact on coke quality as
defined by the NSC Reactivity test. They were:
-
Compressive strength testing of pre-and post-reaction cokes to
allow some understanding of how gasification effects coke
strength.
- The coke
lump shape and surface area to volume ratios provided additional
information on the gasification behaviour of coke; whether it is
preferentially gasified on the surface of the coke lumps or whether
it is a pervasive effect.
- The
structural elements of the coke included porosity and fissure
formation. These elements were considered to impact on coke
reactivity because they have some affect on permeability of the
coke and thus the ability of reducing gases to penetrate into the
coke lumps.
Together these tests provided a more complete
understanding of how the nature of coke influences the coke's
inherent reactivity and strength.
The CRI of the cokes produced in this project
followed the same relationship with CSR as other cokes in
ALS-ACIRL's extensive database of coking test results. The
optimum maximum reflectance range for low CRI and high CSR is 1.3
to 1.5%. General trends, such as decreasing CRI (increasing CSR)
was associated with increasing semifusinite and increasing CRI
(decreasing CSR) was associated with increasing liptinite and
vitrinite, were found in this project. There is a lack of
strong relationships between macerals and coke microtexture; this
was also found by Sharma et al. [2005]. There was no evidence that
more of one microtexture was consumed due to the greater reaction
with CO2 of the crushed coke compared to the cored coke. This
implies that the more reactive microtextures, such as non porous
isotropic derived from inertinite, are consumed early in the test
and the remaining less reactive material in consumed at a constant
rate. It was shown that the difference in reactivity of the crushed
and cored samples is only due to the shape of the particles. Thus,
accicular coke, with a larger surface area compared to more equant
or blocky coke, should give a higher CRI. To determine the
extent that shape will influence the CRI all that is required is
for the equivalent spherical diameter to be calculated for the coke
tested. For the coals used in this project, there was minimal
difference in shape of the crushed cokes and this could be due to
the crushing required to prepare the sample for the reactivity
test.
The compression test proved to be a good
indicator of the inherent strength of coke. The simple sample
preparation and testing allowed the testing of 40 samples of the
pre reactivity cokes and over 30 samples of the post reactivity
cokes. It was shown that the Young's modulus of surface
breakage of the cored particle was lower than the Young's modulus
of the unconfined compression breakage and therefore surface
breakage would be the initial breakage mode of a particle of
unreacted coke for the coals used in this project. This was
not the case for coke after reaction with CO2 where the Young's
modulus of surface breakage was similar or in some cases higher
than the Young's modulus of the unconfined compression. The
Young's modulus for surface breakage of the pre reactivity coke
decreased with porosity and number of pore walls greater than 50
µm.
The relationship between CSR and compressive
strength post reactivity followed the trend as that suggested by
Andriopoulos et al. [2002] where drum related matters moderate the
effect of the bulk material property. The compression test
gave a better differentiation for the higher strength cokes.
Porosity was found to have a major influence on
coke strength, which suggests that coke textures play only a
minimal role in determining coke strength. There was only a
general trend that indicated that there is a minimum in porosity
corresponding to the optimum Romax for maximum pre reactivity coke
strength. The vitrinite concentration had a strong influence
on the formation of large pores and therefore porosity.
The porosity increases after reaction with CO2
thus leading to weakening of the coke structure. This
increase in porosity can not be directly related to the rank or
macerals of a coal, though the increase in porosity depends on rank
of the coal and the microtexture of the coke. There was no evidence
that the increase in porosity was limited to the edges of the
coke. The non-porous isotropic microtexture, roughly related
to the inertinite macerals other than semi fusinite, does
contribute to the loss in strength of the post reactivity coke, but
seems to be rank dependent with the lower rank coals suffering a
greater reduction in coke strength. Coals with a Romax
greater than 1.3% did not show an increase reduction in coke
strength post reactivity test with increasing inertinite less semi
fusinite. Coke microtexture analysis was not done on the post
reactivity coke. However, a subjective visual inspection of
some coke images suggested a decrease in inertinite-derived
material in the post reactivity.
Not all of the reduction in coke strength,
especially for coals with a reflectance below 1.3%, could be
attributed to the amount of non-porous isotropic
microtexture. By examining the different grey levels in the
images it was found that material with a grey level between 75
(reflectance ~ 7%) and 110 (reflectance ~ 9%), called grey 2
material, was lost from the coke structure for coals with a Romax
of less than 1.3%. In this rank range and when the grey 2
material was concentrated in small pore walls then there is a large
reduction in coke strength due to the attack of this material by
CO2. For coals with a reflectance less than 1.3%, this grey 2
material correlated with the sum of the fused carbon domains very
fine and fine.
These results highlight the need to associate the
coke microtexture with the microstructure. That is, the size
and composition of pore walls have a strong influence on the
changes that occur within the coke due to reaction with CO2 and
therefore on coke strength after reaction.