pulverised coal injection

Coal Utilisation Pulverised Coal Injection in Ironmaking Blast Furnaces Angus Thomson, Paul Zulli, Malcolm McCarthy and Keith Horrocks The replacement of coke by pulverised coal injection (PCI) technology requires an understanding of the quality of the coke and pulverised coal as well as their impact on the performance of the blast furnace. This article reviews the past and current trends in PCI and highlights some of the critical factors which will affect the development of the technology and, in particular, how this will relate to coal selection. Figure 1: Blast furnace material balance Introduction The use of auxiliary fuels such as natural gas, oil, coal and other carbonaceous materials for injection into ironmaking blast furnaces is driven primarily by economic factors, especially the benefits derived from reducing the consumption of expensive coals for metallurgical cokemaking, avoiding major expenditure on existing coke batteries or avoiding capital expenditure on new batteries, as well as reducing the overall emissions from the steel plant. From a blast furnace perspective alone, the use of auxiliar y fuels decreases the overall cost of hot metal produced through increased productivity and the possibility of optimal control of the furnace operation, particularly furnace stability. Injected fuels not only replace the carbonaceous content of the metallurgical coke but provide endothermic heat to control the furnace combustion zone energy balance [Burgess]. Fuel injectants lower coke rate as well as allowing use of higher blast temperature. During the 1950s, oil was the preferred injectant due to its low price relative to natural gas and coal. A major reevaluation of alternative blast furnace injectants was sparked by the oil price shocks of the 1970s, and this provided the impetus for investigating the injection of pulverised coal as an alternative technology for ironmaking. It is now recognised that although the capital outlay associated with natural gas injection is lower than that for putverised coal injection, PCI provides the means to ultimately achieve higher coke replacement, simply by permitting higher injection rates. It is generally accepted that the maximum natural gas injection rate is limited to 125 kg per tonne of hot metal produced (kg/thm). Fig.1 shows a simplified material balance over the blast furnace. A survey of current blast furnace practices suggests that a total fuel consumption of approximately 500 kg/thm is representative of a stable operation. Although operating targets differ from steelmaker to steelmaker, the trend in PCI rates is towards a long term operating average of 200 kg/thm, representing 40% of the total fuel rate. Table 1 gives an indication of 1995 average PCI and total fuel rates for a selection of blast furnace operations. Given that (a) the coke price contributes significantly to the overall cost of hot metal, (b) a price differential of approximately A$50 per tonne exists between pulverised coal and coke, and (c) the capital demands are considered (PCI plant, oxygen plant), it will be apparent that PCI represents a substantial incentive for cost reduction in the steel plant. This paper reviews the historical development of PCI as a cost-effective technology for ironmaking and attempts to address several issues associated with the implementation of PCI, including cokemaking strategies, coke quality and pulverized coal quality. Future investigations required to implement PCI in the blast furnace are also discussed. A Historical Perspective Pulverised coal injection has been practised in some blast furnace operations since the early 1960s, for example by the National Steel Corp. in the USA and Shoudu in China. In the 1980s, interest in PCI escalated, prompted mainly by dramatic increases in the price of fuel oil in 1973 and again in 1979, and its potential as a coke replacement began to be realised. In the early 1980s coal injection facilities were installed throughout Europe and Japan, with injection rates normally between 40 and 90 kg/thm (maximum rates up to 435 kg/thm, practiced by Usinor in France and Sboudu in China) and with coke replacement rates of the order of 0.9 kg of coke/kg of coal. From this period, investigations into PCI dealt predominantly with the behaviour of the coal following injection into the blowpipe, particularly the combustion behaviour of the coal. Even at this stage, plant operators were concerned that inefficient combustion in 42 The Australian Coal Review October 1996 Coal Utilisation the blowpipe/raceway zone would result in unburnt coal particles depositing within the void spaces of the burden, hence adversely affecting furnace permeability and productivity. Combustion studies revealed that high combustion efficiency (greater than 75% burnout of the coal) required high volatile content coal (greater than 25%) and a finely ground particle size distribution similar to that used in coal fired power generation plants, viz. 7580% less than 74 micrometres. Not all operators subscribed to this view, namely the Shoudu plant, where anthracite coals were commonly used along with other coal types, and British Steel Corporation at Scunthorpe, with their granular coal injection system (coal size –95% less than 2 mm, 20% less than 75 micrometres). In both cases, combustion efficiencies would have been very low, perhaps less than 30%, yet the operations were apparently successful. During the late 1980s and early 1990s, coal injection rates escalated substantially, with levels of approximately 150 kg/thm relatively common. In some cases, rates as high as 180 kg/thm were used and trials at levels in excess of 200 kg/thm were carried out [Peters et al; Yamaguchi et al 1992; Kamijo et al’ Maki et al]. The objective of these and subsequent plant-related investigations has been to develop the technology with which to operate the blast furnace at the highest possible PCI rate with the highest possible level of coke replacement. This has focussed the research on not only improving the combustibility of coal in the blowpipe/raceway, but also on better understanding the impact of PCI, pulverised coal quality and coke quality on the permeability of the furnace, and hence maintenance of furnace stability and productivity. Implementation of PCI Cokemaking Strategies The future of coke supply is a critical factor for many integrated steel plants. In particular, issues such as coke price and quality, and declining internal and external supplies, are major mediumterm problems to be solved. Coke supply (both quantity and quality) has been constrained mainly by the continual aging and attrition of the existing coke oven population, financial limits on new oven construction and increasingly stringent environmental regulations. To an extent, the coke supply issue has been tempered in recent years by such developments as PCI and continued expansion of electric furnace steelmaking [Hogan]. However, the former presents a number of new challenges for the blast furnace and coke oven operators, especially concerning coke quality. Fig. 2 shows the average age of coke capacity for various countries in 1994. Most of the current coke production capacity, particularly in the US and Japan, has resulted from post-war developments in the 1950s through to the 1970s. In Japan, for example, 44% of the coke ovens still operating are between 21 and 25 years old [Hogan], and even with sound strategies in place to extend the coke oven life, and aims for relatively optimistic targets of 40 years life, by the years 2010 to 2015 many of these batteries will be scheduled for replacement. The capital cost associated with building a battery of coke ovens to the latest environmental standards is significant, with a complete rebuild of an American 820,000 t/a coke battery being quoted as A$375 million [Hogan]. A significant component of the cost of new installations is the cost of compliance with increasingly stringent environmental laws. In Germany, a significant reduction in emissions requires the application of dry quenching of coke which adds in excess of A$175 million to the capital investment for a two million tonne batter y [Hogan], together with the added operating cost of using nitrogen rather than water. With these costs, many steel producers are closing old batteries and buying coke on the open market [Agarwal et al]. This can only be regarded as a short term solution, as the current excess of world coke supply is rapidly dwindling [Nilles] and plans for rebuilding old coke ovens are not progressing fast enough. Table 2 shows estimated and forecast international cokemaking capacity The Australian Coal Review October 1996 43 Coal Utilisation to the year 2005, using an exhaustive survey of production estimates from individual coke producers. While the total cokemaking figure is predicted to be reasonably stable to 2005, the average is only maintained by China and 'Other Asia' countries (outside Japan) predictions. These expansions will not lead to a great increase in available coke, as they are only helping to ser ve a continued growth in steel production in the Asian region. India, South Korea, Taiwan and most importantly China are expected to increase their production of pig iron over the next decade. By 2005, the global demand for coke will be around 15 million tonnes higher than in 1994 [Cleary]. Coke Quality Requirements High quality coke is a key factor for stable blast furnace operation, especially when high productivity operations are practiced, i.e. at high injectant rates and reduced coke rates. Fig. 3 shows a schematic representation of the internal state of the blast furnace. The furnace is a high-temperature, highpressure reactor in which hot reducing gases formed from the combustion of coke and auxiliary fuels in the raceway region, travel counter-current to the burden materials (iron ore, agglomerated ore, coke and fluxes) charged in from the top. As well as providing chemical and thermal energy for the reduction of iron ore, coke also forms part of the inter- nal furnace structure and consequently plays a very significant role in ensuring the furnace remains permeable to gas flow. At high injectant levels, the proportion of coke to ore in the furnace decreases considerably and, hence, the permeability of the burden materials to gas flow decreases. The decreased ore to coke ratio also leads to a lower burden descent velocity for a given iron production rate. Since the residence time of coke inside the furnace increases as PCI increases, it becomes apparent that coke specifications must be reviewed at high injectant rates. The main coke properties relevant to the implementation of PCI include ash content, sulfur content, alkalis, mean size and size distribution, mechanical strength and coke strength after reaction (CSR). The latter is particularly relevant for high PCI operation, as coal injection brings about changes in the mechanism of coke degradation in the blast furnace, especially in the lower regions of the furnace [Steller et al]. As PCI increases and the residence time of coke inside the furnace increases, the reaction of coke carbon with carbon dioxide (known as 'solution loss reaction') increases. This reaction promotes coke degradation and the production of fines in the lower regions of the furnace (deadman and dripping zones, Fig. 3), which then hampers the flow of gas exiting the raceway. It is clear from more recent investigations [Yamaguchi et al, 1996] that measures to suppress coke degradation and fines generation are actively being sought. Pulverised Coal Quality Requirements Pulverised coal quality is important not only in terms of the utilisation inside the blast furnace, but also with respect to the preparation, handling and transportability of the coal. As well as the chemical composition and physical properties, particularly phosphorus, sulfur, alkalis, chlorides and ash fusion temperature, the important properties in pulverised coal utilisation are the combustibility of the coal, calorific value and assimilation characteristics of the remnant coal and char particles including ash. In general, high volatile matter coals (low-rank) have, in the past, been preferred for PCI, since combustibility is improved with increasing volatile matter content (discussed below) [Haywood et al]. On the other hand, the calorific value of the coal, which has a significant effect on the coke replacement ratio and furnace stability (effect on the adiabatic flame temperature), should be as high as possible. Hence, recent studies have been carried out to investigate the use of higher rank coals (higher calorific value) for high-rate PCI. The full realisation of the potential of the high-rank coals as suitable injectants at very high rates is likely to require new developments in injection technology to minimise the impact on furnace operation. The assimilation behaviour relates to both the reactivity and wettability of chars (and coals) in slag and iron, and 44 The Australian Coal Review October 1996 Coal Utilisation is critical where combustion efficiency is low. The required handling characteristics of the coal can be quite variable according to different PCI users since the coal preparation and injection equipment designs vary greatly. The Hardgrove Grindability Index (UGI) should generally be between 30 and 70, this being limited by the ability of the grinding mill to supply, and to minimise breakdown in size during handling and injection. Particle size distribution affects the amount of carrier gas required to transport the coal and abrasion of the conveying equipment, while moisture content of the mill feedstock affects both the energy consumption for drying and the mill output. Current Studies of PCI As discussed above, the replacement of coke by PCI technology requires considerable understanding of the coke and pulverised coal quality requirements, and the impact of these on furnace performance. At BHP Research, a major research program was undertaken on behalf of BHP Australia Coal to examine the suitability of selected Australian coals for high rate PCI [Haywood et al]. Combustion experiments were carried out on the test coals using a large pilotscale, single blowpipe test facility which closely simulated the conditions in the blowpipe/raceway region of the blast furnace, with coal rates up to 300 kg/thm and blast temperatures up to 1300ºC. Fig. 4 shows a schematic of the experimental apparatus used, while Fig. 5 illustrates the general layout of the equipment. Results from the test work illustrated the effect of coal properties on combustibility under high rate PCI conditions. The primary coal property affecting the combustibility was found to be volatile matter content, as shown in Fig. 6. Other experiments (not shown), particularly relating to the off-axis burnout of coal particles, showed that injection lance design and oxygen enrichment could alter the behaviour of a given coal considerably. Two other experimental programs related to the study of remnant char particles (combustion products) were carried out in conjunction with the above combustion studies: chemical reactivity in C02, H20 and 02, by the CSIRO Division of Coal and Energy Technology, and characterisation of the morphology via petrographic analysis, by the University of Newcastle. Together with the development of a numerical model for predicting the total burnout of the coal to the back of the raceway, a comprehensive assess- ment of the combustion performance of the selected coals for PCI application up to 300 kwthm was obtained. Future Studies of PCI In conclusion, the number of blast furnace PCI installations will continue to increase, and injection levels will continue to rise as a consequence of the critical shortages in coke which are expected to emerge by the turn of the century. There will be many operations that will settle for what will be regarded in the near future as a fairly conservative upper injection level of about 150 kg/thm. However, the more ambitious operators or those with more desperate coke supply problems will push beyond today's maximum of 200 kg/thm, aiming for the 'real' injection limit predicted to be at least 250 kg/thm by Yamaguchi and others. Those operations which can sustain such injection levels consistently will reap substantial rewards in terms of further reductions in operating costs and delayed or minimised capital investment in new coke ovens. However, these intense injection operations are unlikely to be achieved without further significant R&D input to fine tune every aspect of the PCI process. This means that the process mechanisms of all the physicochemical interactions between the injected coal, and the furnace burden structure and process streams must be understood in order to provide a guide to the setting and optimising of furnace operating parameters. The critical factors involved in understanding these interactions include: s Coal combustion/gasification • effect of coal type on reaction behaviour with oxygen/air and blast furnace raceway gases, • effect of injection conditions on the combustion including very short reaction times (~10 ms), moderate pressure (~4-5 atm) and high gas temperature (1000–1300ºC), The Australian Coal Review October 1996 45 Coal Utilisation • effect of lance design and method of oxygen injection on coal/oxygen mixing, and the consequent effect on combustion. s Internal furnace processes • mechanism of particle deposition/release within the burden structure — will depend on coal/char properties, ash properties, particle size, burden structure characteristics etc, • reactivity of unburnt coal particles with rising furnace gases, • dissolution/assimilation mechanisms of coal/char/ash particles with descending slag/metal streams, • effect of burden voidage structure on particle deposition/release processes. s Material properties • effect of coal properties such as volatile matter content, char reactivity, particle size and ash properties on combustion and in-furnace processes, • effect of coke quality on blast furnace burden structure for different levels of PCI. Since the late 1970s, many international studies have been conducted on coal combustion under PCI conditions. Much information is now available on this topic though some gaps still remain, particularly concerning the effect of blast furnace pressure and the influence of coal/oxygen mixing following injection, on the combustion process. However, development efforts in a number of research institutions and operating plants are currently addressing these issues. By contrast, mechanisms of the entrapment/dissolution /assimilation processes occurring within the furnace burden structure are understood only qualitatively at best. Similarly, the combined effect of PCI and coke properties on the fundamental processes occurring within the burden structure is largely unknown. Yet it is precisely in this region of the furnace that adverse furnace response to PCI effects will start. Research and development effort directed towards understanding these complex solid/powder/ multi-liquid phase/gas interactions will make a significant contribution towards coal selection for all levels of PCI, and to the development of blast furnace operating strategies suitable for the most arduous of PCI operations. References Agarwal J, Brown F, Chin D, Stevens G, Clark R and Smith D, 1966. The future supply of coke. New Steel, 12, p. 88. Burgess J M, 1979. The potential of coal as a blast furnace injecting. BHP Tech. Bull., 23, p. 66. Cleary P, 1996. The medium term outlook for metallurgical coke. Cokemaking International, 8, pp. 28-32. Haywood R , McCarthy M J, Truelove J S, Mason M B and Thomson A D, 1995. An experiment and theoretical investigation of pulverised coal combustion in blast furnaces. Australian Symposium on Combustion and Fourth Australian Flame Days, University of Adelaide, SA. Hogan W T, 1992. The future world crisis in coke. Iron and Steel Engineer, 69, p. 32. Kamilo T, Takahashi N, Shimizu M, Yoshida Y and Ito R, 1994. Analysis of phenomena in lower part of blast furnace at a high pulverized coal rate operation. Proc. 1st int. Cong. on Science and Technology of lronmaking, ISIJ, Tokyo, P. 505. Nilles P E, 1996. Alternative technologies in iron and steelmaking. Metallurgical and Materials Transactions B, 27B, p. 541. Peters K H, Mohnkern H, and Lungen H B, 1994. Ways and means for low coke blast furnace operation, Proc. Ist lnt. Cong. on Science and Technology of lronmaking, ISIJ, lbkyo, p. 493. Steller J M, Lao D, Lebonvallet J L and Helielsen M, 1996. Development of coal injection in the blast furnace at Usinor Sacilor, injection technology in ironmaking and steelmaking. Proc. IISI Technical Exchange Session, Brussels, p. 15. Yamaguchi K, Ueno H and 'Tamura K, 1992. Maximum injection rate of pulverised coal into blast furnace through tuyeres with consideration of unburnt char, IS]J Int, 32, p. 716. Yamaguchi K, Uno T, Yamamoto T, Ueno H, Konno N and Matsuzaki S, 1996. Coke degradation mechanism and suppression measures during high-rate pulverised coal injection. Tetsu-toHagane, 82, p. 7. Angus Thomson and Malcolm McCarthy are with BHP Research, Newcastle Laboratories. Paul Zuili is with BHP Research at the Port Kembla Laboratories and Keith Horrocks is with BHP Flat Products Division at the Port Kembia Steelworks. 46 The Australian Coal Review October 1996