Several obstacles must be overcome for any industry with the objective of sustainability. Two major hurdles are: radical improvement in energy efficiency and zero waste. To realise this a step change improvement is required from current strategies of energy optimisation and waste minimisation. Many of these improvements cannot be achieved in isolation. TXI (Texas Industries Incorporated) and Hatch have found several solutions to waste and energy problems by applying the practice of By-Product Synergy. This is modelling industry metabolism as a natural ecology- Industrial Ecology. TXI’s cement division and Chaparral Steel developed the STAR (Systems and Technologies for Advanced Recycling) project focusing on the respective cement and steel works with the aim of eliminating waste. Several innovative ideas emerged ranging from the consumption of alternative fuels to recycling of auto shredder fluff. This was the genesis of the CemStarSM technology. Research began in 1994. By 1995 the technology was patented and incorporated into TXI’s Hunter precalciner kiln. Initially all 280,000 TPA of air cooled slag from Chaparral’s North Texas EAF was utilised. Acceptance has grown and by 2002 approximately 340,000 Tonnes of air cooled slags were consumed by the CemStar process in the USA. This technology provides the opportunity to radically increase the consumption of slags in the cement industry. It eliminates a steel making waste problem and provides cement makers with the benefit of precalcined feed.
While ever iron and steel are being produced slag will be produced as a by-product. Approximately 0.23 Tonnes of blast furnace slag are produced for every tonne of iron and 0.1-0.2 Tonnes of steel slag are produced for every tonne of steel. Minor efficiency improvements can be achieved around these figures but dramatic reductions cannot be achieved. The IISI recently announced that global steel production exceeded 900 million tonnes in 2002. Theoretically this would have resulted in generation of approximately 300-400 million tonnes of slag globally however the exact figure is unknown. Slag is often viewed as a waste product with associated disposal costs. Some processes have been developed to produce value added by-products. One of the major uses of slag is as a landfill or for reclaiming land.
Air cooled slags have been used for aggregate, road base, railway ballast, etc. Predominantly Blast Furnace slag mixes are most commonly used for this purpose as they have good stability. Steel slags contain Magnesium oxides and free lime (CaO), which are reactive and can cause volumetric instability1. The instability can largely be overcome by aging the slag stockpile or acidic treatment3. The use of slags for these purposes is analogous to the operation of a quarry. The value of the slag products is comparative to and must economically compete against similar virgin quarry materials. In some areas this results in the slags being disposed free of charge or the steel producer paying a fee for disposal. Operations that attempt to generate revenue from slag often create stockpiles of slag, as demand for the product is highly cyclical. Slag supply often exceeds demand.
Typical Slag Dump Many stockpiles are created and remain even after cessation of steel making activities. Stockpiling of slags requires enormous land areas and can create environmental issues. The principal problem is dust generation in urban areas. Complete suppression of dust on stockpiles in the order of One Million Tonnes plus is almost impossible. To overcome these problems considerable work has been invested in identifying and utilising technologies that convert slags to useful value added by-products. It is common in the Japanese steel industry for steel furnace slags to be reintroduced to the blast furnace to allow recovery of its high residual iron content. This process step requires careful and costly process monitoring and additional energy input. Iron recovery is not always possible due to chemistry requirements or difficulties in fines handling. Plants such as EAF mini-mills usually cannot economically return their slag to the iron producer. Another technique is the expansion or granulation of blast furnace slags. Expansion or granulation has traditionally been performed by passing the molten slag stream through jets of water. Expansion increases porosity and lowers density making the slag suitable for several uses such as lightweight aggregate and building products. Supply and demand is again a problem with granulated slags. A constant supply is assured from the blast furnace however demand is highly cyclical depending upon the construction industry. Granulated slags cannot be reliably stored for more than about 3-6 months as they will cement together.
The intense water quench and rapid temperature change associated with granulation locks the mineralogy in a state that provides the glassy granulate with cementitious properties. An activator such as lime (CaO) or Gypsum (CaSO4) must be added to make use of these properties. Some recent developments have trialled the use of dry granulation equipment. This overcomes the need for a large supply of fresh water for quench (usually >>1kL/T- losses typically exceed 0.6kL/T) and reduces the gaseous emissions. Granulation equipment is typically expensive (Approximately A$5-10M) and can be costly to operate. The granulated product is however a value added material. It is used by the cement industry for aggregate or the manufacture of GGBF blended cements. The granulated slag must be extremely finely ground for use in cements if it is not to have a deleterious effect on cement strength4. Slag has much lower grindability than clinker4. This has the effect of reducing mill throughput at a cement plant so separate slag grinding and cement blending facilities are often constructed. It is possible to make cementitious materials entirely from steel slag and granulated blast furnace slag2. This process however requires a significantly higher capital investment (Approx A$20M-$35M) and the exercise of great care with control of the product chemistry. The market for these products is also limited as only a few product grades are possible. The use of slags for their cementitious properties usually involves the replacement of Ordinary Portland Cement (OPC). Replacement values of up to 80% have been utilised however it is generally accepted that OPC replacement is in the order of 0-50%. When utilised for this purpose slags must compete directly with the more prolific flyash from the burning of coal- usually in thermal power stations. Flyash is a finely divided material that requires little or no grinding and has similar hydraulic properties to GGBF slag. Disposal of flyash is an enormous problem in the power industry. It is quite usual for electrical utilities to develop take or pay flyash contracts with cement or concrete producers.
It has long been recognized that iron and steel slags have similar chemical analysis and mineralogy to cement clinker. Mineral Blast Furnace Slag % BOF/BOS Slag % EAF Slag % Typical Type I Clinker % SiO2 35-39 11-18 8-18 18-25 Al2O3 8-12 1-4 3-10 3-8 FeTOT <1 14-19 20-30 2-6 CaO 36-42 48-54 25-35 60-67 MgO 4-12 1-4 3-9 0-6 SO3 2-3 <0.5 <0.5 1.5-4 P2O5 0.00 0.00 0.00 0.21 TiO2 0.00 0.00 0.00 0.21 Na2O 0.32 0.12 <0.1 0.19 K2O 0.57 0.02 <0.1 0.50 Mineralogy C3S (3CaO•SiO2) 40-70 C2S (2CaO•SiO2) 40 20-40 C3A (3CaO•Al2O3) 7 3-17 C4AF (4CaO•Al2O3•Fe2O3) 34 5-15 It is this similarity that has allowed slags to find uses in the cement industry. Primarily granulated blast furnace slag has been added to and inter-ground in the finish mill with the clinker. Depending on the quantity of GGBF slag added a limited range of different products can be produced. This practice relies on the inherent hydraulic properties of the granulate. The minerals particularly required in cement production are high levels of CaO, Al2O3 and SiO2. There are also several other mineralisers and fluxes required. In particular the presence of iron as FeO and Fe2O3 can be used as a substitute for the addition of iron ore. For the addition of slag into the cement kiln granulation is not necessary. The slag can be supplied into the raw mix as an air cooled and crushed material. The crushed and screened slag is carefully blended with the conventional raw materials of cement manufacture- Limestone, Cement Rock, Shale or Clay and Sand. This blending step is determined by the required chemistry of the clinker product. The amount of slag blended into the mix will vary depending on the slag chemistry and the chemistry of the other raw materials. The blended mix is introduced to the raw mill for grinding to typically –0.2 mm particle size. The major component of clinker production is cement rock or limestone. The raw mill is optimised for this material and is usually a vertical roller mill. This type of mill has very limited capacity for processing materials with very low grindability such as iron and steel slags. Typically up to 4% slag can be added to these mills. Addition rates above this amount will likely affect the mills capacity to process Limestone or Cement Rock. Upon completion of grinding the raw meal is further mixed in a blend silo. This averages the chemistry of the supplied components. There is some likelihood that ground slag may form discrete strata in the silo due to its density difference to the other raw feed materials. Depending on the kiln process the feed is passed from the silo into the preheater, precalciner or slurrying sections and thence to the kiln where the clinker is formed.
TXI pioneered and patented the process whereby crushed air cooled blast furnace and steel slags can be added directly to the cement kiln5. This has many benefits over addition of slag to the raw mill or addition of granulated slag to the finish mill. The grinding of slag in a vertical roller mill is energy intensive and extremely abrasive. Intergrinding slag in a raw mill has an adverse effect on mill life and capacity as well as grinding media and costs. Grinding slag in the raw mill also severely limits the amount of slag that can be added to the kiln. TXI identified through STAR an opportunity to utilise slag in their kilns. Initially the project was based in Midlothian, Texas as both a cement kiln and an EAF are co-located on the same development. This ensured the efficient and economical transportation of slag. TXI undertook a series of laboratory tests to identify if lumps of material could bypass the early process steps and be introduced directly to the kiln. Eventually a method was determined whereby coarse lumps of slag up to 20mm in size could be fed directly to the cement kiln. This enabled TXI to add the crushed air cooled slag directly to the cement kiln after the preheater or precalciner or into the wet slurry on a Long Wet kiln. Traditionally the addition of lump material into a cement kiln has caused many processing problems among which are refractory damage and poor clinker formation. The relatively low melting point of iron and steel slags overcome these issues. Iron and steel slags melt at approximately 1400oC and 1300oC respectively. This is below the typical cement clinker formation temperature of 1450oC.
In melting the slag becomes highly reactive. Care must be exercised to prevent significant damage to refractory materials. The reactivity ensures that the molten slag rapidly combines with the other raw materials in the mix and acts as a flux to drive the clinkerisation reaction forward. A further benefit of adding crushed air cooled slag to the cement kiln is that the slag has been calcined in the blast furnace or steel furnace. The benefit of this is that specific CO2 emission of the cement kiln falls when slag is added7. The cement industry is a large generator of green house gases as the calcination of limestone to lime releases approximately 0.8 tonne of CO2 for every tonne of limestone. CaCO3 > CaO + CO2^ Heat
Calcination of Limestone Air cooled slag is mineralogically similar to clinker. Slag is virtually a proto-clinker with many of the essential Calcium-Silicate matrices formed. The effect of this is that slag added to the cement kiln requires very little energy for clinker formation. The use of slag in the cement kiln ensures that practically all the energy embodied into the slag in the steelmaking process is utilised and does not go to waste. The slag for the CemStarSM process typically needs only be crushed to 100% passing a 20mm screen. This is usually performed as part of the iron unit recovery process. Modifications need to be made to the cement kiln to allow for the addition of the extra raw material stream. Advanced control systems must be implemented to meter the slag into the kiln. Slight deviations on the amount of the addition can have serious consequences for the product or the kiln. The process was implemented full time at each of TXI’s US facilities: TXI’s Hunter kilns in 1995. The Midlothian facility in 1996 and in 1998 it was adopted onto TXI’s 7 Californian long dry kilns6. Since then the technology has been broadly marketed and adopted in the USA. To date on the 15 kilns using CemStarSM technology in the USA the maximum addition rate of slag practiced has resulted in a 15% increase in production. It is theoretically possible to increase this limit however higher limits have not yet been trialled. Generally slag addition rates average 6%. The US geological survey estimates global cement production at 1,643 million tonnes for the year 2000. From these figures it is clear that CemStarSM technology has the capacity to utilise approximately 100 million tonnes of slag which is about a third of the worlds production of iron and steel slags. The cement produced as a result of the CemStar process can still be blended with GGBF slag or flyash to produce blended cements. The blendability of OPC produced by CemStarSM is not reduced. A plant using GGBF slag before the introduction of CemStarSM will use the same and possibly more GGBF slag with its CemStarSM product. This technology allows the cement industry to increase its nett consumption of slag. The addition rate of slag to the raw meal is governed mainly by the chemistry of the raw materials, chemistry of the slag and required chemistry of the clinker.
Essentially target clinker chemistry is established and the quantity of the different raw feed streams is adjusted to meet this goal. The kiln itself is merely a vessel for containing the material and adding the heat required to drive the chemical reactions. It has been found that slags from different steel mills can have vastly different addition rates depending on the cement kiln chemistry. Ideally a slag will have high CaO, MnO3 and modest FeO, Al2O3 and SiO2. A good slag will also have low MgO and alkali content. The ideal slag will also have minimal variation in chemistry from batch to batch. This is not easily acheivable as the aim of steel making is to optimise the chemistry of the iron and steel by allowing the slag composition to vary. It is not possible to concurrently produce quality steel and consistent slag. Slag by definition is a by-product. At present production of iron and steel slags exceeds demand. This results in slag being landfilled or stockpiled. CemStarSM has created a demand for appropriate slags amongst cement producers. Slag has even been mined from the abandoned dump of the defunct Kaiser Steel plant in Fontana, USA. Slag has a different value for each kiln in which it is utilised. The value is driven by the benefit that the kiln derives from the technology and the cost of implementing the technology. In most of the installations to date slag has been acquired in the local area. This ensures that the freight costs are kept to a minimum. Cement kilns are usually sited very close to the limestone quarry to minimise transport costs (approximately 2T of limestone is quarried to produce 1T of clinker). Geographically limestones have an inherent logistical economic advantage over slags. Occasionally by fortunate providence the cement kiln and steel mill are co-located. This is the case with the TXI Chaparral steel mill and TXI cement kiln in Midlothian Recent work however has shown that there are several instances where it is warranted to move the slag internationally. This is being driven either by the poor match of the locally available slags to particular kilns or limited access to slag in the local area. When reviewing slag utilisation opportunities cement kiln operators consider their raw material cost parameters. Cement kilns shall always use their own virgin materials in preference to outside purchased materials. Typically cement kilns will only utilise slags to enhance their production rates while ever excess demand exists. It is likely that capacity expansions in the cement industry may slow over the next few years as cementitious by-products find larger acceptance. Kilns operating on the CemStarSM process are currently sourcing slags from US based suppliers such as Levy Company, Olympic Mill Services, Heckett Multiserv and International Mill Service.