TSpace Research Repository tspace.library.utoronto.

ca

Energy recovery from high temperature slags

M. Barati, S. Esfahani and T.A. Utigard
Version Post-print/Accepted Manuscript

Citation Barati, M., S. Esfahani, and T. A. Utigard. "Energy recovery from high
(published version) temperature slags." Energy 36.9 (2011): 5440-5449. DOI:
10.1016/j.energy.2011.07.007

Copyright/License This work is licensed under the Creative Commons

Attribution-NonCommercial-NoDerivatives 4.0
International License. To view a copy of this license, visit
http://creativecommons.org/licenses/by-nc-nd/4.0/.

How to cite TSpace items

Always cite the published version, so the author(s) will receive recognition through services that track
citation counts, e.g. Scopus. If you need to cite the page number of the author manuscript from TSpace
because you cannot access the published version, then cite the TSpace version in addition to the published
version using the permanent URI (handle) found on the record page.

This article was made openly accessible by U of T Faculty.

Please tell us how this access benefits you. Your story matters.
 Energy Recovery from High Temperature Slags
M. Barati*, S. Esfahani, T.A. Utigard

*) Corresponding author
Assistant professor
University of Toronto, Dept. of Materials Science and Engineering
140 – 184 College Street, Toronto, ON
Canada M5S 3E4
Tel: (416) 978 – 5637
Fax: (416) 978 – 4155
Email: [email protected]

Abstract

Molten slags represent one of the largest untapped energy sources in metal
manufacturing operations. The waste heat of slags amounting to ~ 220 TWh/year at
temperatures in the range of 1200–1600 °C, presents an opportunity to lower the energy
intensity of metal production. Currently, three types of technologies are under
development for utilizing the thermal energy of slags; recovery as hot air or steam,
conversion to chemical energy as fuel, and thermoelectric power generation. The former
route is most developed with its large scale trials demonstrating recovery efficiencies up
to 65%. The latter two are emerging as the next generation methods of waste heat
recovery. An evaluation of these methods shows that for both thermal and chemical
energy recovery routes, a two–step process would yield a high efficiency with minimal
technical risk. For thermoelectric power generation, the use of phase change materials
appears to solve some of the current challenges including the mismatch between the slag
temperature and operating range of thermoelectric materials.

Key words: slag thermal energy, waste heat, slag granulation, thermoelectric, methane
reforming, coal gasification

1. INTRODUCITION
Metal manufacturing industry has achieved tremendous improvements in its energy efficiency in
the past several decades. The specific energy consumption of steel in the U.S., for example, has
decreased from 48 to 20 GJ/ton in the period 1960–2000[1]. This accomplishment has been realized by
implementing numerous technological advancements such as introduction of continuous and chained
operations where the regular cooling–heating cycles between the processing steps are eliminated. In a
study, Fruehan et al.[2] estimated the potential energy saving in steel manufacturing and showed that
there is an opportunity to further reduce the energy consumption involved in making liquid steel by

1
about 20–30%. With established technologies to recover the thermal and chemical energies of process
off–gas, the waste heat of slags represents the last untapped source that may be used for energy
conservation in the rather energy–intensive metals industry.

Metallurgical slags constitute the largest by–product of the high temperature operations
involved in the extraction and refining of metals. Slags are comparable to molten lava and are generally
rich in silica (SiO2), alumina (Al2O3), and lime (CaO). Slag is formed from the refining reactions,
remaining gangue of the ore, the erosion of the furnace refractory, and the added fluxes. The molten slag
is tapped at temperatures up to 1650 °C, carrying a substantial amount of high quality thermal energy.
This energy is usually not recovered, as the slag is tapped and cooled slowly in the slag dumps or is
rapidly quenched by water to make glassy granules that are used as feedstock for cement manufacturing.
Over the past four decades, several processes have been proposed to recover the waste energy of slag as
heat, electricity, and fuel, while none has been commercialized yet. A renewed attempt at evaluation of
the proposed processes is critical, as the metals industry is striving for another major step in improving
its energy efficiency. This article provides an inclusive figure for the energy content of metallurgical
slags, and an overview of the technologies that have been proposed for recovery of that energy. An
evaluation of the methods will be made to suggest the most promising routes towards utilizing the waste
heat of slag.

2. METALLURGICAL SLAGS AND THEIR ENERGY VALUE

2.1. Slag
Metallurgical slags are conventionally divided into ferrous and non–ferrous, based on the metal
produced in each process. The typical chemical composition and generation rate of various slags for the
major metal manufacturing processes are provided in Tables 1 and 2. As seen, the relative amount of
slag produced per ton of iron/steel is lower than the non–ferrous operations. However, the order of
magnitude larger metal production capacity in iron and steel plants gives rise to a substantially larger
slag volume for individual plants.

<<Table 1>>
<<Table 2>>
Traditionally, molten slag is poured into slag pits and is allowed to cool gradually. This slag is
land–filled or crushed and used as a base material for road construction. The ferrous slags, particularly
those of blast furnace (BF) are also increasingly used as a cementitious material due to their high content
of calcium silicates[13], although the use of non–ferrous slags for this purpose has also been
proposed[14]. It has been shown that substituting earth minerals with slag could give rise to nearly 50
percent reduction in the CO2 emissions associated with production of Portland cement[15]. For this
application, the slag is converted into granules that are subsequently ground and charged to the cement
kilns. The slag used as a cement feedstock needs to be in amorphous form (i.e. glassy) to avoid
subsequent swelling of the concrete. Consequently, the granulation processes use pressurized water, wet

2
granulation, or large quantities of air, dry granulation, to prevent slow cooling and crystallization of the
slag during solidification. Cooling rates in excess of 10 °C/s [16] are required to form an amorphous slag
with sufficiently strong hydaulicity.

Despite the effectiveness of wet granulation in producing a glassy slag, the thermal energy
contained in the high temperature slag is not recovered. In addition, the wet process uses a large amount
of water for granulation, typically around 10 tonnes of water of which 1–1.5 tonnes evaporates per tonne
of slag [17-19]. Leaching of alkaline oxides in water and release of H2S gas are other environmental
problems associated with wet granulation. Dry granulation, on the other hand presents an opportunity to
simultaneously generate a glassy slag and recover the sensible heat in the form of hot gases, steam, or
chemical energy. As a result, the attempts to recover slag waste heat have been concentrated on dry
granulation.

The trials for dry granulation of slag date back to 1930[20], but the research surged in 1970s and
1980s in Europe and Japan, resulting in laboratory and pilot scale testing of several methods. In Section
3, the processes that have been proposed for energy recovery from slag based on the use of dry
granulation are described.

2.2. Energy content of slags

The energy consumed in high temperature processing of metals is distributed between metal,
slag, off–gas, and the natural losses to the atmosphere. The slag thermal energy represents about 10–90%
of the output energy depending on the slag/metal ratio and the discharge temperature. For instance, Rodd
et al. [21] concluded that for ferro–nickel production, slag represents approximately 80% of the total
energy input into the electric smelting furnaces. For steel production, on the other hand, the estimated
share of slag energy is 10–15% [22, 23]. Nevertheless, on a world basis the total amount of slag
contained is substantial for all metals (Table 3).

To estimate the energy content of the slags, the generation rates presented above were used
together with FactSage™ thermodynamic calculation of slag enthalpy for typical compositions and
temperatures given previously. The results are summarized in Table 3. The heat value provided in this
table represents the heat released from the slag once cooled from its discharge temperature to the
ambient temperature of 25 °C.

<<Table 3>>

As it is evident from Table 3, ferrous slags account for over 90% of the available energy
associated with slags where BF slag alone constitutes ~ 50% of this energy. It is thus not surprising that
majority of the investigations in this area have focused on BF slags. It should be noted that the values
reported in this table represent the maximum energy content of the slag, when solidified slowly and to
crystalline phase. However, as discussed earlier, a growing number of slag treatment operations consider

3
the subsequent use of slag in cement making, and converting molten slag to a glassy phase. This
approach comes with a disadvantage though, as the heat released from solidification of slag to
amorphous phase is about 17% less than that of a slag forming crystalline structure, Figure 1[26].

<<Figure 1>>
3. ENERGY RECOVERY FROM SLAGS
Attempts to recover energy from metallurgical slags have faced a fundamental constraint that is
the low thermal conductivity of slag, ranging from 1 to 3 W.m–1.K–1 for solid to 0.1–0.3 W.m–1.K–1 for
molten slags at 1400–1500 °C [6, 27, 28]. This explains why the core of a slag pot cooled in air remains
liquid for days. Numerical simulations [10] and experimental measurements [29, 30] of temperature
distribution inside a slag droplet suspended in air point to considerable temperature gradients. As an
example, Figure 2 shows that in the temperature range of 1100–1300 °C, corresponding to the transition
temperature of liquid to glassy slag [27], the difference between core and surface temperatures of a drop
as small as 5 mm, can exceed 200 °C. It is thus apparent that for forming a glassy product and also
efficient heat recovery, the slag must be broken up into small droplets with a large surface available for
heat transfer. Consequently, the earlier attempts [31-37] to recover the heat of slag without granulating it
have been gradually abandoned, as the recovery efficiencies have been low and/or the slag was not
cooled fast enough to become glassy. As a solution, Donald and Pickles [38] have investigated a method
to recover the thermal energy of slags using molten salts. Their experiments involving pouring a stream
of slag into molten salt, measuring the temperature rise of the bath, and relating it to the energy recovery.
They found that efficiencies as high as 66% can be achieved while the slag structure was similar to the
water–quenched samples (i.e. glassy). However, in this direct heat extraction process, dissolution of slag
in the salt and slag–salt separation remain to be addressed. Therefore, practically all active developments
in this area are concentrated on granulation of the slag prior to recovering the energy.

The original investigations on the utilization of slag waste heat were concentrated on thermal
energy recovery, i.e. transferring the heat of slag to hot air, steam or molten salts. Recently, the efforts
have been directed to producing fuel gas, through an endothermic reaction that exploits the slag heat.
Direct generation of electricity from waste heats based on the thermoelectric principle has also been
proposed. In this section, these methods are described.

<<Figure 2>>
3.1. Thermal energy recovery
The differences between various methods are in the way that the slag is fragmented (the use of
centrifugal force or impinging jets), the heat transfer medium (air, steam, or salt), or the heat transfer
mode (i.e. direct–contact in packed/fluidized bed or indirect heat exchange). In this article, the energy
recovery methods are categorized based on the method of slag granulation.

4
Rotary cup atomizer (RCA)
The rotary cup atomizer was first designed in Britain by Pickering et al. [26] and has been
extensively studied since then. It essentially atomizes the molten slag by combined actions of a rotating
cup and air blast, and cools the droplets rapidly to produce a glassy product. The particles are cooled as
they travel through air and are later introduced to two successive fluidized beds for heat recovery. The
suspension of particles with mean diameter of 2 mm in fluidized bed prevents the slag particles from
clustering and at the same time provides rapid cooling to generate slag with over 95% glass content. A
schematic diagram of the RCA is presented in Figure 4. It has been discussed that while the fast rotation
of the RCA at 500–1500 rpm is sufficient to spread the slag into a film and break it up, the annular air jet
supplied around the cup facilitates the formation of smaller and more uniform particles. Based on their
small scale experiments (0.2–0.5 kg slag per second), Pickering et al. proposed a commercial scale
system, that if operated continuously, would yield an energy recovery of 59%. The losses occur because
a) the total latent heat is not released when glassy slag is formed b) the solid slag is discharged from the
heat recovery vessel at 250 °C and c) heat loss in the slag accumulator. On the other hand, the main
advantages of RCA are claimed to be high productivity and controllable slag grain diameter. It has also
been proposed by Akiyama and his co–workers [39-41] that by impinging reactive gas such as mixture
of methane and steam, the sensible heat of the slag can be effectively converted into fuel, as will be
explained later.

Influence of operating conditions such as the cup shape, rotating speed, slag viscosity and gas
flow rate on the grain size was later investigated by Mizuochi et al.. [42, 43] Their results indicated that
the diameter and shape of the slag droplets was strongly dependent on the rotating speed of the cup with
higher rotating speeds resulting in smaller and more uniform spherical particles. In another work by Yu
et al.,[44] it has been found that when exceeding 1000 rpm, the decrease in the granule diameter
becomes less significant and also the temperature of the slag does not have a considerable effect on the
slag granule diameter.

<<Figure 3>>
<<Figure 4>>

The granulation efforts using a rotating cup were continued in 1990s by Siemens VAI (formerly
VAI–UK) and Davy McKee [45-48], although heat recovery was not practiced in these trials.
Commercialization of the technology has not been however realized due to technical challenges such as
formation of slag wool and degradation of the atomizer cup. Recently, rotary cylinder has been tested in
the laboratory scale [49, 50], anticipating a higher efficiency of the granulation energy due to the
reduced slip between slag and metal in this design. The experiments have shown that spherical glassy

5
slag of various sizes can be produced; however, heat recovery from the granulated slag has not been
practiced.

Granulation by spinning disk

The early works involving the use of spinning disks for slag granulation were carried out in
Japan by Sumitomo Metal Industries in 1980s [30], aiming to form a glassy BF slag. Recent
developments in spinning disk granulation include the works of Akiyama and co–workers at Hokkaido
University, Japan, [42, 51, 52], and those of Xie, et al.[18, 19, 53-55] at CSIRO, Australia. The former
group however have proposed the recovery of the slag heat into chemical energy, as will be explained
later.

In this method of granulation, a stream of slag is delivered onto the disk rotating at 1000–3000
rpm. The liquid steam is fragmented due to the impact with the disk and the centrifugal force. The
droplets are scarred radially and collected in the heat recovery chamber where slag is cooled in a
fluidized or a packed bed.

CSIRO has been working on the design, development and scale up of the method since
2002[53]. The early works involved optimization of the granulation method that was later integrated
with heat recovery. Their concept is based on a two–step process; in the first step, molten slag is
granulated using the spinning disk. Through contact with an uprising air flow, the slag droplets freeze
and their temperature drops to about 900 °C, forming an amorphous phase. This slag is then charged to a
packed bed heat exchanger where counter–current flow of secondary air cools the slag to about 50 °C.
The hot air of both steps is at temperatures above 600 °C and may be used to generate steam or be used
for preheating/drying. Figure 5 shows still images of the slag atomization and the final product. As seen,
the granules are comparatively uniform and within the range of 1–1.5 mm. The smaller particle size in
this technology offers several advantages, including faster cooling, higher glassy content, shorter flight
time, and ease of grinding for subsequent sue.

<<Figure 5>>

Their process has been successfully tested on blast furnace slag at rates up to 10 kg/min. Further
developments are underway to first test slag rates up to 100 kg/min and later 1000–2000 kg/min.

Granulation using rotating drum

In the early 80’s, Japanese companies Ishikawajima–Harima Heavy Industries and Sumitomo
Metal developed a process in which a stream of blast furnace slag breaks up as it is impinged onto a
rotating drum [13, 56-58]. The slag particles then fall into a fluidized bed where the heat is recovered.
The process was later modified to lubricate the drum surface with a thin film of water or oil, giving rise
to smaller heat loss and reduction in the particles scattering [30]. Further, pulverized cool slag was

6
injected into the fluidized bed to accelerate the slag cooling, in order to decrease the required flight time
for the droplets and make the plant smaller [29]. It also reduces the agglomeration of slag particles at
high temperature. This process was tested in full scale of 40 t/hr at Wakayama Steel Works [13, 30] with
a reported 50–60 % recovery of the slag sensible heat into hot air. A flowsheet of the process is provided
in Figure 6.

<<Figure 6>>

Another attempt to granulate and recover the heat of BF slag using a twin–drum technique was
made by NKK in the 1980’s [59]. In this method, the slag is poured into the space between two drums. A
film of slag is dragged and frozen on the drums as they rotate outwards. Heat is transferred to a coolant
inside the drums, which is later transferred to a boiler to generate steam. Large scale trials were
conducted at NKK Fukuyama Works, yielding a relatively low 40% heat recovery. The slag glassy
content also reached a maximum of 80%, making the product unsuitable for substituting with cement.

Granulation by solid slag impingement

Between 1979 and 1981, Swedish group Merotec (a subsidiary of the State Steel Company),
developed a process for dry slag granulation based on mechanical disintegration of a liquid slag film by
impinging particles [60, 61]. The method involves striking the slag stream with previously solidified slag
particles, to break up the slag into granules that together with the recycled particles are charged to a
multi–stage fluidized bed, Figure 7. The slag temperature is adjusted to 500–800 °C in the fluidized bed
by controlling the ratio of recycled to liquid slag. The heat is primarily (65%) recovered into high
pressure steam while it is claimed that about 10% additional heat may be recovered from the fluidizing
hot air, that is otherwise exhausted[61]. The slag granulates are typically below 6 mm where the –3 mm
particles are recycled to the granulator.

<<Figure 7>>

Granulation by air blasts

Air blast granulation technique was first developed by Mitsubishi Heavy Industries and Nippon
Kokan KK (now JFE) [59, 62-64]. The early works began in 1977 leading to pilot testing in the mid-80s,
and operating at commercial scale of 80 t/hr until early 90s[21]. In this method, the slag is first pre-
treated to optimize the flow and to adjust the viscosity. It is then poured to a gutter under which the main
air nozzle breaks up the stream of slag spread to a falling film (Figure 8). Other nozzles are also used to
avoid scattering upward and sideward. The air velocity, flow rate, position of the gutter, and the nozzle
size are carefully controlled to achieve acceptable size distribution of the droplets. The heat of the flying
droplets is partially extracted to a waste heat boiler tubes suspended downwards by radiation and
convention. The remainder of the heat is exchanged through a second boiler located at the bottom of the
unit and embedded in the slag granules. The steam from the two boilers is collected into a steam drum

7
while the slag granules (< 3mm) are discharged at temperatures around 200–300 °C. A design for air
blast granulation of BOF slag in the Mitsubishi Heavy Industries which discharges slag from BOF at
intervals of 30 min has included slag granulating capacity of maximum 80t/h with steam generation rate
of 15 t/h. An energy balance of the system has shown that ~41% of the heat is recovered in the steam,
another 39% is carried by the hot air exhausted at 500 °C and could potentially be recovered [64]. The
reasons for ceasing the operation of this process are not clear, although low recovery rate, scattering of
the droplets, and energy intensiveness of the granulation process have been considered as the
shortcomings [18].

<<Figure 8>>

In an experimental study by Liu et al. [16], the efficiency of an embedded boiler has been
measured by flowing a bed of slag down between boiler tubes. According to them, under optimum
conditions, heat recovery can reach a substantial 90%. However, oven–heated slag granules where used
in their work and the overall efficiency of a granulation–heat recovery system is not clear.

3.2. Recovery as chemical energy

Generation of fuels based on the utilization of the slag thermal energy in endothermic reactions
has been recently studied. The following reactions have been proposed [39] to generate CO, H 2, or their
mixtures through gasification of coal, Reactions (1)–(2), reforming of hydrocarbons, Reactions (3)–(5),
or decomposition of methanol, Reaction (6). In addition, the use of hot slag as a medium for heat storage
and carrier has been proposed [65, 66]. Through this method, the thermal energy enables such
endothermic reactions as phase change (such as melting of material) or other heat consuming reactions
(for example decomposition of limestone, 𝐶𝑎𝐶𝑂3 → 𝐶𝑎𝑂 + 𝐶𝑂2 , or melting), that could be later
reversed to release heat.

<<Table 4>>

A thermodynamic analysis [39] of the reactions has shown that decomposition of limestone,
reforming of methane, and gasification of carbon are the most promising heat absorbing reactions, as
they present the least exergy loss among all the reactions studied. The latter two processes have received
more attention due to their potential to generate fuel, which is easy to transport and be used for various
applications.

Energy recovery by methane reforming

In 1997 Kasai et al. [41] proposed the use of hot BF slag for promoting water reforming of
methane (i.e. Reaction (4)). Figure 9 depicts the scheme of this concept. As shown, part of the slag
sensible heat is transferred to the reactants (methane and steam), generating CO and H2. The still molten

8
slag (~ 1250 °C) is then transferred to the granulation step, in which steam is generated by extracting the
remainder of the sensible and latent heat. The steam is fed back to the reformer, thereby increasing the
overall efficiency of the process. The sensible heat of the high temperature reducing gases is recovered
through a heat exchanger that produces steam for power generation. The chemical energy of the gases is
partially recovered in the methanation reactor where the reverse reaction generates methane and steam.
Water vapour is condensed in a second heat exchanger and methane is directed back to the reformer. The
feasibility of the reforming process using hot slag surface [41] and the effect of slag composition on the
reaction rate [40] were examined in two separate studies. It was shown that larger CaO content in the
slag gives rise to higher reaction rates while FeO and sulphur have inhibiting effects. The apparent rate
of reforming was found to be 10–3–10–1 mol.cm–2.s–1. The experimental works provided limited insight
into the energy recovery efficiency, as they involved only examination of the reaction rate over a small
slag surface at a fixed temperature. The heat recovery steps were not included, thus the overall energy
recovery efficiency is not known.

<<Figure 9>>

Maruoka et al. [67] designed a new system in which the slag is first granulated using a rotary
cup and is then accumulated in a packed bed. The steam reforming of methane takes place with the aid
of a Ni–based catalyst. The thermal energy required for the reaction is recovered by direct heat exchange
between the gas mixture and slag granules. In their materials and energy balance analysis, they assumed
that 299 kg of blast furnace slag is generated at 1250 °C per tonne of crude steel. This slag, if cooled to
150 °C through the proposed system, can generate as much as 106 Nm3 of CO+H2 (molar ratio of H2:CO
= 3:1) exiting the reactor at 1100 °C. This would represent chemical energy recovery of 0.24 GJ/ton
crude steel (0.81 GJ/ton slag) or 51% of the slag thermal energy. The gas also contains an additional
thermal energy of 0.155 GJ/ton steel (0.52 GJ/ton slag) that if recovered fully, would increase the total
recovery to approximately 83%. Although this proposed system appears promising, it remains to be
tested. In addition, technical issues such as clustering of slag granules at temperatures as high as 1100
°C, as well as the structure of the granulated slag for subsequent use should be investigated.

<<Figure 10>>

In a more recent experimental work, reforming of methane by CO 2 has been investigated by

Purwanto and Akiyama [68]. They purged a mixture of CO 2–CH4 through a packed bed of slag granules
(~ 2 mm diameter) at a temperature range of 700–1000 °C. Analysis of the exhaust revealed that
increasing the temperature has a substantial effect on the methane conversion extent, which reached a
maximum of 96%. Again, since the focus of this research was only on the conversion kinetics, no
information may be obtained regarding the energy recovery efficiency.

9
Energy recovery by coal gasification
The only documented study on recovering slag heat through coal gasification is a materials–
energy balance analysis by Li et al. [69] in China. In the proposed coal gasification system, CO2 is
injected together with coal into a molten BF slag bath, generating carbon monoxide through Reaction
(1). The sensible heat of off–gas is recovered in a heat exchanger to make steam while the cleaned gas is
used as fuel. According to this analysis, a steel mill generating 10 million tonnes of steel would also
produce 3 Mt slag, from which 0.132 Mt tonnes of CO is generated. Based on an energy content of 1.6
GJ/ton slag, this represents thermal–to–chemical conversion efficiency of 35%. However, the total
energy recovery is greater, because a non–quantified amount of heat is recovered as steam.

3.3. Direct electricity generation

Direct conversion of waste heat to thermoelectric power has been recently reviewed by Rowe
[70]. The method presents an environmentally friendly, safe and reliable technology which can convert
unused heat into electricity. With the emergence of modern semiconductors with Seebeck coefficient in
the order of several hundred microvolts per degree, the technology appears promising, particularly for
high quality heat such as that of slag. Low temperature trials have been successfully carried out on a
laboratory scale and in prototype commercial systems. As an example, an investigation on steel plant
heat recovery has been performed where large amounts of cooling water are discharged at constant
temperatures of around 90 °C. It has been estimated that total electrical power of around 8 MW would be
produced employing currently available modules fabricated using bismuth telluride material technology.

The thermoelectric technology is an emerging area, particularly for high temperature

applications. The choice of thermoelectric materials with large Seebeck coefficient for high temperature
systems is rather limited. However, the slag waste heat may be stored in phase change materials as
proposed by Nomura et al.[65], in the temperature range of ~150–1000 °C, coinciding well with the
operating temperature for a range of thermoelectric materials that may suitably be used for energy
conversion.

4. EVALUATION OF ENERGY RECOVERY METHODS.

Among the various routes for energy recovery from high temperature slags, the thermal methods
have been developed most, and are gradually reaching commercial testing and acceptance. Dry
granulation followed by recovery of heat into steam or hot gases (air) has been practiced in several pilot
and plant trials, although none has been established for long term commercial use.

Table 5 presents a summary of the basic parameters for the thermal recovery methods. As can
be seen, the heat recovery efficiency for these processes is within the range of 40–65%.

<<Table 5>>

10
 The requirement for forming a glassy product while recovering the heat poses a fundamental
challenge in efficient heat recovery. While large quantities of air are required for rapid quenching of
solidifying slag, the temperature (and heat quality) of the heat carrier, steam/air, is reduced at higher
flow rate. The lower gas temperature in turn reduces the recovery of energy as steam or heat source in
the subsequent steps, leading to lower overall efficiency. As a result, the trend in the slag treatment
processes is towards controlled granulation so that small and uniform slag grains are generated. The
larger contact area promotes heat transfer for a given air flow rate, and also facilitates fluidization of the
grains, as well as their final use as cement feedstock. In addition, Bisio [71] has shown that where the
use for hot air exists, such as blast furnace, generation of hot air is more energy efficient than steam.

The recovery of heat from dry–granulated slag grains conventionally takes place in fluidized
bed for improved heat transfer and also preventing agglomeration of the particles. However, the
disadvantage of fluid–bed is that slag is discharged at a temperature typically above 200 °C, causing a
considerable energy loss. An optimum design may include a two–step heat recovery, in which initially
the slag is rapidly cooled in a fluidized bed below the clustering temperature (800–900 °C).
Subsequently, the remainder of the heat is recovered in a packed bed, through embedded boiler tubes or
direct contact of air–slag. Within this context, the proposed process by CSIRO [18, 53, 55] process
appears very promising, as the slag discharge temperature (50 °C) is considerably lower than the
previous methods (~200–300 °C). Such difference in the slag temperature could result in increasing the
recovery efficiency by additional 10–15% compared to those presented in Table 5.

Conversion of slag waste heat into chemical energy offers several advantages over the thermal
technologies:

a) The fuel may be transported over long distances without a loss in the energy content.

b) The metallurgical industry may be linked to the chemical and energy sectors.

c) Fuels have a substantially larger energy density compared to steam or hot air, making
their transportation and storage more economically attractive. As an example, the energy
density of saturated steam at 250 °C (40 bar) is 44 MJ/m3, whereas that of CO–H2 gas
mixtures generated through reactions (1)–(6) and compressed to the same pressure is an
order of magnitude larger (450–500 MJ/m3).

d) The energy of the fuels may be considered of high quality, as the fuels may be
combusted to generate temperatures exceeding 2000 °C, if necessary, whereas the use
temperature for air/steam is limited by their discharge temperature (typically 600 °C or
below).

11
 Despite these advantages, the recovery of the slag waste heat into chemical energy is still in its
early development stages and has not developed beyond laboratory testing of only reforming step. There
is no data on the integrated system consisting of combined heat recovery–reforming, thus a comparison
with the thermal energy recovery methods is not possible. A major challenge in effective conversion of
thermal to chemical energy appears to be the low volume of gas that is generated compared to the slag
mass. As an example, a thermodynamic evaluation shows that at 60% heat recovery, each tonne of gas
mixture (stoichiometric mixture of methane at 25 °C and steam at 200 °C) would require about 10 tonnes
of slag to reach the required reforming temperature of 900 °C. This amount of gas is clearly not
sufficient for quenching the slag fast enough so that a glassy phase is formed. In addition, the long
contact time required for such amount of heat exchange would result is a prohibitively large reactor. As a
result, it is apparent that the process must consist of two steps. In the first step, the slag is granulated and
cooled rapidly to a temperature below its crystallization temperature (~ 1000–1100 °C [72]) using air.
The slag is then charged to a second vessel, where the remaining heat is transferred to the reactants (e.g.
methane and steam/CO2). The heat recovered in the first unit may be utilized to generate the steam
needed for reforming. A schematic diagram of this concept is shown in Figure 11.

<<Figure 11>>

Thermoelectric power generation from slag waste heat is a very attractive alternative to thermal
and chemical energy forms, although this technology is still in its infancy. Currently there are several
challenges against direct conversion of slag heat to electricity, (a) the operating temperature of the
existing thermoelectric materials is considerably lower than the molten slag, (b) high melting point of
slags limits the temperature range at which the heat can be extracted effectively, while molten, and (c)
the slags are extremely corrosive, thus interfacing slag and thermoelectric materials would pose technical
difficulties. For these reasons, it appears that a potential thermoelectric technology can benefit from
PCMs as heat carriers. Recovering the heat of high temperature slags, the PCMs then provide a
continuous supply of thermal energy to the thermoelectric system at lower temperatures. The type of
PCM and its operating temperature range should be matched to the type of the thermoelectric material
selected (typically below 1000 °C). Extensive research is required to answer such question as coupling
of the intermittent slag supply to the power system, conversion efficiency, and cost.

5. SUMMARY AND CONCLUSIONS.

Driven by the energy crisis of 1970s, numerous investigations were made in 1970–1980s to
develop a method of heat recovery from slag, while generating a slag suitable for other applications.
These attempts however diminished with the ease of the energy prices in the 1990s. A renewed interest
in this area has resulted in experimental and pilot trials of several technologies, although none has yet

12
been established industrially. The following are the major findings from a thorough review of the past
and state–of–the–art in waste heat recovery from slags.

1. The world production of slag represents a heat content of 220 TWh per year (2008), with
blast furnace slag accounting for about 50% of this energy.

2. Driven by a demand to produce amorphous slag as a substitute to cement and limited by

the low thermal conductivity of slag, the majority of methods for energy recovery from
slag are relying on dry granulation of slag.

3. Being investigated for over three decades, the thermal energy recovery methods are most
developed among the various schemes proposed for energy recovery from slags.

4. Compared to the rotary drum or the air jet granulation, the techniques employing
spinning disks or rotary cups offer the advantage of generating uniform and small grains
at a lower energy consumption rate. Smaller droplets are quenched faster and require less
coolant, resulting in recovery of more heat and at higher temperature.

5. Because of the cooling rate requirements, a two–step heat recovery consisting of a

fluidized bed, followed by a packed bed would yield higher energy recovery efficiency,
as both high outlet gas temperature and low slag discharge temperature are achieved.

6. The chemical energy recovery methods offer the advantage of high energy density and
zero loss in the energy content during transportation. These methods are under
investigation, with methane reforming and coal gasification receiving more attention as
the feasible routes.

7. An optimum chemical energy recovery system consists of two vessels, where in one the
slag is cooled to bellow the crystallization temperature, and in the other fuel generation
takes place.

8. Thermoelectric power generation based on the slag waste heat appears to be the next
emerging technology in this area. The use of PCMs as heat carriers can address the major
challenges that the technology is currently facing.

ACKNOWLEDGEMENTS
The authors wish to thank Ontario Government, Department of Materials Science and
Engineering (University of Toronto), and NSERC for financially supporting this study.

REFERENCES

13
[1] Stubbles JR. Carbon Dioxide Emissions from the U.S. Steel Industry. In: Pickles CA, editor.
Conference of Metallurgists, Greenhouse Gases in the Metallurgical Industries: Policies, Abatement and
Treatment. Toronto, Canada: MeSoc; 2001. p. 103-16.
[2] Fruehan RJ, Fortini O, Paxton HW, Brindle R. Theoretical Minimum Energies To Produce Steel for
Selected Conditions Washington DC. : U.S. Department of Energy; 2000.
[3] Biswas SK. Principles of Blast Furnace Ironmaking. Brisbane, Australia: Cootha Publishing House,
1981.
[4] Shi C, Krivenko P, Roy D. Alkali-activated Cement and Concretes. London, U.K.: Taylor and
Francis, 2006.
[5] Smirnov LA, Panfilov MI. Metallurgical Slags: Properties and Application. Canadian Metallurgical
Quarterly. 1988;32(2):173-7.
[6] Allibert M, Gaye H, Geiseler J, Janke D, Keene BJ, Kriner D, et al. Slag Atlas. Germany: Verlag
Stahleisen, 1995.
[7] World Steel Association. World Steel Association Yearbook and Factsheets:
http://www.worldsteel.org/Steel. Brussels.
[8] Sudbury MP, Kemp DJ. The Properties and Uses of Granulated Non ferrous Slags. Conference The
Properties and Uses of Granulated Non ferrous Slags, San Antonio, TX, United states vol. 2006.
Minerals, Metals and Materials Society, p. 611-20.
[9] Nell J, 2004. Melting of Platinum Group Metal concentrates in South Africa. The South African
institute of Mining and Metallurgy 2004;104(7):423-8.
[10] Utigard TA, Marin T, Warczok A, Bergman RA, Barati M, Diaz CM. Energy Recovery from High
Temperature Steel, Copper and Nickel Smelter Slags. In: Kapusta J, Davis B, editors. Conference of
Metallurgists (COM2011). Montreal, Canada: 50th Metallurgical Society of CIM; 2011. p. YY-ZZ.
[11] Warner AEM, Diaz CM, Dalvi AD, Mackey PJ, Tarasov AV. JOM World Nonferrous Smelter
Survey, Part III: Nickel: Laterite. JOM. 2006;58(4):11-20.
[12] Warner AEM, Diaz CM, Dalvi AD, Mackey PJ, Tarasov AV, Jones RT. JOM world nonferrous
smelter survey Part IV: Nickel: Sulfide. JOM. 2007;59:58-72.
[13] Nakada T, Nakayama H, Fujii K, Iwahashi T. Heat Recovery in Dry Granulation of Molten Blast
Furnace Slag. Energy Development in Japan. 1983;55:287-309.
[14] Moura WA, Goncalves JP, Lima MBL. Copper slag waste as a supplementary cementing material
to concrete. Journal of Materials Science. 2007;42(Copyright 2007, The Institution of Engineering and
Technology):2226-30.
[15] U.S. Concrete Inc. The Use of Environmentally Friendly Technology in the Concrete Industry,
http://www.us-concrete.com/news/features.asp, Retrieved May 6, 2011. Houston.
[16] Liu J, Yu Q, Dou C, Li R. Experimental Study on Heat Transfer Characteristics of Apparatus for
Recovering the Waste Heat of Blast Furnace Slag In: Jiang Z, Zhang C editors. Conference
Experimental Study on Heat Transfer Characteristics of Apparatus for Recovering the Waste Heat of
Blast Furnace Slag, Zhuhai, China vol. 97-101. Trans Tech Publications, p. 2343-6.
[17] Lee AR. Blast Furnace and Steel Slag. London, U.K.: Edward Arnold, 1974.
[18] Xie D, Jahanshahi S. Waste Heat Recovery from Molten Slags. 4th International Congress on
Science and Technology of Steelmaking. Gifu, Japan: Iron and Steel Institute of Japan; 2008. p. 674-7.
[19] Xie D, Jahanshahi S, Norgate T. Dry Granulation to Provide a Sustainable Option for Slag
Treatment. Sustainable Mining Conference. Kalgoorlie, WA, Australia2010. p. 22-8.
[20] Berger P. Dry Granulation of Molten Blast Furnace Slag in Granulating Mills (Trockene Koernung
von fluessiger Hochofenschlacke in Granuliermuehlen). Stahl und Eisen. 1930;50(51):1775-8.

14
[21] Rodd L, Koehler T, Walker C, Voermann N. Economics of Slag Heat Recovery from Ferronickel
Slags. Sustainability for Profit, Conference of Metallurgists (COM2010). Vancouver, Canada2010. p. 3-
17.
[22] Bengt W. Slag Heat Recovery. International Iron and Steel Institute. 1983:245-56.
[23] Barati M. Energy Intensity and Greenhouse Gases Footprint of Metallurgical Processes: A
Continuous Steelmaking Case Study. Energy. 2010;35(9):3731-7.
[24] U.S. Geological Survey. http://minerals.usgs.gov/minerals/pubs/commodity/. Commodity Statistics
and Information, Accessed Feb 2011.
[25] Niemelä P, Kauppi M. Production, Characteristics and Use of Ferrochromium Slags. The Eleventh
International Ferro Alloys Congress (Infacon XI). New Delhi, India: International Committee on Ferro
Alloys; 2007. p. 171-9.
[26] Pickering SJ, Hay N, Roylance TF, Thomas GH. New Process for Dry Granulation and heat
Recovery From Molten Blast Furnace Slag. Ironmaking and Steelmaking. 1985;12(Compendex):14-21.
[27] K.S. Goto HWN, K.H. Linder. Thermal Conductivities of Blast Furnace Slags and Continuous
Casting Powders in the Temperature Range 100 to 1550 degree C. Stahl u Eisen. 1985;105:1387.
[28] El Gammel T, Jofre SJ, de Loria S. Ermittlung der WarmeleitlUhigkeit in synthetischen
GieBpulvern. Steel Research. 1986;57:620-5.
[29] Yoshinaga M, Fujii K, Shigematsu T, Nakata T. Method of Dry Granulation and Solidification of
Molten Blast Furnace Slag. Tetsu-To-Hagane (Journal of the Iron and Steel Institute of Japan).
1981;67:917-24.
[30] Yoshinaga M, Fujii K, Shigematsu T, Nakata T. Dry Granulation and Solidification of Molten Blast
Furnace Slag Transactions of the Iron and Steel Institute of Japan. 1982;22:823-9.
[31] Okuno R, Fujii H, Toyoda K. Recovery of Heat from Molten Slags from Metallurgical Processes.
U.K. Patent 1,55,4861979.
[32] Nippon Steel Corp. Heat Recovery from Slag. Japan Patent, 81,151,7881981.
[33] Nippon Kokan K.K. Heat Recovery from Molten Slag. Japan Patent, 81,149,5821981.
[34] Nippon Steel Corp. Heat Recovery from Molten Slag. Japan Patent, 8,298,7941981.
[35] Nippon Kokan K.K. Apparatus for Heat Recovery from Slag. Japan Patent, 8,231,7841982.
[36] Nippon Steel Corp. Circulation of Inert Gas in Quenching of Molten Slag. Japan Patent,
5,754,7061982.
[37] Bergkvist B, Koelhi M, Strandell E, Svemar Taeby C. Process and Apparatus for the Recovery of
Heat from Molten Slag. Sweden Patent SE. 448,7391987.
[38] R.Donald J, Pickles CA. Energy recovery from molten ferrous slags using a molten salt medium.
Conference Energy recovery from molten ferrous slags using a molten salt medium, Chicago, IL, USA
vol. 77. Iron & Steel Soc of AIME, p. 681-92.
[39] Akiyama T, Oikawa K, Shimada T, Kasai E, Yagi J-I. Thermodynamic Analysis of
Thermochemical Recovery of High Temperature Wastes. ISIJ International. 2000;40:286-91.
[40] Shimada T, Kochura V, Akiyama T, Kasai E, Yagi J. Effects of Slag Compositions on the Rate of
Methane-steam Reaction. ISIJ International. 2001;41:111-5.
[41] Kasai E, Kitajima T, Akiyama T, Yagi J, Saito F. Rate of Methane-steam Reforming Reaction on
the Surface of Molten BF Slag - For Heat Recovery from Molten Slag by Using a Chemical Reaction.
ISIJ International. 1997;37:1031-6.
[42] Mizuochi T, Akiyama T, Shimada T, Kasai E, Yagi J-I. Feasibility of rotary cup atomizer for slag
granulation. ISIJ International. 2001;41(Compendex):1423-8.

15
[43] Mizuochi T, Yagi J-i, Akiyama T. Granulation of Molten Slag for Heat Recovery. Conference
Granulation of Molten Slag for Heat Recovery, Washington, USA. p. 29-31.
[44] Yu Q-b, Liu J-x, Dou C-x, Hu X-z. Dry Granulation Experiment of Blast Furnace Slag by Rotary
Cup Atomizer. Journal of Northeastern University (Natural Science). 2009;30:1163-5.
[45] Featherstone WB, Macauley D. Slag Granulation. U.S. Patent 5,409,5211995.
[46] Macauley D. Slag Treatment - Time for an Improvement. Steel Times International. 1996;20:S15-6.
[47] Featherstone WB. Slag Treatment Improvement by Dry Granulation. Iron and Steel Engineer.
1998;75(4):42-6.
[48] Featherstone WB, Macauley D. Slag Granulation. U.S. Patent 5,735,9311998.
[49] Kashiwaya Y, In-Nami Y, Akiyama T. Development of a Rotary Cylinder Atomizing Method of
Slag for the Production of Amorphous Slag Particles. ISIJ International. 2010;50(9):1245-51.
[50] Kashiwaya Y, In-Nami Y, Akiyama T. Mechanism of the Formation of Slag Particles by the Rotary
Cylinder Atomization. ISIJ International. 2010;50(9):1252-8.
[51] Akiyama T, Shima T, Kasai T, Yagi J. Feasibility of Waste Heat Recovery form Molten Slag.
China-Japan International Academic Symposium2000. p. 53-65.
[52] Akiyama T, Mizuochi T, Yagi J, Nogami H. Feasibility Study of Hydrogen Generator with Molten
Slag Granulation. Steel Research International. 2004;75(2):122.
[53] Nexhip C, Davidson R, Norgate T, Sanetsis S, Washington B. Dry Granulation of Slags for
Producing Cement Binder. Green Processing. Perth, Australia2004.
[54] Xie D, Washington B, Norgate T, Jahanshahi S. Dry Granulation of Slags - Turning Waste into
Valuable Cement Binder. CAMP-ISIJ. 2005;18:1088-91.
[55] Xie D, Pan Y, Flann R, Washington B, Sanetsis S, Donnelley J, et al. Heat Recovery from Slag
through Dry Dry Granulation. In: Jahanshahi S, Rickards T, editors. 1st CSRP Annual Conference.
Melbourne, Australia: CSIRO Minerals; 2007. p. 29-30.
[56] Sieverding F. Heat Recovery by Dry Granulation of Blast Furnace Slags. Steel Times.
1980;208:469-72.
[57] I.H.I. Ltd. The Dry Granulation and Heat Recovery of the Blast Furnace Slags. 1980 (March). p. 17.
[58] Sumimoto Metals Ind. Treatment of Molten Slags. In: Japan Patent 57, 086, editor.1982.
[59] Yoshida H, Nara Y, Nakatani G, Anazi T, Sato H. The Technology of Slag Heat Recovery at NKK.
SEAISI Conference of Energy Utilization in the Iron and Steel Industry1984.
[60] A New Granulation Technique for Heat Recovery from Molten Slag. Energy Technology: National
Swedish Board for Technical Development; 1979.
[61] Tiberg N. Heat Recovery from Molten Slag, A New Swedish Granulation Technique In: Millhone J,
Wills E editors. Conference Heat Recovery from Molten Slag, A New Swedish Granulation Technique,
Berlin, Germany vol. 2. Springer Verlag, p. 1837-42.
[62] Mitsubishi Heavy Industries Ltd. and Nippon Kokan K.K. Granulating Apparatus for Slag Melts.
In: Japan Patent 5, 915, editor. Japan1982.
[63] Fukuyama Works. Blast Granulation System of BOF Slag and its Products. 1983. p. Nippon Kokan
Technical Report No. 38.
[64] Ando J, Nakahar T, Onous H, Tchimura S, Kondo M. Development of Slag Blast Granulation
Planat Characterized by Innovation of the Slag Treatment Method, Heat Recovery, and Recovery of Slag
as Resources. Mitsubishi Heavy Industries Technical Review1985. p. 136-42.
[65] Nomura T, Okinaka N, Akiyama T. Technology of Latent Heat Storage for High Temperature
Application: A Review. ISIJ International. 2010;50(9):1229-39.

16
[66] Kashiwaya Y, Akiyama T, In-Nami Y. Latent Heat of Amorphous Slags and Their Utilization as a
High Temperature PCM. ISIJ International. 2010;50(9):1259-64.
[67] Maruoka N, Mizuochi T, Purwanto H, Akiyama T. Feasibility Study for Recovering Waste Heat in
the Steelmaking Industry Using a Chemical Recuperator. ISIJ International. 2004;44(2):257-62.
[68] Purwanto H, Akiyama T. Hydrogen Production from Biogas Using Hot Slag. International Journal
of Hydrogen Energy. 2006;31:491-5.
[69] Li P, Qin Q, Yu Q, Du W. Feasibility Study for the System of Coal Gasification by Molten Blast
Furnace Slag. In: Jiang Z, Zhang C, editors. 2009 International Conference on Manufacturing Science
and Engineering, ICMSE 2009, December 26-28, 2009. Zhuhai, China: Trans. Tech. Publications; 2010.
p. 2347-51.
[70] Rowe DM. Thermoelectric Waste Heat Recovery as a Renewable Energy Source. International
Journal of Innovations in Energy Systems and Power. 2006;1(1):13-23.
[71] Bisio G. Energy Recovery from Molten Slag and Exploitation of the Recovered Energy. Energy.
1997;22(5):501-9.
[72] Ryu HG, Zhang ZT, Cho JW, Wen GH, Sridhar S. Crystallization Behaviors of Slags Through a
Heat Flux Simulator. ISIJ International. 2010;50(8):1142-50.

17
List of tables
Table 1- Chemical compositions and production rates of ferrous slags.
Table 2- Typical chemical composition and production rate of non–ferrous slags[8-12]
Table 3- World production of slag and the corresponding energy content [7-12, 24, 25]
Table 4- Endothermic reactions for chemical recovery of waste heats[39]
Table 5- Basic parameters of heat recovery methods based on dry granulation

List of figures
Figure 1- Temperature dependence of heat content for a typical blast-furnace slag [26].
Figure 2- Temperature history of a 5 mm slag drop cooled in fluidized bed[30].
Figure 3- Appearance of slag granulation by RCA with cup rotation speed of 3000 rpm [42].
Figure 4- Schematic diagram of rotary cup atomizer [26]
Figure 5– (a) slag atomization on the spinning disc and (b) granulated slag product. [18, 55]
Figure 6-Process flow diagram of the rotary–drum granulation–heat recovery system [30]
Figure 7- Plant lay–out of Merotec slag granulation process[61]
Figure 8- Schematic diagram of slag granulation by air and heat recovery[21]
Figure 9- Process flow diagram for energy recovery through methane reforming [41]
Figure 10- Steam reforming of methane using granulated slag heat [67]
Figure 11- Proposed process for a two–step chemical energy recovery

18
TABLES
Table 1- Chemical compositions and production rates of ferrous slags.
1
Process Slag composition[3-5] Slag temperature Slag rate
CaO SiO2 Al2O3 MgO MnO FeOx (°C) [6] (t/t. crude steel) [7]
BF 30–56 28–38 8–24 1–18 0.5–2 0.5–1 1400–1450 0.275
BOF 45–60 8-22 1–7 5–17 2–14 10-35 1300 0.126
EAF 35-60 9-20 2-9 5-15 1-8 15-30 1550 0.169
LF 30-60 2-35 5-35 1-10 0-5 0.1-15 1300 0.030
1
BF: blast furnace, BOF: basic oxygen furnace, EAF: electric arc furnace, LF: ladle furnace

Table 2- Typical chemical composition and production rate of non –ferrous slags[8-12]
Metal / Process CaO SiO2 Al2O3 MgO FeO Cr2O3 Slag Slag rate
temperature (t/t. metal)
(°C)
Copper – sulphide ore 2 36 4 1 47 – 1000–1300 2.2
Nickel – sulphide ore 3 36 6 3 46 – 1300–1550 5.3
Nickel (oxide ores)/ FeNi - 54 2 32 11 1 1600 50
Platinum group metals 15 42 5 19 12 2 1350–1600 3700
(PGM)
Zinc 11 24 3 1 47 – 1200–1300 0.01
Lead 20 22 4 1 35 – 1200 0.57

Table 3- W orld production of slag and the corresponding energy content [7 -12, 24, 25]
Estimated slag Slag Heat value
production, 2008 enthalpy (TWh/year)
Metal (process) Million ton (GJ/ton)
Hot metal (BF) 236 1.6 105
Steel (BOF) 112 1.3 40
Steel (EAF) 69 1.8 35
Ferroalloys 58 1.6 26
Copper 34 1.2 11
Nickel 8.4 1.4 3.3
Lead 2.2 1.4 0.9
PGM 1.5 1.5 0.6
Zinc 1.5 1.2 0.5
Total 222
Table 4- Endothermic reactions for chemical recovery of waste heats[ 39]
Reaction Enthalpy Exergy
(kJ/mol) (kJ/mol)
(1) 𝐶 + 𝐶𝑂2 → 2𝐶𝑂 172 122
(2) 𝐶 + 𝐻2 𝑂 → 𝐶𝑂 + 𝐻2 131 91
(3) 𝐶𝐻4 + 𝐶𝑂2 → 2 𝐶𝑂 + 2𝐻2 247 171
(4) 𝐶𝐻4 + 𝐻2 𝑂 → 𝐶𝑂 + 3𝐻2 206 142
(5) 𝐶3 𝐻8 + 3𝐻2 𝑂 → 3𝐶𝑂 + 7𝐻2 498 298
(6) 𝐶𝐻3 𝑂𝐻 → 𝐶𝑂 + 2𝐻2 90 25

19
 Table 5- Basic parameters of heat recovery methods based on dry granulation
Process / Developer Country Status Granulation Recovered Energy Air/steam Ref.
Method energy form efficiency temperature
°C
CSIRO Australia Pilot Spinning Hot air – 600 [18]
disk
British Steel England Pilot RCA Hot air 59 430–520 [26]
Corporation –
University of
Nottingham
Ishikawajima– Japan Plant Single Hot air 50–60 600 [30]
Harima Heavy rotating
Industries & drum
Sumitomo Metal
Merotec Sweden Pilot Slag particle Steam 65 Steam, [61]
impingement 40 bar, 250 C
NKK–Mitsubishi Japan Pilot/Plant Air blast Steam 41 Steam, 250 °C [64]
NKK Japan Plant Twin Steam 40 – [59]
rotating
drums

20
FIGURES

Figure 1- Temperature dependence of heat content for a typical blast -furnace slag [26] .

Figure 2- Temperature histor y of a 5 mm slag drop cooled in fluidized bed[30].

21
 Figure 3- Appearance of slag granulation by RCA with cup rotation speed of 3000 rpm
[42].

hot air
slag supply
cooled launder
wall

slag
granules rotary-cup
air blast
atomizer
primary
fluidized
bed

secondary
fluidized
bed
slag air
discharge

Figure 4- Schematic diagram of rotary cup atomizer [26]

Figure 5– (a) slag atomization on the spinning disc and (b) granulated slag product. [18,
55]

22
 molten slag

granulation

molten granulated slag hot slag granules slag granules

hot air air

crushing

heat exchange

steam sand slag

Figure 6-Process flow diagram of the rotary–drum granulation–heat recovery s ystem [30]

rec
BF slag yc
led screen
sla

slag granules
g
exhaust air
water feed

saturated steam
40 bar, 250 C 0-3 mm >3 mm

granulator exhaust air

bed

heat exchanger

slag granules
fluidizing air

fluidizing air

Figure 7- Plant lay–out of Merotec slag granulation process[61]

23
 saturated steam (or hot water)
(T = 250 C)
steam drum 15 t/hr

granulated
slag
hot air
slag (1600C) 500 C
80 t/hr 3
110,000 Nm /hr
water-cooled walls

boiler
feed water

slag granulator
(high velocity air) slag (~200 C)
Vair ~ 100 m/s oversize
slag
moving bed heat exchanger
(residual heat recovery from slag)

Figure 8- Schematic diagram of slag granulation by air and heat recovery[21]

CH4+H2O = CO+3H2

molten slag molten slag

blast furnace 1523 C 1223 C slag
steam granulation
reformer H2O (g)

heat
steam
exchanger
CH4
(~25 C)
CO+H2 (~25 C)

heat exchanger
methanation
condensed water reaction

hot water /steam CO + 3H2 = CH4+H2O

Figure 9- Process flow diagram for energy recovery through methane reforming [41]

24
 molten slag
CO, H 2

RCA
catalyst

CH4 CH4
H2O H2O

Table feeder
slag storage

Figure 10- Steam reforming of methane using granulated slag heat [67]

molten
slag
hot air

heat
exchanger water
atomizer

exhaust

slag (< 1100 C)

steam
air

methane

Figure 11- Proposed process for a two –step chemical energy recovery

25