European Commission

IarhnIraI sIaaI rasaarrh

Factory-wide control and social and environmental issues
On||ne s|ag ana|ys|s ut|||s|ng contact-free
m|crowave techno|ogy
D. Malmberg
MEFOS
PO Box 812, S-971 25 Lule
C. Grisvard
Arce|or Research
voie Romaine, BP 30320, F-57283 Maizires-ls-Metz
C. Galvo Asceno
ISQ
Apartado 119, P-2781-951 Oeiras
Contract No 7210-PR/302
1 July 2001 to 31 December 2004
F|na| report
Directorate-General for Research
2005 EUR 21787 EN
LEGAL NOTICE
Neither the European Commission nor any person acting on behalf of the Commission
is responsible for the use which might be made of the following information.
A great deal of additional information on the European Union is available on the lnternet.
lt can be accessed through the Europa server (http://europa.eu.intj.

Cataloguing data can be found at the end of this publication.

Luxembourg: Office for Official Publications of the European Communities, 2005

lSBN 92-79-00080-2

European Communities, 2005
Reproduction is authorised provided the source is acknowledged.

P||nted |n /0xemoo0|g

PRlNTED ON WHlTE CHLORlNE-FREE PAPER
Europe D/recI /s a serv/ce Io he/p you f/nd answers
Io your quesI/ons abouI Ihe European Un/on
Freephone number (*j:

(*j Certain mobile telephone operators do not allow access to 00 800 numbers or these calls may be billed.


ABSTRACT
Slags chemical stability due to erosion when deposited, geotechnical properties when used for road
construction, and slag structure have been investigated using cooling rate studies, leaching tests, and
microwave technology. The objective in this project is to simplify slag analysis by utilizing microwave
technology by testing, evaluating and introducing the technology as a fast and reliable contact free
technology for on-line slag analysis.

The expected benefits by introducing a new technology to be used for on-line slag analysis in
metallurgical processes are a faster analysis and a simplified analysis procedure. In its extension this
will result in an improved process control as well as a better yield regarding slag use both for process
control as for road construction material.

The work reported here has been carried out as a multi-partner, multi-national, project by ARCELOR,
ISQ and MEFOS with the latter acting as a project coordinator.

The geotechnical properties were investigated using BOF slag mixed with scrap to obtain three different
cooling options, i.e. three different slag thicknesses. Each mixture was melted in a 6t pilot EAF, and
tapped during controlled cooling rate conditions. To accomplish the aimed slag basicities lime was
added to the melted slag. Two different slag basicities and two slag tapping temperatures were
characterized. Each slag sample was thereafter crushed to comply to those fractions required for road
applications and placed outdoors to age for 7 months with the objective to study changes in the
chemical composition of the slag during this time.

The XRF analysis of the samples gives small variations of the chemical composition of the different
slag samples during the 7 months of ageing period. The variations observed are equivalent to 1
relative %, rather due to the representativeness of the different samples, than the ageing effect. The
same pattern is valid for the iron oxidation evolution.

The investigation shows that an increase of the cooling rate damages the geotechnical properties
especially related to Micro-Deval and Los-Angeles characteristics and an average cooling rate between
15 to 20 C/min during the first 10 minutes following directly after tapping provided a final product
having the properties required for road applications.
Leaching tests with EAF and Ladle slags were performed, using a flow through test (down-flow) and
the standard batch test DIN 38414-S4 (agitated leaching vessel).
The tests were carried out with EAF and ladle slags from the same steel producer. The samples were let
to age by natural weathering at the slag storage facility during six months and the slag composition of
the samples were determined after 3 and 6 months of ageing.
Slag samples submitted to those trials presented very low total leaching levels. The most extracted
elements are calcium and magnesium, probably due to that free calcium and magnesium oxides are not
completely solved into the slag. In flow-through test, leaching levels are below 1.2 g/kg for calcium,
0.06 g/kg for magnesium and 0.6 mg/kg for other elements. Leachates obtained with DIN 38414-S4
present higher leaching values; however, these are lower than 18 g/kg (Ca), 0.06 g/kg (Mg) and 10
mg/kg (other elements).

Along with the leaching investigation also an Environmental Impact Assessment of Steel Slag Reuse
has been performed. In this a state-of-art inventory based on bibliographic references of slags final
destiny as well as environmental impact, processing and further use is presented.

Microwave technology is a versatile technology that can be used in many different applications. Two
applications have been tested and evaluated in previous reported ECSC funded projects. It is also used
on laboratory scale to investigate molecules due to bond length and thus also the molecular structure for
gaseous compounds. In this project microwave technology has been used to investigate solid and liquid
slag, synthetic as well as production slag, due to theirs dielectric properties and relate these to the
chemical structure represented by the slag basicity. As a result of the investigation the dielectric
3


properties, expressed as the refractive index, for different frequency bands for the investigated slags
have been derived. In this respect the refractive index should be considered as a band index rather then
a single frequency index.

In the correlation made between the refractive index and the slag basicity no explicit correlation
between the refractive index and the slag composition, on a statistical significant level, could be
established. To establish such a correlation on a statistical significant level the number of trials must be
increased largely.

The industrial trials performed showed that it is possible to use the technology to measure the slag
thickness in a ladle though the results are very much dependent on the slag condition; rough surface
and/or not dissolved slag formers appearing in the ladle. Therefore, before the technology could be used
on a production scale these problems must be addressed and solved.

4


TABLE OF CONTENTS
Page


ABSTRACT 3
1 OBJECTIVES OF THE PROJECT 7
2 COMPARISON OF INITIALLY PLANNED ACTIVITIES AND WORK
ACCOMPLISHED 7
2.1 MEFOS 7
2.2 ARCELOR 8
2.3 ISQ 8
3 DESCRIPTION OF ACTIVITIES AND DISCUSSION 8
3.1 MEFOS 8
3.1.1 TRIAL PERFORMANCE 8
3.2 ARCELOR 13
3.2.1 SLAG SAMPLES AND TRIAL PERFORMANCE 13
3.2.2 SLAG TEMPERATURE EVOLUTION DURING COOLING 14
3.2.3 SLAG SAMPLES TREATMENT 15
3.2.4 EVOLUTION OF COMPOSITION OF THE SLAG SAMPLES DURING AGEING 15
3.2.5 GEOTECHNICAL TESTS RESULTS 16
3.3 ISQ 16
3.3.1 PHYSICAL-CHEMICAL ANALYSIS OF SLAGS FROM EXPERIMENTAL TRIALS 16
3.3.2 ANALYSIS AND CHARACTERIZATION OF SLAGS FOR RECYCLING PURPOSES 17
3.3.3 FLOW THROUGH TEST RESULTS 20
3.3.4 DIN 38414-S4 DETERMINATION OF LEACHABILITY BY WATER 21
4 CONCLUSIONS 22
5 EXPLOITATION AND IMPACT OF THE RESEARCH RESULTS 23
5.1 MEFOS 23
5.2 ARCELOR 24
5.3 ISQ 24
6 LIST OF REFERENCES 25
APPENDICES: SCIENTIFIC AND TECHNICAL DESCRIPTION OF THE RESULTS 26
LIST OF FIGURES 27
LIST OF TABLES 28




5







1 OBJECTIVES OF THE PROJECT
The objective in this project has been to simplify slag analysis by utilizing microwave technology by
testing, evaluating and introducing the technology as a fast and reliable contact free technology for on-
line slag analysis. In this novel work investigations of the dielectric properties of slag have been
performed on laboratory/pilot scale and correlated against slag composition, slag state, and slag
temperature with the aim to develop procedures for on-line slag analysis on production scale. The way
to achieve this has been:

investigations of the dielectric properties of different slags at different states (phases), as a function
of frequency and polarisation
correlation of these properties against slag temperature, slag composition and slag geotechnical
properties
using the established correlation to evaluate the process progress on production scale
establishing procedures for fast in-situ slag analysis during production
using characteristic data achieved for correlation of slag capacities and physical and chemical
properties to future develop active models for process control

On a production scale, trials were performed in the EAF, BOF, LF, as well as in ladle equipment for the
in-situ quick analysis of the molten slags produced. The expected benefits by introducing a new
technology for on-line slag analysis in metallurgical processes are a faster analysis and a simplified
analysis procedure. In its extension this will result in an improved process control as well as a better
yield regarding slag use both for process control and for road construction material. Using microwave
technology for slag analysis is a novel approach towards an instrument that could be used for slag
analysis on-line during production. Microwave technology is a versatile technology which has been
used in different ECSC funded project with a good outcome [1,2].

2 COMPARISON OF INITIALLY PLANNED ACTIVITIES AND WORK
ACCOMPLISHED
2.1 MEFOS
MEFOS part of the project has been focused on using microwave technology to investigate the
dielectric properties, i.e. refractive index, of both synthetic as well as production slag in solid and liquid
state. The derived dielectric data has been correlated towards slag temperature, slag composition (i.e.
slag basicity), as well as geotechnical properties. Furthermore, derived data has also been used during
production trials investigating the correlation between on-line production slags in order to evaluate
procedures for fast in-situ slag analysis during production.

Originally a frequency window ranging from 2 to 40 GHz was selected for the investigation. However,
above 18 GHz no microwave investigation of the slag was possible due to lack of signal reflection or
with a signal transmission with such a low signal to noise ratio that no useful information could be
extracted. In this respect 18 GHz was set to be the maximum frequency limit for the measurements and
no trials from frequencies above 18 GHz will therefore be reported.

During the project one of the contractors, RWI AB, went into bankruptcy. Because of this MEFOS took
over the RWI part of the work for the remaining time of the project. The microwave related part of the
project was therefore reorganized and delayed a few months. However, the tasks in the project could be
terminated in due time of the project.

7


2.2 ARCELOR
During the project time IRSID changed its name to ARCELOR RESEARCH.

The initial planned activities of ARCELOR have been completed and include the following activities:

solid slag re-melting, composition adjustment and casting for solid slag samples preparation
the preparation of slag samples for geotechnical applications
the monitoring of slag ageing and estimation of slag geotechnical properties
the analysis of results of microwaves tests performed by MEFOS on ARCELOR slags obtained
with controlled cooling rate

2.3 ISQ
All planed ISQ activities were carried out and includes:

- Chemical characterization of slags produced from experimental trials (supporting laboratory),
performing analysis with traditional equipment such as ICP/AES spectrometry, in order to
confirm/validate the data achieved with the microwave system;
- Leaching tests were conducted on several industrial slag samples, using lysimeter techniques
static and dynamic mode;
- Environmental Impact Assessment of steel slags recycling alternatives.

3 DESCRIPTION OF ACTIVITIES AND DISCUSSION
3.1 MEFOS
Trials have been performed on solid and liquid, synthetic as well as production, slag with the objective
to evaluate the dielectric properties, i.e. the dielectric constant. The dielectric properties will from now
on be expressed as the refractive index of the investigated slags. The trials were performed on
laboratory, pilot and production scale and the investigation of the slag properties were performed in
seven different frequency bands, each with a bandwidth of 4 GHz, ranging from 2 to 18 GHz.
In general the same type of microwave instrumentation was used for both liquid and solid investigations
on laboratory as well as on pilot scale. The microwave equipment used is presented below.

Microwave instrumentation

- Vector Network Analyser (VNA) type 360B, No 226003
- Test set Frequency Converter type 3630A, No 383002
- Signal Generator type 6769B, No 33001
- Broad band horn antenna 2-18 GHz
- Two linearly antennas 20-40 GHz
- Coaxial cables
- PC computer

3.1.1 Trial performance
Solid industrial slag on laboratory scale

Slag samples from LD converter, ladle slag and Blast Furnace from one low alloy integrated steel
producer and samples from the AOD converter as well as from the EAF from two different European
stainless steel producers have been investigated. The experimental set up for solid slags is shown in
Figure 1. In Figure 2 the instrumentation, a VNA and computer, used to collect data is shown. To avoid
sample contamination during trial performance each slag sample was kept in individual sample baskets
throughout the trial procedure. During the investigation the slag is exposed to microwave radiation and
8


a circularly polarised signal transmitted from the antenna penetrates the slag and is thereafter reflected
towards an aluminium foil at the bottom of the sample basket. The foil serves as a reference for the
bottom surface of the slag. The reflected signal is received at the same antenna but in opposite
polarisation mode and thereafter forwarded to the computer for signal processing and determination of
the refractive index.

The measurements were performed on solid slags samples with a particle size varying between 1 and
20 mm. The high alloy steel slags had in general a smaller fraction compared with low alloy steel slags.
Each slag sample did undergo three consecutive tests and the sample basket where rotated 120 between
every test. After each trial performance the average of the three tests were calculated and tabulated as
the refractive index. This trial procedure was used with the objective to minimize unwanted interference
from a coarse slag surface arising from the orientation of the individual slag particles as well as
impurities of excessive size inside the slag volume. Before the measurements started each slag sample
was defined in terms of thickness, volume and weight.

The chemical composition of the industrial slags investigated can bee seen in Table 1. In Table 2 the
results derived for each slag are presented as the refractive index related to each frequency band
investigated. All the reported values of the refractive indexes fall inside the limit for an accepted
signal/noise ratio better than 7. In Figure 3 and Figure 4 the derived refractive indexes are shown
graphically. It can bee seen from Table 2 that the average refractive index is somewhat higher for slags
from low alloy steel production compared with the refractive index for slags from high alloy steel
production and that it also varies due to the frequency bands. These variations that can be seen in
Figures 3 and 4 could to some extent be explained as an impact from a coarse slag surface with a
disadvantageous orientation of the slag particles. From the figures it can also be seen that the refractive
indexes for high alloy steel slags varies less then for low alloy steel slags. This could probably depend
on that the high alloy steel slags investigated had a smaller fraction compared with the low alloy steel
slags. However the origins of the variations are not statistically significant, mostly due to the limited
number of tests performed, and the statement of their origin can therefore not be conclusive.

In Figure 5 is the refractive index plotted as a function of the slag basicity, defined as
(CaO+MgO)/(Al
2
O
3
+SiO
2
), for all the investigated solid slags. Nevertheless, no explicit link between
the slag basicity and refractive index can be seen. In Figure 6 is the index plotted in a similar way for
low alloy steel however nor in this figure an explicit link between the slag basicity and the refractive
index is seen.

Liquid synthetic and industrial slag on pilot scale: campaign 1

Two campaigns were performed in an induction furnace with the microwave antennas suspended above
the mouth of the furnace. In Figure 7 and Figure 8 is the experimental set-up shown. In Table 3 the
furnace data is presented. The same microwave instrumentation used for trials on solid slag was also
used for the liquid measurements.

A descriptive picture of the experimental set-up is presented in Figure 9. It can be seen that some
geometrical constrains apply to the experimental set-up as the opening angle of the antenna(s) are large
and the radius of the furnace is relatively small. This means that the antenna(s) must be positioned close
to the furnace opening to assure measurements free from interferences from reflecting signals. However
if the antenna is positioned too close to the furnace, the antenna itself will become a part of the furnace
and the measurements will be spoiled. Another aspect that must be considered is that if the antenna is
positioned too close to the furnace then the heat radiation will be a problem. To circumvent the heat
radiation problem the furnace is equipped with a ceramic lid, see Figures 8 and 9, which is transparent
to microwave radiation and almost opaque to heat radiation. From the above implications it was found
that the smallest inner diameter of the furnace, that could be used as experimental set-up, was 25 cm.
The furnace was lined with standard refractory material during campaign 1 and with a graphite crucible
during campaign 2.

9


Trial procedure

Each slag sample was charged into the furnace in which a 150 kg low alloy steel melt had been
prepared. The steel melt serves as a reflecting surface for the microwaves. Before charging the slag
sample, aluminium was added to the steel melt to assure low oxygen potential and a minimum of
reoxidation reactions between steel and slag during the experiments. The melting of the slag required
between 30 and 60 minutes depending on the slag quality. When the slag sample was totally melted and
the temperature was between 1550 and 1600 C the microwave measurements started.

The amount of slag added was controlled by the condition that each investigated liquid slag should have
an approximate thickness of 5 cm, throughout the experimental sequence. The furnace was flushed with
argon to prevent contamination of the slag during the experiments. Slag samples were extracted at the
beginning and at the end of each measurement sequence.

Results and discussion

During the first campaign 10 synthetic slag mixtures in the system Al
2
O
3
-CaO-SiO
2
were planned to be
investigated, see Table 4, but due to problems with refractory wear during the trials only six slags could
be fully investigated, see Table 5. The refractory wear was to some extent anticipated due to the basicity
range of the synthetic slag mixtures investigated, see Table 4, however the proportion of it was of a
magnitude requiring a relining of the furnace which was not anticipated.

The serious refractory wear encountered during the experiments is illustrated in Figure 10 where the
slag compositions at the beginning of melting, start and end of measurement are shown in an isothermal
section of the Al
2
O
3
-CaO-SiO
2
phase diagram at 1600
o
C. In the figure are all the intended synthetic slag
compositions shown. The compositions in Figure 10 are normalized to the Al
2
O
3
-CaO-SiO
2
system,
which explains why some slag compositions are in two phase regions of the diagram. Initially seven
industrial slags were at hand but due to a too high melting point, for the induction furnace to cope with,
the stainless steel slags had to be omitted from the liquid trials.

Results from the slag analyses are shown in Table 6. Samples numbered L406-1 to L414-1 were
extracted at the start of measurement sequence and the samples numbered L406-2 to L414-2 at the end
of the sequence.

From Figure 10 as well as Table 6 it is clear that the slag composition does not differ considerably
between the start and the end of the measurements except for synthetic slag 2 which exhibits a large
change in alumina content throughout the melting. However the initial slag composition changes
considerably during the melt down phase until the microwave measurements start for all synthetic slags.

The industrial slags did not show any abnormal refractory wear probably depending on the higher slag
basicity combined with a higher MgO content for these slags. The results of the refractive index
observed for the different slags at different frequency region are presented in Table 7.

No data could be obtained for synthetic slag 2. The reason for this has not been possible to trace
however a combination of impenetrable slag and the signal performance is possible. There were also
some problems obtaining values in different frequency region for the other investigated slags. As can be
seen from Table 7 the number of approved signals decreases with increasing frequency, however for the
frequency window 20-40 GHz the signal intensity was without any exception of such a low quality that
the data collected had to be omitted in the evaluation. The uncertainty regarding the reported values
were divided into three different groups depending on the signal/noise ratio, accepted S/N >7, probable
3<S/N<7 and ambiguous S/N<3. By analysis of the results seen in Table 7 it may be found that the
frequency regions 2-6 GHz, 4-8GHz and 6-10GHz have been most successful obtaining acceptable
signal/noise ratios. This could imply that those frequency regions are the most suitable when
performing microwave investigations on the refractive index on liquid slags on production scale.

10


In order to illustrate the refractive index, n, dependency of the frequency the results for synthetic slag 7
and the results for the ladle slag are graphically presented in Figure 11 and Figure 12 respectively. In
Figure 13 the result for all investigated slags are presented in the same way. One would anticipate that
the refractive index, n, should decrease with a higher frequency, however this dependency is not clearly
seen from the results from the measurements. This could be due to the rather small frequency region
investigated from 2 to18 GHz or other physical slag properties not addressed in this report.

In order to show the slag composition dependency of the refractive index, indexes in three different
frequency bands, 2-6 GHz, 4-8 GHz and 6-10 GHz, are plotted against different definitions of the slag
basicity, see Figures 14, 15 and 16.

In the figures all slags are represented, with the exception of synthetic slag 3, in the frequency band 4-8
GHz. In Figure 14 the slag basicity is defined as the ratio between CaO and SiO
2
and an increase of the
refractive index is shown when the basicity increases. A similar behaviour is observed in Figure 15
where the slag basicity is defined as the ratio (CaO+MgO)/SiO
2
and also in Figure 16 where the basicity
is defined as the ratio (CaO+MgO)/(Al
2
O
3
+SiO
2
). However in both Figure 14 and Figure 15 a clear
discrepancy from the trend of increasing refractive index with respect to the basicity is shown for one
slag sample corresponding to the LD converter slag. This sample shows the largest refractive index
observed in the frequency region 2-6 GHz, n=3.6, and the slag basicities are 1.2 for the definition,
CaO/SiO
2
, and 1.8 for the definition, (CaO+MgO)/SiO
2
. On the other hand this divergence from the
observed trend is not shown in Figure 16 where the defined slag basicity includes alumina.

The LD converter slag has the lowest alumina content, ~4.7 %, of all the investigated slags and this
could influence the result of the observed refractive index. Alumina is amphoteric in nature, which
means that it can act as both an ionic ion and a network forming ion depending on the slag composition.
This complicates the situation regarding the structure of the liquid slag and consequently also the results
of the measured refractive index. It may be difficult to penetrate the reasons why the LD slag deviates
from the other slag samples as the measured signal for the LD slag was in the accepted limit with a
signal/noise ratio better than 7.

Generally the refractive index of a polar matter should exhibit a higher value than a non polar matter.
The observed increase of refractive index with increasing slag basicity is in line with this as a higher
basicity would give a break-up of the covalent bindings between silica tetrahedras and therefore give a
more polar matter.

There are some uncertainties in the determination of the refractive index in particular with respect to the
changing slag composition during the measurements. This was mainly due to the excessive refractory
wear. Another source of experimental uncertainties arises from the low signal/noise ratio observed for a
number of slags in different frequency regions. Nevertheless a trend may be noticed where the
refractive index seems to increase with the slag basicity. The tendency with an increase of the refractive
index appears to be mostly visible in the frequency region 2-6 GHz. This tendency would indicate a
connection between the slag composition and the refractive index. However, to quantify this relation-
ship a large number of further controlled experiments are needed.

Liquid and solid synthetic slag on pilot scale: campaign 2

With respect to the abnormal refractory wear during the first campaign on liquid slags a second
campaign, with a slightly different approach, was performed. The furnace was modified to hold a
graphite crucible with an inner diameter of 28 cm, see Figure 7. By using graphite as refractory
material, the 150 kg low alloy steel melt used in the previous trials was changed to a FeSi melt with a
silicon content of 75 % and by using graphite lining no further refractory wear was expected.

Trial procedure

The trial procedure for the second pilot campaign was in most similar with the first campaign. The
melting of the slag samples required between 30 and 60 minutes and the microwave measurement
started when the slag material was totally melted and the temperature was between 1550-1600C.
11


Directly after each measurement was finished, a slag sample was extracted and the remaining slag was
removed from the crucible. Thereafter a new measurement sequence started with charging of a new slag
material. Chemical analysis of the slag samples was carried out elsewhere.

The main examined slag system in this investigation was Al
2
O
3
-CaO-SiO
2
. Three trials were also
performed with the Al
2
O
3
-CaO-MgO-SiO
2
slag system. The amount of slag in all trials with liquid slag
was adjusted to a slag layer thickness of 5 cm. All slags were made out of pure Al
2
O
3
, SiO
2
, CaO and
burned dolomite lime. Basically the same approach of slag mixtures presented in Table 5 was used here
with addition of burned dolomite for three slag mixtures.

During this campaign measurements on solid slags were performed prior to furnace trials.

Results and discussion

The melting of the slag materials worked very well and no wear of the graphite crucible could be
detected, except from the rim of the crucible that was partly oxidised. Under the ferrosilicon bath level
some accretions, possible made out of silicon carbides, could be detected. The chemical analysis of the
liquid slag samples are summarised in Table 8. The impurity (e.g. Na
2
O or K
2
O) levels were low in all
the trials.

The microwave measurements of the refractive indexes of the solid slag samples are summarised in
Table 9 and the refractive indexes of the liquid slag samples in Table 10.

The derived refractive index for liquid slags is plotted versus different definitions of slag basicity and
frequency bands in Figures 17, 18 and 19. In Figure 17 the basicity is defined as the ratio of CaO to
SiO
2
(B2). Since MgO is also a basic slag constituent, the (CaO+MgO) to SiO
2
ratio (B3) is more
appropriate, which is presented in Figure 18. In Figure 19 the refractive index versus the (CaO+MgO)
to (Al
2
O
3
+SiO
2
) ratio (B4) is plotted. As mention above, alumina is amphoteric and can act both as a
basic or an acid constituent depending on the slag composition. From the figures it can be seen that
between some frequency bands there is a tendency of increasing refractive index with increasing
basicity however not consistently throughout the frequency bands and this observation can therefore not
be conclusive. In this respect further trials are necessary to reach such a conclusion.

Solid and liquid production slag on pilot scale

Generally the refractive index is lower and more constant with respect to the frequency bands for the
solid slags compared with the liquid slags. Two of the slags investigated in the solid trials were the
same slags as investigated in the trials with liquid slags. A comparison between the solid and liquid
result regarding the frequency dependent of the refractive index for the LD slag and ladle slag are
presented in Figure 20. From the figure it is seen that refractive index for the liquid LD slag is higher
than the solid LD slag. The result for the ladle slag shows an opposite relationship where the refractive
index is higher for the solid than for the liquid slag. It may be difficult to draw any more conclusions
from this figure but the overall impression would be that more trials are necessary to find an
explanation for the observed results. The numeric values of the refractive indexes can be seen in
Tables 2 and 7.

If a plot is done using only the slags from the low alloy steel a similar result as seen for the liquid slags
may be observed as in Figure 6. Though, the same discrepancy concerning the LD slag as found in the
liquid slag trials may be seen for the solid slags. As only two values for the refractive index could be
establish in the analysed frequency ranges 2-6 GHz, 4-8GHz and 6-10 GHz, it may be troublesome at
the present to find a scientific explanation to the observed discrepancy.

Slag from ARCELOR on pilot scale

Slag samples after 7 months of ageing from ARCELOR were remelted and investigated. The slag was
remelted in the induction furnace described in Table 3 and schematically shown in Figure 7. The
melting procedure was identical to the previous ones on synthetic and industrial slags. The refractive
12


indexes, presented in Table 11, were calculated from the microwave measurements. The magnitude of
the indexes derived are higher compared with the ones derived on similar production slags in the
previously investigations on industrial slags however no scientifically explanation to this can presently
be given.

Trials on production scale

Production trials have been performed on ladle slag prior to and after treatment in the CAS-OB(2) at
SSAB in Lule. The trials were performed on a moving lade to evaluate the possibility to speed up the
slag analysis during production. Based on the data gained during trials on pilot scale three different
frequency bands, 2-6 GHz, 4-8 GHz and 6-10 GHz were used during trials performance. In Figures 21
and 22 the antenna unit suspended above a ladle can bee seen.
Throughout the trials the slag prior to the treatment in the CAS-OB(2) was in general not fully
dissolved and thus with a very coarse and varying slag surface. This affected the individual
measurements and the results derived could not be used in the subsequent work in the correlation of
process data against data accomplished during pilot trials.

The slag after treatment was dissolved and occasionally thin, however some slag thicknesses could be
calculated see Table 12. The previously derived refractive indexes were used to find a correlation
towards the slag composition however the outcome of this work was not satisfactory and no explicit
relations cold be found.

3.2 ARCELOR
3.2.1 Slag samples and trial performance
Different slag samples have been obtained using liquid slag prepared in the 6 ton pilot EAF equipped
with a chute used to tap the liquid slag into moulds of different volume, see Figure 23. Two different
slag basicities and two slag tapping temperatures have been characterized.

The liquid slag has been produced by using solid BOF slag mixed with approximately 5.3 tons of scraps
into the 6 ton pilot EAF. Depending on the aimed slag basicity (3.2 and 4.2), some lime has been added.
To obtain the different slag cooling conditions, each liquid slag type has been tapped into three moulds
having different thicknesses: 3, 7 and 13 cm thus producing a total of nine slag samples. The individual
dimensions of these three moulds are: 3 x 96 x 150 cm: 7 x 80 x 80 cm and 13 x 80 x 80 cm, see Figure
24. Specifically designed equipment was used to measure the slag temperature evolution during the slag
cooling, after tapping, in the middle of the slag layer in the mould, see Figure 25.

The procedure applied to perform these trials was:

melting of 950 kg of BOF slag mixed with 5.3 t of scrap
Casting of the 1
st
slag (CaO/SiO
2
= 3.3) into three different moulds (3 different
thicknesses) at a temperature around 1690C, see Figures 26 and 27
addition and melting of 400 kg of BOF slag
Casting of the 2
nd
slag (CaO/SiO
2
= 3.5) into three different moulds (3 different
thicknesses) at a temperature around 1630C
addition and melting of 500 kg of BOF slag and 100 kg of lime
Casting of the 3
rd
slag (CaO/SiO
2
= 4.2) into three moulds (3 different thicknesses) at a
temperature around 1670C.

The compositions of the slags are given in Table 13 and the casting conditions (thickness of the slag
layer, casting temperature, physical state of the slag) in Table 14.
The slag compositions are different depending on the trials:

13


slag 1 and 2: have a basicity around 3.3 to 3.4, different casting temperature and iron oxide
content, 1680C and 15 % for slag 1 and 1630C and 25 % for slag 2,
slag 1 and 3: have different basicities and iron oxide content respectively: 3.3 - 15 % for slag 1
and 4.2 26 % for slag 3. The casting temperature is approximately 1670C.
slag 2 and 3: have different basicities and casting temperatures 3.5 1630C for slag 2 and 4.2
1665C for slag 3. The iron oxide content varies around 25 %.

Concerning the slag oxidation, it has not been possible to correctly control the iron oxide content in the
slag for trials 2 and 3. This iron over oxidation was mainly due to the air intake in the EAF during the
melting and preparation phases of slag 2 and 3. In fact the composition of slag 2 and 3 is rather
characteristic of a converter slag after oxygen post blowing rather than a standard BOF slag without
oxygen post blowing. The physical state of the slag has been determined using an ARCELOR
mathematical model which is able to describe the proportion of the liquid slag and the nature of the
present solid phases. For these trials, the slag was practically always liquid except the first slag where
there was some solid MgO. The amounts of solid slag used for each slag sample ranges between 40 to
230 kg.

3.2.2 Slag temperature evolution during cooling
The temperature evolution during the slag cooling in each mould was recorded during the complete
cooling period. The different cooling curves for slag 1, 2 and 3 are shown in Figures 28, 29 and 30
respectively.

Comments about the slag cooling temperature evolution:

for each slag tapping, the thermocouple has been put in position around 2 to 3 minutes after the
slag casting started. This means that there is no information regarding the temperature evolution
during the first 2 to 3 minutes of the slag cooling period. In Figures 28, 29 and 30, the
temperature indicated at time 0, thus corresponds to the temperature measured in the electric arc
furnace at the beginning of the slag casting.
for slag 2, see Figure 29, the significant slag temperature fall during the first 35 minutes in the 13
cm slag thickness mould is unusual. This temperature fall is explained by the formation of gas
pockets inside the slag layer, see Figure 31. These gas pockets caused a bad contact between the
thermocouple and the slag. However, after the destruction of the gas pockets in combination with
sand addition near the thermocouple to improve the contact with the slag, the temperature fall
stabilized and a plausible slag temperature evolution was obtained after 90 minutes.
for slag 3, see Figure 30, concerning the 3 cm thickness, the bad contact between the
thermocouple and the slag was due to the rapid solidification of the slag and in this case it was not
possible to insert the thermocouple in the slag layer. The almost normal slag temperature
evolution observed after around 60 minutes, is due to sand addition to improve the contact
between the thermocouple and the slag (similar as in trial with slag 2 and 13 cm thickness).
Figures 32, 33 and 34 show the temperature evolution during slag cooling for three different slag layers:
3, 7 and 13 cm. Using these graphs, it is possible to evaluate an average slag cooling temperature versus
the slag layer thickness. Here 3 trials, for which the temperature evolution was unusual (slag 3 -
thickness 3 cm, slag 1 thickness 7 cm and slag 2 thickness 13 cm), have not been considered. In
Figure 33, the lower temperature evolution of slag 1, compared to the temperature evolution of slag 2
and 3, is explained by the fact that the thermocouple was not immersed into the middle of the slag layer
but merely positioned directly under the slag surface.
The average slag cooling temperature evolutions for different slag thickness are shown in Figure 35 and
Table 15.

14


3.2.3 Slag samples treatment
The massive solid slag samples have been ground and sieved to obtain a grain size in the range between
6.3 to 15 mm as required for road applications. The grinding was carried out at ARCELOR using in
house grinders. Due to the small size of the grinders, it was necessary to carry out this operation in
several consecutive steps by reducing the breaker jaw distance, at each step, from 45 mm to 8 mm. At
each grinding step, the amount of slag with a grain size bigger than 15 mm, was recycled and the
grinder was adjusted to a smaller breaker jaw distance each time. At the end of the grinding operation,
the proportion of crushed slag in 6.3 to 15 mm range is in between 57 to 75 % and the proportion of
fines with grain size lower than 6.3 mm is in between 23 to 41 % range.

With the present slag samples, neither the slag quality (basicity, tapping temperature) nor the cooling
rate seems to have any effect on the fines formation during the grinding operation, see Figures 36 to 38.
For ageing, the different crushed slags were arranged outdoors in layers of approximately 10 to 15 cm
thickness.

3.2.4 Evolution of composition of the slag samples during ageing
The chemical composition of the different slag samples was determined after a 7 months ageing period
(first semester 2003), during which the slag samples were kept outdoors and also during a period with
exceptional little rainfall.

The chemical compositions of the different slag samples corresponding to the initial liquid and solid
slags, and after 2, 4 and 7 months of ageing, are given in Tables 16 to 18 concerning XRF analyses and
Tables 19 to 21 concerning the iron oxidation level and the free lime content. The XRF results show
almost no changes of the chemical composition of the different slag samples during the 7 months of
ageing period. The differences observed are equivalent to 1 relative % rather due to the
representativeness of the different samples, than the ageing effect. The same goes for the iron oxidation
evolution.

Concerning more especially the free lime content, see Tables 19 to 21, the analysis has been performed
by a chemical method. This method is based on slag sample dissolution with ethylene-glycol, filtration
and measurement of Ca by ICP-OES. During the 2 first months of ageing, one can observe a free CaO
content decrease depending on the different slag samples. It seems that the free lime content decrease is
more significant for the slag samples cooled down with a low cooling rate (corresponding to 13 cm slag
thickness). This decrease has not been confirmed by the analyses made during the rest of the ageing
period (after 4 and 7 months) as can bee seen in the Figure 39.

In fact this chemical method gives the total calcium that has been dissolved by the solvent and not only
the Ca coming from free CaO. The Ca dissolved can come from different phases such as:

quicklime (CaO),
calcium hydroxide (Ca(OH)
2
),
calcium carbonate (CaCO
3
),
combined with others elements as iron (CaFeOx) or manganese (CaMnOx) with an important
CaO content, higher than 90 %.
In parallel with the chemical method, an XRD analysis was performed on the different slag samples.
This type of analysis method is more a qualitative than quantitative and able to determine the proportion
of each phase present in the slag sample. It can bee seen that the principal phases present in the slag are
larnite (Ca
2
(SiO
4
)), iron calcium oxide (Fe
2
O
3
(CaO)
2
) and hematite (Fe
2
O
3
) and that lime (CaO) is
detected only in 2 out of 3 slag samples as shown in Table 22. In the table are the results of the XRD
analysis on crushed slag sampled at the beginning of ageing period presented. These results show that
the mineral composition of the different slag samples does not vary extensively despite the conditions
of the solid slag samples due to basicity, tapping temperature and cooling rate.

15


The XRD analysis, performed on slag sampled late during the ageing period, has not shown real
changes of the different phases during the 7 months of ageing. The comparison of the results obtained
by the chemical method and by the XRD method seems to confirm that the evolution of those results
obtained by the chemical method is mainly due to the slag samples representativeness rather than an
evolution of the free CaO content, because the total dissolved Ca has not really change during the
ageing period, see Figure 39. Perhaps there was a modification of the phases nature from CaO to
(Ca(OH)
2
) and/or (CaCO
3
) without possibility to make a quantitative evaluation of each concerned
phases using XRD method.

3.2.5 Geotechnical tests results
The geotechnical tests permits evaluation of the physical characteristics of slag lumps in relation to their
use for road construction.
Two different geotechnical tests have been performed:

Los Angeles (LA) :
This test evaluates the shock-resistance by measurement of the grain size evolution.
It is a standardized test where slag lumps (5 kg) and steel balls (1 cm diameter) are put in a sphere
and the result of the test is provided as a new grain size of the slag after a defined number of rotations
of the sphere.
Micro Deval (MDE) :
This test evaluates attrition (wear, abrasion and fragmentation of slag lumps by mutual friction) by
measurement of the grain size evolution.
It is a standardized test where slag lumps (1 kg) are mixed with small steel balls and water in a
sphere and thereafter measure the new grain size of the slag after a defined number of rotations of the
sphere.

Due to the insignificant evolution of the slag composition, these tests were only performed on slag n1
(basicity 3.2 and iron content around 16 to 17 %) sampled after 3 months ageing. The choice of this slag
was done because it was the only slag that had a composition of a normal LD slag at the end of a
converter blowing without oxygen post blowing.

The obtained results show that the increase of the cooling rate decreases the slag quality regarding the
geotechnical properties and decrease also the bulk density of the slag, see Table 23.

Concerning our test, only the slag with a lower average cooling rate around 15 to 20 C/min during the
10 minutes after the slag tapping (corresponding to the slag thickness of 130 mm) would have the
properties required for road applications. For comparison, see Table 24 in which the LA and MDE
coefficients for different other materials are presented.

3.3 ISQ
3.3.1 Physical-chemical analysis of slags from experimental trials
Laboratory analysis of several slag samples from the experimental trials at MEFOS were carried out
using traditional analysis equipment such as ICP/AES, in order to support the data achieved with the
microwave system. During 2002 and 2003, several samples were sent by MEFOS, in order to quantify
the following elements of interest: FeO, CaO, SiO
2
, Al
2
O
3
, MgO, TiO
2
and MnO. The results obtained
are presented in Tables 25 to 27.

Prior to chemical analysis, it was necessary to crush and grind all samples to an adequate granulometry.
The samples were then digested in a chloridric and nitric acid mixture. The digestion is followed by an
oxidation process with perchloric acid, in order to precipitate all silica. The solution is then filtered and
the silica precipitate is burned and volatilised by fluoridric and sulphuric acid to determine its silica
content.

16


The remaining solution is then diluted in adequate proportions in order to fulfil the requirements of
ICP/AES determination methods, previously studied (detection and quantification limits and matrix
interferences).
Five determinations were made for each sample, paying particular attention to homogeneity criteria in
sample composition and allowing the calculation of standard deviation and result variation coefficient.

Simultaneously, ISQ performed determinations in Certified Reference Materials (CRM) with chemical
compositions similar to the samples of interest, following the same analytical procedures and work
methodology.

3.3.2 Analysis and characterization of slags for recycling purposes
Slags are long viewed as valuable by-products of steel industry. Some of the recycling alternatives used
worldwide include cement incorporation, production of construction, road base and sub-base
aggregates, soil improvement, and others. The application of these recycling techniques is now a reality
for several decades.

ISQ performed an inventory of the state-of-art of steel slags final destinies as well as an environmental
impact assessment associated with their processing and further application. This study was based on
several bibliographic references, reporting previous studies on the reuse of these materials.

Environmental Impact Assessment of steel slags reuse

The idea that steel slags are a waste product to be disposed of is long gone. During the early phases of
its commercialisation, these materials have traditionally been utilised only in very low value
applications, occasionally at a cost to the steel producer. These early applications were ways/roads
consolidation and paving.

In the last years, the range of applications has grown due to several worldwide efforts made by the steel
industry in order to reach sustainable development, improving the environmental performance and the
added value of its by-products, turning them into valuable raw-materials to other activities and markets.
Therefore, many new recycling alternatives have been developed as well as several slag processing
units, within steel plants or not, to comply with specific needs in the market. In Table 28 the main
applications of EAF and BOF slags worldwide are resumed.

In Europe, the most popular applications are cement incorporation and production of road base and sub-
base aggregates. According to EUROSLAG, the European Association of Slag Producers and
Processors, these applications account for 42% (18.8 MTon) and 37% (16.7 MTon) of total produced
slag for road construction and cement incorporation, respectively.

Although the application of such recycling techniques is a reality for several decades, in Europe there
are still significant slag quantities being dumped into landfills (7%, 3.4 MTon) or stockpiled for long
periods at steel plants (3%, 1.2 MTon).

These values are justified by the fact that the characteristics of some slag types do not comply with the
high quality demands of some applications (namely in terms of volume stability) and their processing is
not economically feasible. Ladle slags for instance have a wide range of possible chemical
compositions and are produced in small quantities and therefore, these slags are usually processed
together with EAF slags or, simply, sent to landfill.

Regarding slag stockpiles, these are invariably generated due to the process of slag stabilisation by
weathering. Stockpiles increase during periods of weak demand or minimal infrastructure investment.

Steel slags always had fewer applications than blast furnace slags, because steel slags are most difficult
to deal with due to the wider range of possible chemical compositions and the high concentrations of
17


reactive free lime and magnesium oxide in their structure, causing volumetric instability and thus
implying its treatment. This reality is however being inverted due not only to the decrease of integrated
steel plants world wide, but also to the development of the steelmaking process, optimising slag quality
and slag processing operations.

Recent developments made possible the use of steel slags in areas previously served only by BF slags.
In the area of slag cement materials for instance, steel slags are blended with BF slag for cement
incorporation because steel slags are richer in iron oxide.

Analysing the information presented in Table 28, iron and calcium oxide recycling in the blast furnace
and the recycling of entrained metals into the metallurgical circuit are alternatives with decreasing
tendency. On one hand they are restricted to integrated steel plants that have the capacity to tolerate the
recirculating feedstock at the blast furnace, on the other hand due to the decrease of blast furnace plants
worldwide and the growing demands for steel quality.

The landfill also follows a decreasing tendency due to stricter environmental requisites and, on the other
hand, the growing demand for these materials as raw materials for cement industry and civil
engineering.

Recent references refer the application of steelmaking slags in waste stabilization, mine acid effluent
neutralization and use as raw material in the ceramic and glass industries. The information concerning
these new applications is however very scarce and needs further R&D.
Environmental impacts associated with processing and reuse
Steel slags are subjected to several treatments, basically to remove free lime and magnesia and metal
inclusions. For the majority of applications, it is essential a basic preparation of the slag, which is
performed in slag processing units, inside the steel plant or not. This preparation includes procedures
such as cooling (by natural weathering or dry or wet quenching), crushing, screening, and magnetic
separation of metal inclusions such as iron or steel.

The possible environmental impacts associated with the processing and reuses of these materials are
reported subsequently. Reported impacts are very rare and are normally associated with bad industrial
practices.
Emissions to air
In slag processing units, emissions of total suspended particles (TSP) and metal oxides can occur during
operations such as chilling, quenching with air or steam, grinding, sieving, blending (with other
materials) and granulation.

Regarding landfill and stockpile operations, generation of dust storms (TSP) from the slag stockpiles
during high winds must be considered.

In cement industry, the use of slags has a beneficial effect in reducing carbon dioxide emissions at the
kiln by utilising the pre-calcined lime in the slag rather than burning more limestone.
Emissions to water
When exposed to rainwater, steel slags generate alkaline leachates due to the hydration of free CaO an
MgO present in the slag matrix. The reaction of these two oxides with water generates calcium
hydroxide and magnesium hydroxide respectively, responsible for the high pH of the leachates. These
hydroxides can reach pH values higher then 11. Some authors report cases of corrosion of certain
materials in contact with these leachates, such as aluminium and galvanized steel [3].
While the slag is stored at the steel plant, since it is exposed to the elements, soluble components of the
slag leach out of the matrix and percolate through the soil, possibly reaching underlying aquifers,
process that can take a number of years. Besides the major chemical components of steel slags, such as
18


calcium, silicium, aluminium or magnesium oxides, these materials can have other elements considered
as toxic, although in trace amounts, such as arsenic, cadmium, chromium, nickel and lead. Chromium is
the most preoccupant, ranging from 1000 to 3000 mg/kg in BOF slag and 5000 to 22000 mg/kg in EAF
slag [3].

Another phenomenon associated with slags exposure to the environment is the formation of calcium
carbonate precipitates, formed due to reaction of calcium hydroxide with atmospheric carbon dioxide.
Sporadic occurrences of water courses obstruction have been previously reported in literature.

The knowledge of the chemical properties and leaching behaviour is important not only for slags alone,
but also for slag blended materials, especially when these are supposed to be in contact with water.
About incorporation in concrete, for instance, a previous study of the American Nuclear Society
reported the leaching of several elements by mass desegregation in slag cement materials exposed to
deionised water and synthetic seawater [4].
Generation of Waste
Regarding waste production associated with slag recycling, the only environmental impact to consider
is the eventual destruction of structures with incorporated steel slag, due to the above referred
expandable tendency of this material. This phenomena can however be avoided by an effective
stabilization of the slag prior to its reuse.
Noise
Noise emissions are associated solely during processing of slags in operations such as crushing and
grinding. These can be avoided using several sound isolating techniques.
Landscape
The negative effects on landscape are basically caused by big slag stockpiles at steel plants. Stockpiles
and landfills cause negative impacts on surrounding populations which can affect the communitys
perception of the entire steelworks. These alternatives also cause the occupation of large areas of land
which could be much more usefully and profitably used for other purposes. Legacy issues can also
occur on cessation of operations.
Presently, many old slag stockpiles are being reclaimed in order to process the slag for further
applications, thus improving the landscape, freeing large areas of land and saving natural resources.
Socio - economy
The processing and reuse of steel slags brings anything but positive impacts to socio-economy, such as
the reduction of demand for virgin materials thus saving natural resources and the creation of direct and
indirect employment. The use of slag in some applications (such as cement making) captures all of the
energy that has been invested into the slag.

There are however some limiting factors for some applications such as obtaining a suitable slag within
reasonable transport distance and the constant rather than cyclical demand for slag.

For instance, granulated material is subject to fluctuations in market demand. On the other hand, ground
granulates must compete not only against Portland cement but also fly ash, which is readily and cheaply
available in many regions.

Granulation process requires not only a significant capital investment but also access to large quantities
of fresh water. However, there have been some recent developments in granulating with reduced water
consumption, due to development of dry granulation processes, now in the early stages of
commercialisation.

19


Mitigation measures
The first step for slag management must be performed inside the steel shop, with the development of
production techniques and practices towards the reduction of slag per ton of steel. Since zero waste
production is not yet realistic, opportunities for slag minimization must be studied including the
reduction of slag volume through better control of lime input to the furnace, etc.

The previous treatment of steel slags prior to their reuse is nowadays a common and positive practice,
in order to assure the protection of the environment. As referred above, in slag processing units there
are several unit operations that can generate impacts in the environment if not properly managed.
Therefore, during slag processing, monitoring plans should be maintained according to applicable
legislation in each country, in order to monitor the emissions to air and water.

Long term stockpiling of slags inside the steel plants is still a common practice in many countries. In
order to protect underlying soils, slags should be deposited in impermeable pavement, ideally with a
system to collect and treat the leachates generated.

The evaluation of slag reuse is a low cost, high value option for steel plants. The technology for
reducing slag volume and increasing its value to other industries exists. It is dependent more on steel
chemistry and operating practices than on capital investment.

Leaching tests

The environmental impact of most concern when handling and reusing steel slags is the leaching of
soluble components from the slag matrix, when these materials are in direct contact with soil, exposed
to rainwater or when in permanent contact with surface and sea waters.

Therefore, the knowledge of leachates chemical composition and their behaviour is environmentally
relevant. Several leaching tests were conducted in order to monitor the leachates composition
throughout slags aging process, and thus assess their compatibility with environmental regulations and
their sound use in several applications.

Leaching tests were carried out with EAF and ladle slags from Siderurgia Nacional (SN), a Portuguese
steel producer. Slag samples were let to age by natural weathering at SN slag storage facility in Porto,
in the period of December 2002 to July 2003. The chemical composition of the slags has been
determined after 3 and 6 months ageing. Results are presented in Tables 29 and 30.

During the referred ageing period, slags were exposed to high atmospheric precipitation levels,
especially in December (280 l/m
2
), January (250 l/m
2
), March (110 l/m
2
) and April (130 l/m
2
) [3].

Despite the high precipitation levels verified in Porto, especially during the first three months of aging
by weathering, the results show no significant evolution of the chemical composition. The differences
observed for ladle aged three months are probably due to the different homogeneity (representativeness)
of the sample.

3.3.3 Flow through test results
The test is constructed as a column or a lysimeter, simulating site-specific conditions, namely the
landfill of a material and the natural leaching of pollutants from the solid matrix. The experimental set
is shown in Figure 40A (EAF slag) and Figure 40B (ladle slag).

Trials were conducted with EAF and ladle slag samples, from the same heat, in three different ageing
levels, referenced in Tables 29 and 30, for a period of 30 weeks. During laboratory trials, samples were
watered every week with one litre of distilled water and the leachates collected 3 days after.

20


Several phenomena normally occurring in these trials were observed, specifically, channeling, system
clogging (in ladle samples), biological growth and oxidation of metallic iron inclusions (in EAF
samples), see Figure 41.

The leachates were analyzed in order to monitor relevant parameters such as heavy metals (Fe, Mn, Ti,
Al, Ca, Mg, Cr), pH and electrical conductivity. The results are presented in Figures 42A to 42G for
EAF slags and Figures 43A to 43F for ladle slags. Each graphic presents the total leached quantity of an
element in mg/kg of dry sample (cumulative values) during 30 weeks of trials.
EAF leachates presented a variation of pH ranging from 7.5 to 10.5 and conductivity, very steady along
the ageing period, with values around 20 mS/m. Regarding heavy metals, the leaching of Ti, Cr and Al
(Figures 42B, 42D, 42F) doesnt seem to be affected by the ageing process, while the releases of the
other monitored elements increase in older samples.
Ladle leachates presented less stable values of pH (7.5 to 12) and conductivity, which decreases from
360 mS/m to 30 mS/m. Heavy metal releases increase with ageing, although the rate of leaching
decreases over time.
Over the entire duration of the trials, the individual concentrations in mg/l appeared to be independent
of time for the majority of the chemical species, i.e. there is no specific tendency but an oscillating
behaviour. This fact was also verified in previous similar tests reported in literature [4].
Despite the general increasing trend of heavy metals release with ageing, after more than one year of
trials, these slag samples present very low leaching levels. Calcium and magnesium leached from solid
slags are below 0.5% and all other metals below 0.05 %.

3.3.4 DIN 38414-S4 Determination of leachability by water
DIN batch test was used to evaluate leachability of heavy metals by water. Samples are mixed with
distilled water in a liquid to solid ratio of 10:1 during 24 hours, using a rotative agitator, see Figure 44.
The sample is left to decant and the leachate obtained is filtered and analyzed. The method allows
repeated leaching of the residue (solid fraction).

Trials were conducted with fresh EAF and ladle slags, see Tables 29 and 30. The leachants were
analysed in order to monitor relevant parameters such as heavy metals (Fe, Mn, Ti, Al, Ca, Mg, Cr), pH
and electrical conductivity. The results are presented in Figures 45A to 45F for EAF samples and
Figures 46A to 46F for ladle samples. Each graphic presents the total leached quantity of an element in
mg/kg of dry sample (cumulative values) during eight extraction cycles.

Solubility behaved very unstably in the first extraction cycles and consequently the trials were repeated
eight times but no sample stabilization occurred.

In EAF samples, pH and conductivity values are very similar to those obtained with flow thought test,
although the methods cannot be compared. Conductivity ranges from 15 to 9 mS/m in a decreasing
trend line and pH ranges from 8 to 10. Ladle samples present higher and more stable pH values along
the several cycles (between 11.5 and 12.5) and conductivities between 130 mS/m and 450 mS/m.

Extraction levels are, as expected, higher than flow-through test, since DIN 38414-S4 is a forced
extraction method. Nevertheless, in what concerns leachability by water, these slag samples present
very low levels, ranging from 0.003 % (Fe) to 1.1 % (Ca) in EAF and 0.009 % (Mn) to 4.7 % (Ca) in
ladle.

The presence of other heavy metals such as arsenic, lead, cadmium, copper, mercury, nickel and zinc,
usually present in trace amounts, was also investigated in fresh samples. The concentrations obtained
are presented in Table 31.

The results presented in Table 31 were compared with the limit values defined by Council Decision
2003/33/EC, that establishes criteria and procedures for the acceptance of waste at landfills, defining
leaching limit values (LLV) for several heavy metals in wastes. Has observed, all values are below the
21


LLV for inert wastes, except for chromium in the ladle. It must be refereed however that these results
refer to fresh samples that have not been submitted to subsequent stabilization, i.e., unprocessed/fresh
slag.

The low solubilities verified in these trials are most probably related to mineral components bounded
within stable crystalline phases and also the leaching conditions, namely the basic pH. In fact, the pH
of a solution greatly affects the solubility of most minerals like iron, zinc, copper, manganese and
cadmium, which decreases for higher pH values [5].

Normally, chromium concentrations in the leachates tend to be low because ions are bound within
stable crystalline phases. However, there are certain types of soil (for instance, soils with high content
of oxidized manganese or other species that work as electron receptors) that promote oxidation of Cr
3+

to Cr
6+
. On the other hand, if leachates are put in contact with oxygenated waters with pH levels above
5, its probable the presence of Cr
6+
, because this specie is stable in such conditions [6, 7].

The trails carried out confirm the information consulted in several references, referring that except for
chromium (particularly in the hexavalent form), the leaching of other elements is not relevant in terms
of environmental impact.

The reuse of steel slags is highly recommended not only to increase the added value of this steel
industry by-product but also to save natural resources and, in some applications to have energy savings,
like cement industry. Low value applications, such as dumping, stockpiling or land filling must be
replaced by other valuable alternatives, which imply slag treatment and processing before its reuse.
Nowadays the pre-treatment of slag or its stabilization before reuse is, in fact, a common practice, due
to the growing demand of slags for high quality applications.

4 CONCLUSIONS
The investigations on slag have shown:

That the dielectric properties, expressed as a numeric value of the refractive index, for synthetic
as well as industrials slag could be derived within different frequency bands. The indexes
should be considered as band indexes rather than an index related to one single frequency.
That three different frequency bands 2-6, 4-8 and 6-10 showed the best results on liquid slag.
This knowledge could, dependent on the slag condition; rough surface and/or not dissolved slag
formers appearing in the ladle, be used to evaluate the thickness of the slag during production.
If the slag is too thin however the uncertainty in the measurement is high and the significance
of the value low.
The dielectric properties, refractive index, in combination with polarization could presently not
be used to identify the slag composition by correlation towards the slag basicity.
Slags geotechnical properties were investigated using BOF slag mixed with scrap, melted and
tapped during controlled cooling rate conditions. An average cooling rate between 15 to
20 C/min during the first 10 minutes directly after tapping provided a final product having the
properties required.
The investigation shows that an increase of the cooling rate damages the geotechnical
properties especially related to Micro-Deval and Los-Angeles characteristics.
During the first two months of ageing the free CaO content decreases for the individual slag
samples and with a more significant decrease for the sample with a low cooling rate
(corresponding to 13 cm slag thickness). However sample analysis after 4 and 7 months of
ageing shows a varying evolution pattern including both increasing and decreasing values of the
analysed free CaO content.
The free lime content behaviour could to some extent be explained by the chemical analysis
method using ethylene glycol as slag solvent, which can be distorted, since the dissolution of
Ca is not limited to free CaO but also to Ca from calcium hydroxide, calcium carbonate and
22


calcium combined with iron and manganese oxides. The use of XRD analysis confirms this but
it is presently not able to deliver quantitative evaluation of the free CaO, only responsible of the
subsequent slag expansion. Some complementary works on the XRD analysis are still necessary
to obtain a quantitative determination of the free lime content in the converter slag.
The cumulative leaching values of heavy metals obtained after more than one year of trials, are
very low, which is positive in terms of the environmental compatibility of these materials. The
trials performed with DIN 38414-S4 batch test are particularly optimistic since this is a worst
case scenario method i.e. it can exaggerate the leaching potential.

The results obtained also confirm that the natural leaching of soluble components from slags is
not a continuous process but an intermittent one. Leaching increased along slags ageing period
and, generally, no stabilization occurred after 30 weeks, meaning that it is in fact a slow process
that can take months, even ears to complete.

Other factors rather than metal concentrations have to be considered when accessing the
environmental impact of slags, such as accumulation effects resulting from long exposing
periods, the solubility and mobility of chemical species, their biological availability and the
properties of solid matrix (porosity, permeability).

Therefore, the reuse of slags without a proper previous treatment or the uncontrolled dumping
can have negative effects on the environment in the long term. Therefore, the pre-treatment of
slag or its stabilization before reuse is recommended, even for low value applications.

5 EXPLOITATION AND IMPACT OF THE RESEARCH RESULTS
5.1 MEFOS
Presently there are no installations made where the microwave technology is used for slag analysis.
Such an installation requires more dielectric data regarding the different slag qualities produced in a
steel plant than what has been investigated in this project. The industrial trials performed showed that it
is possible to use the technology to measure the slag thickness in a ladle though the result was very
much dependent on the slag condition; rough surface and/or not dissolved slag formers appearing in the
ladle. Therefore, before the technology could be used on a production scale these problems must be
addressed and solved.

As mentioned previously it was not possible to explicit determine the slag compositions using
microwave technology during the duration of the project. However, the observations shown in Figures
14, Figure 15 as well as Figure 16, were the refractive index is plotted against different definitions of
the slag basicity, illustrates an increase of the refractive index with an increase of the slag basicity.
Observations can also be seen in Figures 17-19. This implies that the results from the microwave
measurements could be used in the definition of a "microwave basicity" similar to the concept of
"optical basicity".

Optical basicity were used extensively during the 70's and early 80's to classify different slag properties
e.g. sulphide capacity or phosphorus capacity. One big drawback of this approach was that the optical
basicity was calculated from the slag composition and therefore did not prove a method for direct slag
classification.

By further develop the concept of microwave basicity as a direct classification method of different slag
properties instead of trying to directly determine the slag composition could be most successful. This
could improve:

process control
process economy
reduce the slag amount in different process
simplify the treatment of slag as a waste product
23


simplify the use of slag as a product that could be used for other applications, e.g. controlling
the free lime in slag

It needs to be pointed out that in order to achieve this further development and refinement of the
concept is necessarily.

A conference paper Preliminary microwave measurements on liquid slag presented at Molten Slags
Fluxes & Salts, No VI, Cape Town, The Republic of South Africa, 25-28 January 2004 contains some
information gained in the microwave part of the project. This paper will also be published in Iron
making and Steelmaking", 2005 vol.32 No1.

The principals of MEFOS will be informed about the results accomplished in the project through
internal meetings.

Information regarding the results of the project will also be future distributed at conferences and/or in
trade papers however not yet decided where and when.

5.2 ARCELOR
The exploitation of the results obtained requires in a first step the knowledge of the current liquid slag
cooling rate in the industrial slag pits and in a second step to use this knowledge to define or adapt the
current tapping process to obtain the cooling rate required for road applications use.

Concerning the improvement of the geotechnical properties of the slag in relation with its cooling rate,
the results are restricted to pilot test and not easy to transfer to an industrial scale especially the control
of the slag cooling rate.

Only the result obtained about the free CaO characterisation will be used by some steel shops to
characterise their converter slag and define the solid slag treatment in regard of its valorisation.

No publications or patent have been done/proposed.

5.3 ISQ
ISQ is considering making a scientific publication about the performed work in CHEMPOR 2005, the
Portuguese Annual Conference on Chemical Engineering, to be held in September.

Regarding dissemination of the results a demonstration of the technology used during the project will be
held at SN on 6
th
April. To attend the demonstration, ISQ and SN will invite people from the other steel
plant (theres only two in Portugal), associations connected to the metallurgical and foundry industries,
universities and other R&D enterprises.

Knowing or predicting steel slags leaching behavior along their ageing period is relevant information to
define the adequate treatment prior to their reuse in a specific application. Therefore this information is
useful not only for steel producers but also for slag processing units and industries using slags as raw
material, such as the cement industry.

The processing and reuse of steel slags brings anything but positive impacts to socio-economy, such as
the reduction of demand for virgin materials thus saving natural resources and the creation of direct and
indirect employment. The use of slag in some applications (such as cement making) captures all of the
energy that has been invested into the slag.

Many standards are available to perform short and long term leaching tests. In order to harmonize this
procedure in the European Union, standard EN 12457:2002 (referring to waste leaching behavior tests)
has been published by CEN. Other standards are presently under development referred to the simulation
of the leaching behavior of a waste material under specific conditions.

24


6 LIST OF REFERENCES
[1] RADIO WAVE INTERFEROMETRY FOR BOS SLAG CONTROL
Contract No 7210-CB/905 reference No 96-C1.04 a,b,c.
[2] Improvement of the EAF performances through an optimization of the foaming slag practice
7210-CB/903 (95-C1.04).
[3] Instituto de Meteorologia Climatologia, Portugal, www.meteo.pt, last consulted in 11/08/04.
[4] Anigstein, R. et all, Potential Recycling of Scrap Metal from Nuclear Facilities, Part I,
Volume 1, Unites States Environmental Protection Agency, September, 2001.
[5] Miller, Raymond W., Donahue, Roy L., Soils in Our Environment, Prentice Hall, 7
th
edition,
1995.
[6] European Commission, Summary report on RTD on iron and steel slag: Development and
perspectives, Directorate-General, Science, Research and Development, 1999.
[7] Motz H., Production and use of air-cooled blastfurnace and steel slags, Proceedings of the
3
rd
European Slag Conference, EUROSLAG, July 2003.
25


APPENDICES: SCIENTIFIC AND TECHNICAL DESCRIPTION OF THE RESULTS
MEFOS

Within the project MEFOS has used microwave technology to perform trials on solid and liquid,
synthetic as well as production, slag with the objective to evaluate the dielectric properties, i.e. the
dielectric constant of the slag. The dielectric data accomplished were recalculated and thereafter
expressed as the refractive index for each frequency band and slag tested. Trial performance and data
accomplished are described in chapter 3.

From data it can be seen that three frequency band provides better results and these band were therefore
used in the subsequent trials used to find a correlation between the refractive index and the slag
composition. An indication of this correlation can be seen however not explicit and with a low
statistical significance. In future investigations of slag structures the data accomplished in this project
will serve as a data base.

ARCELOR RESEARCH

With the objective to investigate the properties of slag to be used for geotechnical applications trials
have been performed. Slag samples have been prepared by re-melting and preparing BOF slag the 6 ton
pilot EAF. The liquid samples were tapped into moulds of different volume and two different slag
basicities and two slag tapping temperatures have been characterized.

From trial data it can bee seen that An average cooling rate between 15 to 20 C/min during the first 10
minutes following directly after tapping provided a final product having the properties required for road
applications. Furthermore, the investigation shows that an increase of the cooling rate damages the
geotechnical properties especially related to Micro-Deval and Los-Angeles characteristics.

Ageing trials shows that during the 2 first months of ageing, one can observe a free CaO content
decrease depending of the different slag samples. It seems that the free lime content decrease is more
significant for the slag samples cooled down with a low cooling rate (corresponding to 13 cm slag
thickness). This decrease has not been confirmed by the analyses made during the rest of the ageing
period after 4 and 7 months.

ISQ

The work performed includes Chemical characterization of slags produced from experimental trials
(supporting laboratory) by performing analysis with traditional equipment such as ICP/AES
spectrometry in order to confirm/validate the data achieved with the microwave system. Furthermore,
leaching tests have been conducted on several industrial slag samples, using lysimeter techniques.

The cumulative leaching values of heavy metals obtained after more than one year of trials, are very
low, which is positive in terms of the environmental compatibility of these materials. The trials
performed with DIN batch test are particularly optimistic since this is a worst case scenario method
i.e. it can exaggerate the leaching potential.

The results obtained also confirm that the natural leaching of soluble components from slags is not a
continuous process but an intermittent one. Leaching increased along slags ageing period and,
generally, no stabilization occurred after 30 weeks, meaning that it is in fact a slow process that can
take months, even ears to complete.

Other factors rather than metal concentrations have to be considered when accessing the environmental
impact of slags, such as accumulation effects resulting from long exposing periods, the solubility and
mobility of chemical species, their biological availability and the properties of solid matrix (porosity,
permeability). Therefore, the reuse of slags without a proper previous treatment or the uncontrolled
dumping can have negative effects on the environment in the long term. Therefore, the pre-treatment of
slag or its stabilization before reuse is recommended, even for low value applications.
26


LIST OF FIGURES
Figure 1 Experimental set up, Antenna and slag sample
Figure 2 Instrumentation
Figure 3 Results of the refractive index dependency of the frequency for low alloy
steel slags.
Figure 4 Results of the refractive index dependency of the frequency for high alloy
steel slags
Figure 5 The refractive index for all solid slags as a function of slag basicity defined
as (CaO+MgO)/(Al
2
O
3
+SiO
2
).
Figure 6 The refractive index for the slags from low alloyed steel as a function of slag basicity
defined as (CaO+MgO)/(Al2O3+SiO2).
Figure 7 Furnace, schematic
Figure 8 Antennas suspended above HF induction furnace
Figure 9 Experimental set-up, schematic
Figure 10 Chemical composition of the synthetic slag during the experiments
Figure 11 Results of the refractive index dependency of the frequency for
synthetic slag 7
Figure 12 Results of the refractive index dependency of the frequency for
ladle slag
Figure 13 Results of the refractive index dependency of the frequency for all
investigated slags
Figure 14 The refractive index as a function of the slag basicity defined as CaO/SiO2.
Figure 15 The refractive index as a function of the slag basicity defined as
(CaO+MgO)/SiO2.
Figure 16 The refractive index as a function of the slag basicity defined as
(CaO+MgO)/(Al2O3+SiO2).
Figure 17 Refractive index versus B2, CaO/SiO
2
ratio for different frequency bands
Figure 18 Refractive index versus B3, (CaO+MgO)/SiO
2
ratio for different
frequency bands
Figure 19 Refractive index versus B4, (CaO+MgO)/(Al
2
O
3
+SiO
2
) ratio for different
frequency bands
Figure 20 Comparison of the refractive index for solid and liquid slags.
Figure 21 Antenna encapsulation
Figure 22 Antenna encapsulation above ladle
Figure 23 The 6 t. pilot electric arc furnace equipped with the chute allowing to tap the liquid slag
in the different moulds
Figure 24 The different moulds for liquid slag tapping with specific equipment designed for slag
temperature measurement during its cooling.
Figure 25 The specific equipment with thermocouple used for slag temperature measurement during
its cooling.
Figure 26 Slag casting in the mould from the electric arc furnace
Figure 27 Mould equipped with temperature measurement during slag cooling
Figure 28 Slag cooling temperature evolution during trial with slag 1
Figure 29 Slag cooling temperature evolution during trial with slag 2
Figure 30 Slag cooling temperature during trial with slag 3
Figure 31 Pockets gas formed during the slag cooling in the mould of 13 cm thickness with slag 2
Figure 32 Slag cooling temperature evolution of the different types of slag for a slag layer of 3 cm
Figure 33 Slag cooling temperature evolution of the different types of slag for a slag layer of 7 cm
Figure 34 Slag cooling temperature evolution of the different types of slag for a slag layer of
13 cm
Figure 35 Average slag cooling temperature evolution versus the thickness of the slag layer
Figure 36 Evolution of the fines proportion during the grinding of the slag 1
Figure 37 Evolution of the fines proportion during the grinding of the slag 2
Figure 38 Evolution of the fines proportion during the grinding of the slag 3
Figure 39 Free lime content evolution in the different slag samples during 7 months ageing
Figure 40 Experimental set for flow-through test.
27


Figure 41 Phenomena occurred during trials (channeling, system clogging, biological growth and
oxidation of metallic iron inclusions).
Figure 42 EAF slag leachates evolution in flow through test
Figure 43 Ladle slag leachates evolution in flow through test
Figure 44 Dynamic extraction method DIN 38414
Figure 45 EAF slag leachates obtained with DIN 38414 S4 (cumulative values)
Figure 46 Ladle slag leachates obtained with DIN 38414 S4 (cumulative values)

LIST OF TABLES
Table 1 Industrial slag composition investigated during the solid slag measurements
Table 2 Refractive index observed for the different solid slags at different frequency
bands (rounded values in the table)
Table 3 Furnace data
Table 4 Initial synthetic slag mixtures
Table 5 Investigated synthetic slags
Table 6 Chemical analysis of investigated liquid slags before and after trial
Table 7 Refractive index observed for the different liquid slags at different frequency bands
(rounded values in the table)
Table 8 Chemical analysis of liquid slag samples
Table 9 Measured refractive index of solid slag
Table 10 Measured refractive index of liquid slag
Table 11 Measured refractive index of liquid ARCELOR slag
Table 12 Calculated slag thicknesses, SSAB slag
Table 13 Composition of the three tested slags, sampled in the electric arc furnace just before
the slag casting
Table 14 Conditions of the casting and slag characteristics
Table 15 The different average slag cooling kinetics (C/min) after slag tapping.
Table 16 Evolution of the slag sample composition of slag 1 during 7 months ageing
Table 17 Evolution of the slag sample composition of slag 2 during 7 months ageing
Table 18 Evolution of the slag sample composition of slag 3 during 7 months ageing
Table 19 Evolution of the slag sample composition of slag 1 during 7 months ageing
Table 20 Evolution of the slag sample composition of slag 2 during 7 months ageing
Table 21 Evolution of the slag sample composition of slag 3 during 7 months ageing
Table 22 Results of the XRD analysis of the slag samples at the beginning of the ageing
period
Table 23 Results of Los Angeles and Micro Deval tests performed on the slag n1
Table 24 Los Angeles and Micro Deval coefficient for different material including standard LD slag
Table 25 Analytical results of slag samples sent in 2002
Table 26 Analytical results of slag samples sent in 2002 (continued)
Table 27 Analytical results of BOF and Ladle slag samples sent in May 2003
Table 28 Actual state-of-art of steel slags final destinies
Table 29 Chemical composition of EAF slag samples submitted to leaching tests at ISQ (% wt)
Table 30 Chemical composition of ladle slag samples submitted to leaching tests at ISQ (% wt)
Table 31 Trace heavy metal concentrations in fresh slags samples
28


TABLES

Table 1 Industrial slag composition investigated during the solid slag measurements

Slag quality Slag analysis wt %
Al
2
O
3
CaO Cr
2
O
3
FeO MgO MnO SiO
2
TiO
2

LD converter 4.69 38.93 n/a 1.85 18.24 1.53 32.34 1.60
Ladle low alloy steel 20.78 33.64 n/a 0.95 18.42 0.58 24.16 1.01
Blast Furnace 13.09 33.76 n/a 0.49 15.96 0.48 32.88 2.05
AOD converter 1 1.4 51 0.5 0.5 7.7 0.3 30 0.9
EAF 1 5 44 10 0.7 8.8 1.1 25 2.1
AOD converter 2 1.4 56.7 0.5 0.3 5.5 0.3 30.7 0.39
EAF 2 6.3 47.9 2 0.8 5.6 2 32.4 1.37

Table 2 Refractive index observed for the different solid slags at different frequency
bands (rounded values in the table)

Slag origin Frequency bands (GHz)
2-6 4-8 6-10 8-12 10-14 12-16 14-18
Refractive index
LD converter 2.50 1.98 - 2.20 2.49 2.54 1.98
Ladle low alloy steel 2.25 1.91 1.95 2.13 2.51 2.78 2.14
Blast Furnace 1.99 1.75 2.14 2.35 1.81 2.10 2.68
AOD converter 1 2.08 2.10 2.12 2.09 2.10 2.08 2.08
EAF 1 1.91 1.93 1.90 1.81 1.93 1.92 1.80
AOD converter 2 1.65 1.81 1.72 1.67 1.75 1.72 1.64
EAF 2 2.29 2.21 1.95 1.50 1.38 1.63 1.70

Table 3 Furnace data

Nominal Capacity 150 kg
Power 90 kW
Frequency 2200 Hz
Inner Diameter 0.28 m
Depth 0.56 m

Table 4 Initial synthetic slag mixtures

Mol. Fraction Wt (%) Weight (Kg)
No xAl
2
O
3
xCaO XSiO
2
%Al
2
O
3
wt%CaO %SiO
2
kgAl
2
O
3
kgCaO kgSiO
2

1 0.00 0.35 0.65 0.00 33.45 66.55 0.00 2.41 4.79
2 0.00 0.55 0.45 0.00 53.29 46.71 0.00 3.84 3.36
3 0.25 0.75 0.00 37.74 62.26 0.00 2.72 4.48 0.00
4 0.45 0.55 0.00 59.80 40.20 0.00 4.31 2.89 0.00
5 0.15 0.20 0.65 23.33 17.11 59.57 1.68 1.23 4.29
6 0.20 0.35 0.45 30.41 29.27 40.32 2.19 2.11 2.90
7 0.35 0.35 0.30 48.66 26.76 24.58 3.50 1.93 1.77
8 0.15 0.55 0.30 23.84 48.07 28.09 1.72 3.46 2.02
9 0.35 0.55 0.10 49.20 42.52 8.28 3.54 3.06 0.60
10 0.20 0.70 0.10 31.06 59.79 9.15 2.24 4.30 0.66

29


Table 5 Investigated synthetic slags
Slag Mol. Fraction Wt (%) Weight (Kg)
No xAl
2
O
3
xCaO xSiO
2
%Al
2
O
3
%CaO %SiO
2
kgAl
2
O
3
kgCaO kgSiO
2

2 0.00 0.55 0.45 0.00 53.29 46.71 0.00 2.66 2.34
3 0.25 0.75 0.00 37.74 62.26 0.00 1.89 3.11 0.00
4 0.45 0.55 0.00 59.80 40.20 0.00 2.99 2.01 0.00
6 0.20 0.35 0.45 30.41 29.27 40.32 1.52 1.46 2.02
7 0.35 0.35 0.30 48.66 26.76 24.58 2.43 1.34 1.23
8 0.15 0.55 0.30 23.84 48.07 28.09 1.19 2.40 1.40

Table 6 Chemical analysis of investigated liquid slags before and after trial

Charge protocol Slag analysis wt %
Sample No slag type Al
2
O
3
CaO FeO MgO MnO SiO
2

L.406-1 synthetic slag 2 18.49 35.31 3.00 0.4 1.3 40.09
L.406-2 32.66 24.42 2.78 0.37 1.34 36.63
L.407-1 synthetic slag 7 47.36 19.59 2.73 0.18 1.41 27.84
L.407-2 48.26 18.9 2.64 0.16 1.73 27.68
L.408-1 synthetic slag 6 35.53 21.67 2.12 0.19 1.81 38
L.408-2 36.57 20.02 3.15 0.16 2.2 36.57
L.409-1 synthetic slag 4 56.7 29 1.43 0.3 0.78 10.79
L.409-2 58.8 26.22 1.26 0.22 0.75 11.71
L.410-1 synthetic slag 8 20.89 35.13 1.79 6.94 1.74 32.88
L.410-2 20.9 34.35 1.27 9.73 1.43 31.57
L.411-1 synthetic slag 3 30.78 46.81 0.60 11.27 0.17 9.82
L.411-2 29.5 44.07 0.86 14.93 0.33 9.93
L.412-1 EAF slag 9.99 29.08 8.98 13.05 8.83 28.04
L.412-2 8.97 29.43 8.08 13.72 8.92 28.72
L.413-1 LD slag 4.45 39.43 2.34 17.4 1.84 31.87
L.413-2 4.92 38.42 1.36 19.08 1.22 32.8
L.414-1 Ladle slag 21.09 33.98 1.11 17.75 0.68 23.92
L.414-2 20.47 33.29 0.80 19.08 0.47 24.39

30


Table 7 Refractive index observed for the different liquid slags at different frequency bands (rounded
values in the table)
Slag quality Frequency step(GHz)
2-6 4-8 6-10 8-12 10-14 12-16 14-18
Refractive index
Synthetic slag 2 NA NA NA NA NA NA NA
Synthetic slag 7 1.61 2.06 2.15 1.74 2.25** 1.81** 1.81**
Synthetic slag 6 1.30 1.41 1.50 NA 2.04** 2.41** 2.13**
Synthetic slag 4 1.56 2.24* 2.52** NA 1.88* NA NA
Synthetic slag 8 2.19** 1.95 2.02 2.82** NA NA 3.88**
Synthetic slag 3 3.18* NA 2.69* 3.72* 4.20* NA NA
EAF slag 1.77** 1.80 2.45 NA NA NA NA
LD slag 3.57 2.40 2.65 NA NA NA NA
Ladle slag 1.53 1.39** 1.92 1.55 1.60 1.57 2.05**
Accepted S/N>7, *Probable 3<S/N<7, **Ambiguous S/N<3, NA= Not Aviable

Table 8 Chemical analysis of liquid slag samples
Slag sample no
1 2 3 4 5 6 7 9 101 102 103
(%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%)
SiO
2
66.1 54.3 72.9 49.0 72.9 49.0 38.6 28.6 53.3 44.2 40.1
Al
2
O
3
6.0 3.9 13.5 23.2 13.5 23.2 36.5 32.6 6.2 25.2 19.1
CaO 26.5 40.9 12.9 26.9 12.9 26.9 23.8 37.9 24.7 18.6 24.9
K2O 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
MgO 1.1 0.8 0.7 0.5 0.6 0.8 0.9 0.8 15.6 11.8 15.8
MnO 0.0 0.0 0.1 0.2 0.1 0.0 0.0 0.0 0.1 0.1 0.1
Na
2
O 0.0 0.0 0.1 0.1 0.1 0.1 0.2 0.1 0.0 0.1 0.1
P
2
O
5
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
TiO
2
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Sum 100 100 100 100 100 100 100 100 100 100 100

B2 0.40 0.75 0.38 0.46 0.18 0.55 0.62 1.32 0.46 0.42 0.62
B3 0.42 0.77 0.39 0.47 0.18 0.56 0.64 1.35 0.76 0.69 1.01
B4 0.38 0.72 0.29 0.33 0.16 0.38 0.33 0.63 0.68 0.44 0.69
B2 = CaO% to SiO
2
% ratio, B3 = (CaO% + MgO%) to SiO
2
% ratio
B4 = (CaO% + MgO%) to (Al
2
O
3
% + SiO
2
%) ratio

31


Table 9 Measured refractive index of solid slag
Frequency band
Slag no
2-6 GHz 4-8 GHz 6-10 GHz 8-12 GHz 10-14 GHz 12-16 GHz 14-18 GHz
Refractive index (n)
1 1.70 1.59 1.59 1.59 1.71 1.74 NA
2 1.68 1.56 1.68 1.57 1.62 1.65 1.45
3 2.20 2.07 2.07 2.12 2.10 1.78 1.72
4 2.25 1.91 2.01 1.78 1.72 2.57 NA
5 1.72 1.73 1.87 1.65 o o NA
6 2.18 2.21 2.12 2.12 1.98 2.22 2.26
7 2.09 2.31 2.11 1.96 2.10 2.16 2.07
8 1.79 1.68 1.79 1.79 1.68 1.62 NA
9 2.04 2.25 2.13 2.07 2.02 2.13 2.02
10 2.17 2.20 2.09 2.15 2.24 1.96 1.59
101 2.60 2.18 2.30 1.98 2.33 NA NA
102 2.54 1.87 2.08 2.21 2.04 1.69 2.06
103 NA 1.90 1.94 1.91 2.09 NA NA
o = Ambiguous determination NA = Not Available

Table 10 Measured refractive index of liquid slag
Slag no Frequency band
2-6 GHz 4-8 GHz 6-10 GHz 8-12 GHz 10-14 GHz 12-16 GHz 14-18 GHz
Refractive index (n)
1 3.82 3.76 3.84 3.84 3.71 NA 3.68
2 2.11 2.91 2.60 NA 2.88 NA NA
3 3.05 NA NA NA NA 1.75 NA
4 NA NA 1.40 2.54 2.08 NA NA
5 1.91 1.92 1.98 1.91 2.37 NA NA
6 NA 2.56 NA 2.63 3.05 NA NA
7 NA 2.14 NA NA NA NA NA
9 NA 2.24 2.20 2.27 2.34 2.02 NA
101 2.79 2.96 3.31 2.54 2.43 3.39 NA
102 2.33 NA NA NA NA NA NA
103 3.83 3.76 3.84 3.84 3.71 NA 3.68
NA = Not Available

Table 11 Measured refractive index of liquid ARCELOR slag


Trials Mould no Frequency band
2-6 GHz 4-8 GHz 6-10 GHz
Refractive index (n)
Slag 1 3 2.25 3.69 3.63
Slag 3 1 3.05 3.85 1.77
32


Table 12 Calculated slag thicknesses, SSAB slag
Signal strength Slag level Signal strength Steel level Slag thickness
( S/N>7 ) ( mm ) ( S/N>7 ) ( mm ) ( mm )
8 3272 7 3287 15
8 3272 7 3287 15
9 3258 9 3279 21
10 3221 9 3255 34
11 3176 11 3186 10
9 3328 10 3337 9
20 3308 23 3333 25
15 3212 14 3230 18
10 3193 10 3205 12
24 3189 23 3189 0
26 3209 25 3231 22
8 3235 8 3252 17
23 3324 22 3332 8
7 3262 7 3285 23
18 3290 21 3321 31
18 3294 15 3317 23
32 3239 31 3239 0
12 3458 10 3498 40
11 3195 11 3195 0
16 3204 16 3208 4
7 3394 7 3394 0
25 3415 26 3452 37

Table 13 Composition of the three tested slags, sampled in the electric arc furnace just before the slag
casting



Trials CaO/SiO2
Fe
%
SiO
2
%
CaO
%
Al
2
O
3
%
TiO
2
%
MgO
%
P
2
O
5
%
MnO
%
Cr
2
O
3
%
Slag 1 3.3 15.0 14.2 47.0 3.54 1.01 9.05 2.07 3.75 0.31
Slag 2 3.5 25.1 11.6 40.2 2.28 0.81 6.69 1.99 3.51 0.36
Salg 3 4.2 26.4 9.85 41.3 2.01 0.70 7.3 1.60 2.58 0.29
33


Table 14 Conditions of the casting and slag characteristics

Trials Moulds
Slag thickness
(cm)
CaO/SiO2
Casting
temperature
(C)
Liquid phase
(%)
Nature of solid
phase
1 3 3.3 1690 99.6 MgO
2 7 3.3 1670 98.5 MgO
3 13 3.3 1670 98.5 MgO
1 3 3.5 1639 100 -
2 7 3.5 1630 100 -
3 13 3.5 1620 100 -
1 3 4.2 1670 100 -
2 7 4.2 1665 100 -
3 13 4.2 1660 100 -
Slag 3
Slag 2
Slag 1


Table 15 The different average slag cooling kinetics (C/min) after slag tapping

Cooling period
Slag thickness
3 cm
Slag thickness
7 cm
Slag thickness
13 cm
0 to 2.5 min 417 - -
0 to 5 min 228 121 22
0 to 10 min 132 77 17
0 to 20 min 70 44 13


34


Table 16 Evolution of the slag sample composition of slag 1 during 7 months ageing

Slag 1
Fe
(%)
SiO
2

(%)
CaO
(%)
MgO
(%)
Al
2
O
3

(%)
TiO
2

(%)
P
2
O
5

(%)
MnO
(%)
Cr
2
O
3

(%)
Slag
thickness
Initial liquid
slag
15.0 14.21 47.0 9.05 3.54 1.01 2.07 3.75 0.31
Initial crushed
slag
15.4 14.12 46.6 10.21 3.40 1.01 2.07 3.75 0.31
After 2
months
ageing
15.2 13.83 46.3 9.77 3.58 1.05 2.01 3.76 0.31
After 4
months
ageing
15.7 13.82 46.0 10.13 3.32 1.10 2.03 3.79 0.34
3 cm
After 7
months
ageing
14.9 14.28 46.3 10.53 3.34 0.97 2.04 3.81 0.36
Initial crushed
slag
17.6 13.17 44.7 9.74 3.35 1.00 2.11 3.97 0.34
After 2
months
ageing
17.3 13.11 44.4 9.57 3.29 1.02 2.10 3.97 0.36
After 4
months
ageing
17.4 12.96 43.9 9.86 3.23 1.08 2.07 4.01 0.38
7 cm
After 7
months
ageing
17.4 13.70 44.9 9.52 3.46 0.96 2.12 3.98 0.40
Initial crushed
slag
17.9 12.94 44.4 9.45 3.46 1.05 2.07 4.06 0.36
After 2
months
ageing
17.9 12.91 43.9 9.16 3.45 1.01 2.04 4.15 0.36
After 4
months
ageing
17.5 13.20 44.0 9.73 3.16 1.07 2.15 4.05 0.39
13 cm
After 7
months
ageing
17.9 13.55 44.1 9.89 3.47 0.99 2.10 4.03 0.40


35


Table 17 Evolution of the slag sample composition of slag 2 during 7 months ageing

Slag 2
Fe
(%)
SiO
2

(%)
CaO
(%)
MgO
(%)
Al
2
O
3

(%)
TiO
2

(%)
P
2
O
5

(%)
MnO
(%)
Cr
2
O
3

(%)
Slag
thickness
Initial liquid
slag
25.1 11.57 40.2 7.70 2.28 0.81 1.99 3.51 0.36
Initial
crushed slag
24.6 11.74 40.1 8.00 2.24 0.82 1.97 3.52 0.37
After 2
months
ageing
24.8 11.59 39.5 7.71 2.39 0.85 1.94 3.51 0.37
After 4
months
ageing
24.5 11.87 39.7 8.71 2.21 0.85 2.04 3.50 0.40
3 cm
After 7
months
ageing
11.79 40.1 8.36 2.32 0.81 1.96 3.52 0.41
Initial
crushed slag
24.9 11.57 40.0 7.82 2.28 0.85 1.96 3.53 0.37
After 2
months
ageing
24.7 11.55 39.7 7.59 2.33 0.86 1.96 3.51 0.38
After 4
months
ageing
27.9 11.60 39.9 7.92 2.23 0.85 1.99 3.52 0.41
7 cm
After 7
months
ageing
24.4 12.27 40.1 8.43 2.30 0.78 2.01 3.51 0.42
Initial
crushed slag
27.1 10.95 38.1 7.24 2.27 0.81 1.87 3.52 0.38
After 2
months
ageing
26.4 11.18 38.0 7.91 2.10 0.81 1.90 3.52 0.39
After 4
months
ageing
26.8 11.13 38.0 7.37 2.09 0.84 1.93 3.55 0.43
13 cm
After 7
months
ageing
26.3 11.63 38.8 7.55 2.14 0.73 1.97 3.52 0.44


36


Table 18 Evolution of the slag sample composition of slag 3 during 7 months ageing

Slag 3
Fe
(%)
SiO
2

(%)
CaO
(%)
MgO
(%)
Al
2
O
3

(%)
TiO
2

(%)
P
2
O
5

(%)
MnO
(%)
Cr
2
O
3

(%)
Slag
thickness
Initial liquid
slag
26.4 9.85 41.3 7.35 2.01 0.70 1.60 2.58 0.29
Initial
crushed slag
26.6 9.85 41.2 8.06 2.04 0.73 1.57 2.55 0.30
After 2
months
ageing
26.3 10.01 40.8 8.19 1.96 0.73 1.61 2.59 0.30
After 4
months
ageing
26.1 9.94 40.7 8.26 1.98 0.79 1.60 2.61 0.36
3 cm
After 7
months
ageing
26.3 10.05 41.2 8.15 2.08 0.73 1.55 2.58 0.34
Initial
crushed slag
26.5 9.95 41.5 7.66 2.06 0.74 1.58 2.51 0.30
After 2
months
ageing
26.3 9.63 40.5 8.23 2.03 0.75 1.52 2.52 0.31
After 4
months
ageing
23.5 9.55 40.5 8.76 1.98 0.80 1.53 2.58 0.34
7 cm
After 7
months
ageing
25.9 10.04 41.1 8.65 2.15 0.70 1.55 2.58 0.34
Initial
crushed slag
27.3 9.78 41.8 6.34 2.16 0.77 1.51 2.54 0.28
After 2
months
ageing
26.0 9.87 39.7 9.94 1.73 0.67 1.55 2.61 0.31
After 4
months
ageing
26.6 9.74 40.9 8.02 1.97 0.78 1.55 2.56 0.35
13 cm
After 7
months
ageing
26.3 10.72 41.4 8.36 2.05 0.70 1.58 2.56 0.33


37


Table 19 Evolution of the slag sample composition of slag 1 during 7 months ageing

Slag
thickness
SLAG 1
Fe total
(%)
Fe metal
(%)
Fe
2+
(%)
Fe
3+
/
(Fe
2+
+Fe
3+
)
(%)
Free lime
(%)
Initial crushed slag 15.40 1.20 10.00 0.30 6.10
After 2 months ageing 15.25 1.85 10.45 0.22 5.80
After 4 months ageing 15.75 2.10 12.80 0.06 5.24
3 cm
After 7 months ageing 14.90 0.90 9.95 0.29 4.70
Initial crushed slag 17.60 0.95 12.05 0.28 5.65
After 2 months ageing 17.35 1.60 12.20 0.23 5.22
After 4 months ageing 17.45 3.00 11.50 0.20 2.51
7 cm
After 7 months ageing 17.40 0.90 12.00 0.27 4.89
Initial crushed slag 17.95 1.30 16.40 0.01 5.50
After 2 months ageing 17.95 1.65 13.15 0.19 3.28
After 4 months ageing 17.55 2.00 12.75 0.18 4.51
13 cm
After 7 months ageing 17.90 0.85 13.10 0.23 4.94


Table 20 Evolution of the slag sample composition of slag 2 during 7 months ageing

Slag
thickness
SLAG 2
Fe total
(%)
Fe metal
(%)
Fe
2+
(%)
Fe
3+
/
(Fe
2+
+Fe
3+
)
(%)
Free lime
(%)
Initial crushed slag 24.60 1.10 17.60 0.25 3.50
After 2 months ageing 24.80 2.25 17.70 0.22 3.21
After 4 months ageing 24.50 2.40 17.40 0.21 2.87
3 cm
After 7 months ageing 1.05 17.80 3.56
Initial crushed slag 24.95 1.15 17.70 0.26 3.85
After 2 months ageing 24.70 2.20 17.80 0.21 3.05
After 4 months ageing 27.90 2.20 18.00 0.30 2.20
7 cm
After 7 months ageing 24.40 0.95 17.70 0.25 3.33
Initial crushed slag 27.10 1.95 18.10 0.28 2.70
After 2 months ageing 26.45 2.70 19.95 0.18 1.47
After 4 months ageing 26.85 2.20 18.45 0.25 2.38
13 cm
After 7 months ageing 26.30 1.20 17.00 0.31 2.51




38


Table 21 Evolution of the slag sample composition of slag 3 during 7 months ageing

Slag
thickness
SLAG 3
Fe total
(%)
Fe metal
(%)
Fe
2+
(%)
Fe
3+
/
(Fe
2+
+Fe
3+
)
(%)
Free lime
(%)
Initial crushed slag 26.60 1.90 17.65 0.29 8.60
After 2 months ageing 26.30 3.35 18.30 0.20 7.78
After 4 months ageing 26.15 3.40 18.00 0.21 5.60
3 cm
After 7 months ageing 26.35 2.15 17.80 0.26 7.64
Initial crushed slag 26.50 2.35 11.95 0.51 8.50
After 2 months ageing 26.30 3.20 17.60 0.24 8.34
After 4 months ageing 23.50 3.80 18.10 0.08 7.85
7 cm
After 7 months ageing 25.90 1.75 18.35 0.24 3.93
Initial crushed slag 27.35 3.25 15.30 0.37 7.75
After 2 months ageing 26.00 2.95 18.60 0.19 4.75
After 4 months ageing 26.65 3.00 18.00 0.24 7.81
13 cm
After 7 months ageing 26.30 1.80 18.80 0.23 8.78




39


Table 22 Results of the XRD analysis of the slag samples at the beginning of the ageing period.

T
h
e

d
i
f
f
e
r
e
n
t
p
h
a
s
e
s
S
l
a
g
1
S
l
a
g
2
S
l
a
g
3
3

c
m
7

c
m
1
3

c
m
3

c
m
7

c
m
1
3

c
m
3

c
m
7

c
m
1
3

c
m
L
a
r
n
i
t
e
C
a
2
(
S
i
O
4
)
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
I
r
o
n
c
a
l
c
i
u
m
o
x
i
d
e
F
e
2
O
3
(
C
a
O
)
2
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
F
o
r
s
t
e
r
i
t
e
M
g
2
S
i
O
4
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
W
u
e
s
t
i
t
e
F
e
0
.
9
0
2
O
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
A
l
u
m
i
n
i
u
m

i
r
o
n

o
x
i
d
e
A
l
F
e
O
3
+
+
+
+
+
+
+
+
+
P
e
r
o
v
s
k
i
t
e
C
a
T
i
O
3
+
+
+
+
+
+
+
+
+
I
r
o
n

o
x
i
d
e
F
e
O
-
-
-
-
-
-
-
-
-
H
a
e
m
a
t
i
t
e
F
e
2
O
3
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
A
l
u
m
i
n
i
u
m

o
x
i
d
e
A
l
2
O
3
-
-
-
-
-
-
-
-
-
-
C
a
l
c
i
u
m

h
y
d
r
o
x
i
d
e
C
a
(
O
H
)
2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
C
a
l
c
i
u
m

i
r
o
n

o
x
i
d
e
C
a
F
e
O
3
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
C
a
l
c
i
u
m

i
r
o
n

o
x
i
d
e
C
a
F
e
O
2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
L
i
m
e
C
a
O
-
-
-
-
-
-
T
h
e

d
i
f
f
e
r
e
n
t
p
h
a
s
e
s
S
l
a
g
1
S
l
a
g
2
S
l
a
g
3
3

c
m
7

c
m
1
3

c
m
3

c
m
7

c
m
1
3

c
m
3

c
m
7

c
m
1
3

c
m
L
a
r
n
i
t
e
C
a
2
(
S
i
O
4
)
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
I
r
o
n
c
a
l
c
i
u
m
o
x
i
d
e
F
e
2
O
3
(
C
a
O
)
2
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
F
o
r
s
t
e
r
i
t
e
M
g
2
S
i
O
4
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
W
u
e
s
t
i
t
e
F
e
0
.
9
0
2
O
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
A
l
u
m
i
n
i
u
m

i
r
o
n

o
x
i
d
e
A
l
F
e
O
3
+
+
+
+
+
+
+
+
+
P
e
r
o
v
s
k
i
t
e
C
a
T
i
O
3
+
+
+
+
+
+
+
+
+
I
r
o
n

o
x
i
d
e
F
e
O
-
-
-
-
-
-
-
-
-
H
a
e
m
a
t
i
t
e
F
e
2
O
3
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
A
l
u
m
i
n
i
u
m

o
x
i
d
e
A
l
2
O
3
-
-
-
-
-
-
-
-
-
-
C
a
l
c
i
u
m

h
y
d
r
o
x
i
d
e
C
a
(
O
H
)
2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
C
a
l
c
i
u
m

i
r
o
n

o
x
i
d
e
C
a
F
e
O
3
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
C
a
l
c
i
u
m

i
r
o
n

o
x
i
d
e
C
a
F
e
O
2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
L
i
m
e
C
a
O
-
-
-
-
-
-
T
h
e

d
i
f
f
e
r
e
n
t
p
h
a
s
e
s
T
h
e

d
i
f
f
e
r
e
n
t
p
h
a
s
e
s
S
l
a
g
1
S
l
a
g
1
S
l
a
g
2
S
l
a
g
2
S
l
a
g
3
S
l
a
g
3
3

c
m
3

c
m
7

c
m
7

c
m
1
3

c
m
1
3

c
m
3

c
m
3

c
m
7

c
m
7

c
m
1
3

c
m
1
3

c
m
3

c
m
3

c
m
7

c
m
7

c
m
1
3

c
m
1
3

c
m
L
a
r
n
i
t
e
L
a
r
n
i
t
e
C
a
2
(
S
i
O
4
)
C
a
2
(
S
i
O
4
)
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
I
r
o
n
c
a
l
c
i
u
m
o
x
i
d
e
I
r
o
n
c
a
l
c
i
u
m
o
x
i
d
e
F
e
2
O
3
(
C
a
O
)
2
F
e
2
O
3
(
C
a
O
)
2
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
F
o
r
s
t
e
r
i
t
e
F
o
r
s
t
e
r
i
t
e
M
g
2
S
i
O
4
M
g
2
S
i
O
4
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
W
u
e
s
t
i
t
e
W
u
e
s
t
i
t
e
F
e
0
.
9
0
2
O
F
e
0
.
9
0
2
O
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
A
l
u
m
i
n
i
u
m

i
r
o
n

o
x
i
d
e
A
l
u
m
i
n
i
u
m

i
r
o
n

o
x
i
d
e
A
l
F
e
O
3
A
l
F
e
O
3
++
++
++
++
++
++
++
++
++
P
e
r
o
v
s
k
i
t
e
P
e
r
o
v
s
k
i
t
e
C
a
T
i
O
3
C
a
T
i
O
3
++
++
++
++
++
++
++
++
++
I
r
o
n

o
x
i
d
e
I
r
o
n

o
x
i
d
e
F
e
O
F
e
O
--
--
--
--
--
--
--
--
--
H
a
e
m
a
t
i
t
e
H
a
e
m
a
t
i
t
e
F
e
2
O
3
F
e
2
O
3
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
A
l
u
m
i
n
i
u
m

o
x
i
d
e
A
l
u
m
i
n
i
u
m

o
x
i
d
e
A
l
2
O
3
A
l
2
O
3
--
-
-
-
-
--
-
-
-
-
-
-
-
-
--
--
C
a
l
c
i
u
m

h
y
d
r
o
x
i
d
e
C
a
l
c
i
u
m

h
y
d
r
o
x
i
d
e
C
a
(
O
H
)
2
C
a
(
O
H
)
2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
C
a
l
c
i
u
m

i
r
o
n

o
x
i
d
e
C
a
l
c
i
u
m

i
r
o
n

o
x
i
d
e
C
a
F
e
O
3
C
a
F
e
O
3
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
C
a
l
c
i
u
m

i
r
o
n

o
x
i
d
e
C
a
l
c
i
u
m

i
r
o
n

o
x
i
d
e
C
a
F
e
O
2
C
a
F
e
O
2
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
L
i
m
e
L
i
m
e
C
a
O
C
a
O
-
-
-
-
-
-
-
-
-
-
-
-

40


Table 23 Results of Los Angeles and Micro Deval tests performed on the slag n1

Average cooling rate (C/min) Slag
thickness
for
cooling
(mm)
Until 5
min
Until 10
min
Until 20
min
Los
Angeles
Micro
Deval
Bulk
density
30 228 132 70 19.5 2.5
70 121 77 44 21.0 12.0 2.8
130 22 17 13 12.7 4.9 3.4


Table 24 Los Angeles and Micro Deval coefficient for different material including standard LD slag

Materials Los
Angeles
Micro
Deval
Quality
increase
BF slag 30 35
Limestone 25 25 - 30
Porphyry
(
*
)
15 6 10
Standard LD slag 12 6 8

(
*
)
usually used for upper road layer

41


Table 25 Analytical results of slag samples sent in 2002







Sample reference
Element
L 410 -1 L 410 -2 L 409 -1 L 409 -2 L 408 -1 L 408 -2 L 407 -1 L 407 -2 L 406 -1
FeO (%) 4.3 3.6 4.6 3.6 6.5 7.3 6.8 7.7 6.4
Std Dev 0.13 0.16 0.06 0.12 0.11 0.28 0.13 0.22 0.30
% CV (1) 3.1 4.4 1.4 3.2 1.7 3.8 1.9 2.9 4.6
MnO (%) 1.8 1.5 0.9 0.8 1.9 2.3 1.4 1.9 1.3
Std Dev 0.05 0.03 0.02 0.01 0.07 0.05 0.05 0.05 0.06
% CV (1) 2.6 2.0 2.2 1.7 3.7 2.0 3.3 2.6 4.5
TiO
2
(%)
0.2 0.2 0.5 0.5 0.5 0.6 0.2 0.3 0.7
Std Dev 0.02 0.01 0.01 0.01 0.01 0.02 0.01 0.01 0.03
% CV (1) 8.8 5.9 1.6 1.7 1.4 3.3 2.7 3.3 4.4
Al
2
O
3
(%)
20.2 19.5 56.3 46.8 34.7 35.7 31.5 37.1 18.8
Std Dev 0.49 0.49 0.83 0.86 0.45 1.14 0.35 0.55 0.67
% CV (1) 2.4 2.5 1.5 1.8 1.3 3.2 1.1 1.5 3.6
CaO (%) 39.4 37.8 31.1 28.1 25.5 22.6 22.5 21.9 39.0
Std Dev 1.07 0.72 0.39 0.47 0.37 0.68 0.43 0.34 1.68
% CV (1) 2.7 1.9 1.2 1.7 1.4 3.0 1.9 1.5 4.3
MgO (%) 7.7 10.9 0.3 0.2 0.7 0.3 0.2 0.2 0.3
Std Dev 0.23 0.26 0.003 0.01 0.01 0.01 0.003 0.01 0.02
% CV (1) 3.1 2.4 1.1 4.0 1.5 3.1 1.4 5.1 6.3
SiO
2
(%)
29.4 29.2 10.9 23.6 36.4 36.1 39.4 33.9 35.3
Std Dev 1.33 1.00 0.74 1.36 1.13 1.31 1.86 1.03 1.42
% CV (1) 4.5 3.4 6.8 5.7 3.1 3.6 4.7 3.1 4.0
42


Table 26 Analytical results of slag samples sent in 2002 (continued)

Sample reference
Element
L 406 -2 L 411 -1 L 411 -2 L 412 -1 L 412 -2 L 413 -1 L 413 -2 L 414 -1 L 414 -2
FeO (%) 6.7 3.2 3.5 11.9 10.3 5.6 3.4 4.2 3.3
Std Dev 0.08 0.28 0.13 0.56 0.42 0.18 0.10 0.06 0.11
% CV (1) 1.2 8.7 3.7 4.7 4.1 3.3 3.0 1.4 3.5
MnO (%) 1.5 0.3 0.2 8.5 8.8 1.9 0.9 0.8 0.5
Std Dev 0.04 0.03 0.01 0.31 0.55 0.07 0.03 0.02 0.02
% CV (1) 2.6 7.9 3.7 3.7 6.3 3.8 3.7 2.3 4.1
TiO
2
(%)
1.1 0.2 0.2 0.7 0.7 1.6 1.6 1.1 1.0
Std Dev 0.02 0.02 0.01 0.01 0.04 0.05 0.04 0.01 0.04
% CV (1) 2.2 6.7 5.6 2.0 5.5 3.4 2.8 0.9 4.1
Al
2
O
3
(%)
35.7 27.6 27.2 8.8 6.6 4.1 4.1 18.0 16.7
Std Dev 0.94 0.87 0.51 0.16 0.54 0.16 0.17 0.07 0.31
% CV (1) 2.6 3.1 1.9 1.8 8.2 3.8 4.1 0.4 1.9
CaO (%) 28.1 48.1 47.0 31.4 32.9 42.2 42.0 38.1 36.6
Std Dev 0.31 3.21 2.03 0.52 1.86 1.48 0.73 0.39 1.29
% CV (1) 1.1 6.7 4.3 1.7 5.7 3.5 1.7 1.0 3.5
MgO (%) 0.4 12.6 15.8 14.0 14.6 18.2 20.2 18.5 19.5
Std Dev 0.01 0.56 0.36 0.32 0.92 0.76 0.34 0.14 0.41
% CV (1) 2.2 4.4 2.3 2.2 6.3 4.2 1.7 0.7 2.1
SiO
2
(%)
29.4 10.8 10.7 26.8 29.7 32.7 34.4 27.6 25.8
Std Dev 1.28 0.60 0.78 0.66 1.23 1.76 1.08 0.47 1.03
% CV (1) 4.3 5.6 7.3 2.4 4.1 5.4 3.1 1.7 4.0




43


Table 27 Analytical results of BOF and Ladle slag samples sent in May 2003

BOF 1 BOF 2 LADLE 1 LADLE 2
FeO (%) 25.5 25.1 16.6 15.0
Std Dev 1.05 0.49 0.92 0.10
% CV (1) 4.1 2.0 5.6 0.7
MnO (%) 3.4 3.5 6.1 6.3
Std Dev 0.06 0.05 0.23 0.14
% CV (1) 1.7 1.4 3.8 2.2
TiO
2
(%)

1.4 1.5 0.8 0.8
Std Dev 0.03 0.03 0.02 0.01
% CV (1) 1.9 1.8 2.8 1.5
Al
2
O
3
(%)

1.6 1.1 17.2 17.6
Std Dev 0.13 0.07 0.33 1.72
% CV (1) 8.4 6.5 1.9 9.8
CaO (%) 38.9 40.1 36.6 37.0
Std Dev 0.81 0.73 0.67 0.21
% CV (1) 2.1 1.8 1.8 0.6
MgO (%) 10.7 11.1 7.8 7.8
Std Dev 0.23 0.17 0.15 0.04
% CV (1) 2.1 1.5 1.9 0.6
SiO
2
(%)

8.7 9.2 9.5 8.9
Std Dev 0.64 0.61 0.86 0.43
% CV (1) 7.4 6.7 9.1 4.8



44


Table 28 Main applications for EAF and BOF slags

Application EAF BOF Tendency Notes
Iron recovery and lime carrier in BF X
Recycling of entrained metals into the
metallurgical circuit
X X
Higher quality demands
on steel
Road construction
- Layers with/without binder
- Road cover, road base, ways consolidation
- Hot mix asphalt pavement (plant mixes,
surface seal, slurry applications)
X X
Incorporation in concrete (limestone
substitution)
X X
Granular based aggregate bases X X
Hydraulic engineering
- Waterway constructions on rivers and dams
- Fine grained mixtures as mineral filter
- Coarse grained armour stones
X X
High volume stability
required

Liming agent in agriculture (soil conditioner) X na pH control
Landfill, dumping X X
Growing demand for
other added value
applications
Marine applications (artificial reefs, marine
blocks, sediment capping)
X X
Mineral wool insulation X na
Railroad ballast X X na
Waste stabilization Emerging technique
Ceramic manufacturing / Porcelain industry Emerging technique
Mine acid effluent neutralization Emerging technique
na not available

Table 29 Chemical composition of EAF slag samples submitted to leaching tests at ISQ (% wt)

FeO MnO TiO
2
Al
2
O
3
CaO MgO Cr
2
O
3

41.2 3.95 0.39 5.16 18.4 3.91 0.95
Aged 3 months 41.4 3.93 0.38 4.63 18.5 3.84 1.13
Aged 6 months 39.3 3.81 0.40 4.76 18.2 3.62 0.99

45


Table 30 Chemical composition of ladle slag samples submitted to leaching tests at ISQ (% wt)

FeO MnO TiO
2
Al
2
O
3
CaO MgO Cr
2
O
3

2.01 0.38 0.24 4.08 54.1 7.37 0.03
Aged 3 months 8.48 2.40 0.22 3.07 38.2 8.01 0.24
Aged 6 months 1.29 0.39 0.21 3.29 46.0 6.11 0.06

Table 31 Trace heavy metal concentrations in fresh slags samples

Trial results (mg/kg) Leaching Limit Values
(2)
(mg/kg)

EAF
(1)
Ladle
(1)
Inert Non hazardous Hazardous
total Cr 0.46 0.59 0.5 10 70
As < 0.5 < 0.5 0.5 2 25
Pb < 0.5 < 0.5 0.5 10 50
Cd < 0.04 < 0.04 0.04 1 5
Cu 0.36 0.52 2 50 100
Hg < 0.01 < 0.01 0.01 0,2 2
Ni 0.18 0.32 0.4 10 40
Zn 0.26 0.98 4 50 50
(1) Average values of two samples.
(2) Leaching limit values according to Council Decision 2003/33/EC that establishes criteria and procedures for
the acceptance of waste at landfills (L/S = 10 l/kg; leachates extracted according to DIN 38414 S4).

46


FIGURES





Figure 1 Experimental set up, Antenna and slag sample



Figure 2 Instrumentation


47



0
0.5
1
1.5
2
2.5
3
2-6GHz 4-8GHz 6-10GHz 8-12GHz 10-14GHz 12-16GHz 14-18GHz
Frequency band( GHz )
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(

n

)
LD converter ladle low alloy steel Blast Furnace


Figure 3 Results of the refractive index dependency of the frequency for low alloy
steel slags.

0
0.5
1
1.5
2
2.5
3
2-6GHz 4-8GHz 6-10GHz 8-12GHz 10-14GHz 12-16GHz 14-18GHz
Frequency band ( GHz )
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(

n

)
AOD converter 1 EAF 1 AOD converter 2 EAF 2

Figure 4 Results of the refractive index dependency of the frequency for high alloy
steel slags
48



0
0.5
1
1.5
2
2.5
3
0.00 0.50 1.00 1.50 2.00 2.50
Basicity(CaO+MgO)/(Al2O3+SiO2)
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(

n

)
2-6GHz 4-8GHz 6-10GHz

Figure 5 The refractive index for all solid slags as a function of slag basicity defined
as (CaO+MgO)/(Al
2
O
3
+SiO
2
).

0
0.5
1
1.5
2
2.5
3
0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80
Basicity(CaO+MgO)/(Al2O3+SiO2)
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x


(
n

)
2-6GHz 4-8GHz 6-10GHz

Figure 6 The refractive index for the slags from low alloyed steel as a function of slag basicity defined
as (CaO+MgO)/(Al2O3+SiO2).

49





Figure 7 Furnace, schematic



Figure 8 Antennas suspended above HF induction furnace

50



5 cm 5 cm 5 cm
52 cm
Steel
Slag
52 cm
Steel
Slag
52 cm 52 cm 52 cm 52 cm
Steel
Slag
25 cm 25 cm
Argon inlet
Antenna position
Lid


Figure 9 Experimental set-up, schematic
0.00 0.25 0.50 0.75 1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
SiO
2
Al
2
O
3
9
5
10
CaO
S
i
O
2
synthetic slag 2
synthetic slag 7
synthetic slag 6
synthetic slag 4
synthetic slag 3
synthetic slag 8
C
a
O
Al
2
O
3
2
1
3 4
8
7
6


Figure 10 Chemical composition of the synthetic slag during the experiments

51



1.613
2.064
2.147
1.738
2.247
1.81 1.81
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
2-6GHz 4-8GHz 6-10GHz 8-12GHz 10-14GHz 12-16GHz 14-18GHz
Frequency band ( GHz )
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(

n

)
Ambiguous due to low signal level

Figure 11 Results of the refractive index dependency of the frequency for
synthetic slag 7
1.526
1.387
1.924
1.538
1.595
1.569
2.046
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
2-6GHz 4-8GHz 6-10GHz 8-12GHz 10-14GHz 12-16GHz 14-18GHz
Frequency band ( GHZ )
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(

n

)
Ambiguous

Figure 12 Results of the refractive index dependency of the frequency for
ladle slag


52



0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
2-6GHz 4-8GHz 6-10GHz 8-12GHz 10-14GHz 12-16GHz 14-18GHz
Frequency band ( GHz )
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(

n

)
Ladle slag LD slag EAF slag Synthetic slag 3
Synthetic slag 4 Synthetic slag 6 Synthetic slag 7 Synthetic slag 8

Figure 13 Results of the refractive index dependency of the frequency for all
investigated slags

0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
Basicity CaO/SiO2
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(
n
)
2-6GHz 2002 4-8GHz 2002 6-10GHz 2002

Figure 14 The refractive index as a function of the slag basicity defined as CaO/SiO2

53



0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00
Basicity (CaO+MgO)/SiO2
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(
n
)
2-6GHz 2002 4-8GHz 2002 6-10GHz 2002

Figure 15 The refractive index as a function of the slag basicity defined as
(CaO+MgO)/SiO2

0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80
Basicity (CaO+MgO)/(Al2O3+SiO2)
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(
n
)
2-6GHz 2002 4-8GHz 2002 6-10GHz 2002

Figure 16 The refractive index as a function of the slag basicity defined as
(CaO+MgO)/(Al2O3+SiO2)
54



0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40
B2 (CaO/SiO2)
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(
n
)
2-6GHz 4-8GHz 6-10GHz 8-12GHz 10-14GHz 12-16GHz 14-18GHz

Figure 17 Refractive index versus B2, CaO/SiO
2
ratio for different frequency bands

0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60
B3 (CaO+MgO)/SiO2
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(
n
)
2-6GHz 4-8GHz 6-10GHz 8-12GHz 10-14GHz 12-16GHz 14-18GHz

Figure 18 Refractive index versus B3, (CaO+MgO)/SiO
2
ratio for different
frequency bands

55


0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
B4 (CaO+MgO)/(Al2O3+SiO2)
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(
n
)
2-6GHz 4-8GHz 6-10GHz 8-12GHz 10-14GHz 12-16GHz 14-18GHz

Figure 19 Refractive index versus B4, (CaO+MgO)/(Al
2
O
3
+SiO
2
) ratio for different
frequency bands

0
0.5
1
1.5
2
2.5
3
3.5
4
2-6GHz 4-8GHz 6-10GHz 8-12GHz 10-14GHz 12-16GHz 14-18GHz
Frequency band ( GHz )
R
e
f
r
a
c
t
i
v
e

i
n
d
e
x

(

n

)
LD converter solid LD converter liquid ladle solid slag ladle liquid slag

Figure 20 Comparison of the refractive index for solid and liquid slags


56






Figure 21 Antenna encapsulation


57




Figure 22 Antenna encapsulation above ladle


Figure 23 The 6 t. pilot electric arc furnace equipped with the chute allowing to tap the liquid slag in
the different moulds
58





Figure 24 The different moulds for liquid slag tapping with specific equipment designed for slag
temperature measurement during its cooling



Figure 25 The specific equipment with thermocouple used for slag temperature measurement during
its cooling
59




Figure 26 Slag casting in the mould from the electric arc furnace


Figure 27 Mould equipped with temperature measurement during slag cooling
60



Slag 1
0
200
400
600
800
1000
1200
1400
1600
1800
0 60 120 180 240 300 360 420 480 540 600
Time (min)
T
e
m
p
e
r
a
t
u
r
e

(

C
)
Slag thickness 3 cm
Slag thickness 7 cm
Slag thickness 13 cm
Break-up of the thermocouple

Figure 28 Slag cooling temperature evolution during trial with slag 1


Slag 2
0
200
400
600
800
1000
1200
1400
1600
1800
0 60 120 180 240 300 360 420 480 540 600
Time (min)
T
e
m
p
e
r
a
t
u
r
e

(

C
)
Slag thickness 3 cm
Slag thickness 7 cm
Slag thickness 13 cm
Bad temperature measurement due to gas
pockets formation during the slag cooling

Figure 29 Slag cooling temperature evolution during trial with slag 2
61


Slag 3
0
200
400
600
800
1000
1200
1400
1600
1800
0 60 120 180 240 300 360 420 480 540 600
Time (min)
T
e
m
p
e
r
a
t
u
r
e

(

C
)
Slag thickness 3 cm
Slag thickness 7 cm
Slag thickness 13 cm
Bad contact between the slag and the thermocouple

Figure 30 Slag cooling temperature during trial with slag 3


Figure 31 Pockets gas formed during the slag cooling in the mould of 13 cm thickness with slag 2
62




Figure 32 Slag cooling temperature evolution of the different types of slag for a slag layer of 3 cm

Figure 33 Slag cooling temperature evolution of the different types of slag for a slag layer of 7 cm
63




Figure 34 Slag cooling temperature evolution of the different types of slag for a slag layer of 13 cm

0
200
400
600
800
1000
1200
1400
1600
1800
0 100 200 300 400 500 600
Time (min)
S
l
a
g

t
e
m
p
e
r
a
t
u
r
e

(

C
)
3 cm 7 cm 13 cm
Slag thickness

Figure 35 Average slag cooling temperature evolution versus the thickness of the slag layer
Slack thickness 13 cm
0
200
400
600
800
1000
1200
1400
1600
1800
0 60 120 180 240 300 360 420 480 540 600
Time (min)
T
e
m
p
e
r
a
t
u
r
e

(

C
)
Slag 1
Slag 2
Slag 3
Bad temperature measurement due to gas
pockets formation during the slag cooling
64


0
10
20
30
40
50
0 15 30 45
Grinder adjustment (mm)
G
r
a
i
n

s
i
z
e

l
o
w
e
r

t
h
a
n

6
.
3

m
m

(
%
)

.


3 cm 7 cm 13 cm Slag 1

Figure 36 Evolution of the fines proportion during the grinding of the slag 1
0
10
20
30
40
50
0 15 30 45
Grinder adjustment (mm)
G
r
a
i
n

s
i
z
e

l
o
w
e
r

t
h
a
n

6
.
3

m
m

(
%
)

.

3 cm 7 cm 13 cm Slag 2

Figure 37 Evolution of the fines proportion during the grinding of the slag 2

0
10
20
30
40
50
0 15 30 45
Grinder adjustment (mm)
G
r
a
i
n

s
i
z
e

l
o
w
e
r

t
h
a
n

6
.
3

m
m

(
%
)

.

3 cm 7 cm 13 cm Slag 3


Figure 38 Evolution of the fines proportion during the grinding of the slag 3


65



Slag 1
1
2
3
4
5
6
7
0 2 4 6 8
Ageing (months)
F
r
e
e

L
i
m
e

(
%
)

.
Thickness 3 cm Thickness 7 cm Thickness 13 cm
Slag 2
0
1
2
3
4
5
6
0 2 4 6 8
Ageing (months)
F
r
e
e

L
i
m
e

(
%
)

.
Thickness 3 cm Thickness 7 cm Thickness 13 cm
Slag 3
4
5
6
7
8
9
10
0 2 4 6 8
Ageing (months)
F
r
e
e

L
i
m
e

(
%
)

.
Thickness 3 cm Thickness 7 cm Thickness 13 cm


Figure 39 Free lime content evolution in the different slag samples during 7 months ageing
66







Figure 40 - Experimental set used in flow-through test. A EAF slag; B Ladle slag. Layers from top
to bottom: glass wool, slag sample, granite bed, perforated plate (leachate drain).




Figure 41 - Phenomena occurred during trials (channeling, system clogging, biological growth and
oxidation of metallic iron inclusions).

A B
67




Figure 42 EAF slag leachates evolution in flow through test
EAF - Fe
0,00
0,05
0,10
0,15
0,20
0,25
0,30
0,35
0,40
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

F
e

/

k
g
EAF - Ti
0,000
0,005
0,010
0,015
0,020
0,025
0,030
0,035
0,040
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

T
i

/

k
g
EAF - Ca
0
50
100
150
200
250
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

C
a

/

k
g
EAF - Cr
0,000
0,005
0,010
0,015
0,020
0,025
0,030
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

C
r

/

k
g
EAF - Mn
0,00
0,05
0,10
0,15
0,20
0,25
0,30
0,35
0,40
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

M
n
/

k
g
EAF - Al
0,00
0,05
0,10
0,15
0,20
0,25
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

A
l

/

k
g
EAF - Mg
0
10
20
30
40
50
60
70
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

M
g

/

k
g
fresh 3m 6m
A B
C D
E
F
G
68



Figure 43 Ladle slag leachates evolution in flow through test

Ladle - Fe
0,000
0,005
0,010
0,015
0,020
0,025
0,030
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

F
e

/

k
g
Ladle - Ti
0,00
0,02
0,04
0,06
0,08
0,10
0,12
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

T
i

/

k
g
Ladle - Mg
0,00
0,10
0,20
0,30
0,40
0,50
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

M
g

/

k
g
Ladle - Al
0,0
1,0
2,0
3,0
4,0
5,0
6,0
7,0
8,0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

A
l

/

k
g
Ladle - Ca
0
200
400
600
800
1000
1200
1400
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

C
a

/

k
g
Ladle - Cr
0,00
0,01
0,02
0,03
0,04
0,05
0,06
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
week
m
g

C
r

/

k
g
fresh 3m 6m
A B
C
D
E
F
69
70
71
European Commission
EUR 21787 - Factory-wide control and social and environmental issues
On||ne s|ag ana|ys|s ut|||s|ng contact-free m|crowave techno|ogy
D. Ma|moe|g, C. G||sva|d, C. Ga|vo Asceno
Luxembourg: Office for Official Publications of the European Communities
2005 - 71 pp. - 21 x 29.7 cm
Technical steel research series
lSBN 92-79-00080-2
Price (excluding vATj in Luxembourg: EUR 15