Don`t waste your Secondary Resources

Don‘t waste your
Secondary Resources
7th European Slag Conference
9th - 11st October 2013, IJmuiden
The Netherlands
Proceedings
EUROSLAG Publication
No. 6
Organized by:
Tata Steel, Harsco and
Pelt&Hooykaas
1
Preface
It is a great pleasure for EUROSLAG to welcome again friends and experts in slag
issues to the 7th European Slag Conference held in IJmuiden which is facilitated by
Tata Steel at Dudok Huis Conference Centre. We are delighted that about 200 people have been able to accept the invitation of EUROSLAG. It is the clearest indication
possible of just how many people, both in Europe and throughout the world, are professionally involved in slag and are interested in exchanging knowledge with one another concerning the production, properties, new technologies, standardisation, legislation and marketing of blast furnace and steel slag.
The IJmuiden conference is the successful continuation of a series of EUROSLAG
conferences started 1998 in Marseilles, followed by conferences in Düsseldorf, Nottingham, Oulu, Luxembourg and the last one in Madrid in 2010.
EUROSLAG deliberately chose The Netherlands as next conference venue in 2013
as Tata Steel is Europe's second largest steel producer with main steelmaking operations in The Netherlands and UK too. Together with Harsco Metals & Minerals and
Pelt & Hooykaas, IJmuiden they are producing, processing and marketing slag as
qualified products in different fields of application. So, EUROSLAG is very happy to
discuss again slag issues in IJmuiden together with all the experts present during the
three days conference.
The second reason to come to IJmuiden was that EUROSLAG also wants to give all
participants the opportunity to get to know the beautiful Netherlands with all its historical and cultural highlights.
The success of a conference depends on the lectures which will be presented. EUROSLAG therefore hopes that the content of the papers and presentations which
have chosen will be of interest for all participants, and that they may also be a source
of new ideas for own activities.
2
It is a lot of work to prepare the documents and to present results within a short time
of 20 minutes. Therefore EUROSLAG would like to thank all authors for their contributions to the conference in advance.
Taking into account the conference program, EUROSLAG has succeeded again in
receiving very interesting reports on the production, properties, use, marketing,
standardisation and legislation concerning blast furnace and steel slags.
The lectures have been divided into five sessions with the following themes. Each
session is chaired by a professional who is long term working in slag issues and cooperating with EUROSLAG since many years.
Theme 1:
Slag as By-product,
chaired by Hans Kobesen, Tata Steel, The Netherlands
Theme 2:
Metallurgy and Processing,
chaired by Dirk Mudersbach, FEhS-Institute for Building Materials,
Germany
Theme 3:
Research and Applications,
chaired by Nick Jones, Harsco Metals Group Limited, UK
Theme 4:
From Research to Applications
Part 4/1: chaired by Marko Mäkikyrö, Ruukki Metals Oy, Finland
Part 4/2: chaired by Karen Kiggins, National Slag Association, USA
Theme 5:
Environmental Affairs chaired by Heribert Motz, EUROSLAG
Knowing, what enormous work has to be done to organize a conference, EUROSLAG would like to thank the company management of Tata Steel for the acceptance of execution, the hospitality and especially Hans Kobesen and his colleagues for their very involved preparation.
Heribert Motz
Chairman of EUROSLAG
3
Content
Theme One
1. Resource efficient use of by-products and secondary raw materials with
regard to strengthening European legislation in Germany, G.Endeman
2. Study of Volume Stability and Recycling of BOF Slag at China Steel, YuChen Lee.
3. Efficiency of quartz addition on EAF slag stability, D.Mombelli*, C. Mapelli, S. Barella, A. Gruttadauria
Theme Two
4. Sustainable stabilisation and reuse of ladle furnace slag from electric
steelmaking, H. Schliephake, B. Dettmer, K. Schulbert, T. Zehn, T. Rekersdrees , P. Drissen, D. Mudersbach
5. Dry Slag Granulation – The Environmentally Friendly Way to Making
Cement, I. McDonald and A.Werner.
6. Mixing method for cooling and full vitrification of BFS, H. Kappes
7. Controlled cooling of BOF slag to enhance Fe-recovery, D. Poirier, M.
Gotelip Barbosa, W.Xuan, J. Poirier, G. Thevenin, D. Bulteel
Theme Three
8. Industrial Utilization of EAF Slag as Aggregate, A.E. Yıldızçelik, A. Ünal, O.
Yücel
9. Stabilization of CaO-SiO2-MgO (CSM) Slags by Recycled Alumina, H. Epstein, R. I. Iacobescu, Y. Pontikes, A. Malfliet, L. Machiels, P.T. Jones, B.
Blanpain
10. Recent
and
former
European
RFCS
slag
research
projects
Slag treatment and utilisation, I. Unamuno and A. Morillon
11. Construction of Test Sections to Evaluate Performance of Basic Oxygen
Furnace (BOF) Steel Slag as Aggregate in Stone Mastic Asphalt, J.S.
Chen, S.F. Chen, M.C. Liao, W.C. Chen, T.L. Tao, B.L. Hsu, T.K. Hsu
12. Slag properties – Easy access using new dedicated software, E. Nagels,
S. Arnout and B. Soete
4
13. Behaviour of Slag Bound Mixtures in Road Construction, Dr N Ghazireh,
B Kent and J Smith
Theme Four, part one
14. Development of new CEM X cements based on Ground granulated blast
furnace slag, fly ash and clinker, V. Feldrappe, A. Ehrenberg
15. Hydration Properties of Rapidly Air-Cooled Ladle Furnace Slag with
Gypsum, J.M. Kim,S.M. Choi,H.S. Kim,S.H. Lee,S.Y. Oh
16. Using Granulation (Pelletizing) to Icrease the Usage of Slags, J. Roininen
V. Kuokkanen
17. Investigation of BOF slag use for potato and tomato cultivation with saline irrigation water in Italy, T.A. Branca, C. Pistocchi, V. Colla, G. Ragaglini,
C. Tozzini and L. Romaniello
18. Removal of Phosphorus from Wastewater by Steel Slag Filter Systems,
P. Drissen, F. Chazarenc, M. Fixaris, M. Rex, H. Rustige, St. Troesch
Theme Four, part two
19. Does stored granulated blast furnace slag lose its reactivity?, A. Ehrenberg
20. Agricultural Utilization of Iron and Steel Slag in the USA, J. J. Yzenas Jr
21. Accelerated weathering of LD-slag using water and CO2., SR van der
Laan, JBA Kobesen, EJ Berryman, AE Williams-Jones
22. Development of ECO Slag Processing Technology for Iron Recovery and
Value-Added Products in Steelmaking, I. Sohn , J.I. Hwang, H.S.Kim , J.S
Choi, Y.S. Jeong, H.C. Lee
23. Global Opportunities of Steel-making Slag Materials as a Source of Silicon-based Fertilizers, M.C. Provance-Bowley , S.R. Miranda
5
Theme Five
24. Developing a Quality Protocol for Steel Slags, E Poultney, N Ghazireh, N
Jones, C Laskey, M Davies, P Redfern, J.Barritt
25. Sustainable reuse of iron and steel slags in road applications - Technical
requirements for environmental acceptance in France, J.Domas
26. Legislation vs. Regulation in the USA, C.Ochola
27. Development of Leachings tests for By-products and other seconday
construction products in Europe, E.Onstenk
28. Sustainability of construction works - European standards and implications for secondary materials, A. Schuurmans
6
Theme 1
Slag as By-product
7
G. Endemann
Resource efficient use of by-products and secondary raw materials
with regard to strengthening European legislation in Germany
Wirtschaftsvereinigung Stahl/Steel Institute VDEh, Sohnstraße 65,
D-40237 Dusseldorf, Germany
Abstract:
With the title “A resource-efficient Europe – Flagship initiative of the Europe 2020
Strategy” the European Commission requests a sustainable and efficient use of all
materials and resources. This is also the aim of the German resource efficiency
program called ProgRess.
The presentation gives a short overview on resource efficiency in steel making.
It describes aims and demands of above mentioned strategies and existing European
legislation, e.g. the Waste Framework Directive. This directive excludes by-products
from the waste regime and defines end-of-waste criteria. Experience therefrom is
evaluated with regard on progress regarding resource efficiency.
Legislation leaves space for national interpretation and implementation. But there are
also national initiatives and new legal developments on strengthening soil and water
protection in Germany. These could lead away from the integrated approach of sustainability and jeopardise the aims of resource efficiency.
8
1. What is resource efficiency
Wikipedia defines resource efficiency as “Maximizing the supply of money, materials,
staff, and other assets that can be drawn on by a person or organization in order to
function effectively, with minimum wasted effort or expense” [1].
The European “Roadmap to a Resource Efficient Europe” and the strategy “A resource-efficient Europe – Flagship initiative of the Europe 2020 Strategy” of the European Commission request a sustainable and efficient use of materials and all resources [2, 3]. But both miss to define resource efficiency. Nevertheless they propose
a number of different measures and approaches to increase resource efficiency and
to exploit synergies and address trade-offs. Help can be found at the “Online Resource Efficiency Platform” (OREP) [4]: “Resource efficiency means using the Earth's
limited resources in a sustainable manner.” Further on they promise that increasing
resource efficiency is a key to securing growth and jobs for Europe. It should bring
major economic opportunities, improve productivity, drive down costs and boost
competitiveness.
A sustainable and efficient use of materials and all resources is also the aim of the
German resource efficiency programme called “ProgRess” that aims “in particular to
minimise adverse effects of raw materials extraction and processing on environmental media” [5]. Unfortunately, a definition of resource efficiency is missing, too.
2. Resource efficiency and the steel industry
Steel making is material and energy intensive. This means that steelworks need
about 16 GJ of energy per ton of manufactured steel (Germany 2010). Additionally
German steelworks mainly use about 1 ton of iron ore, 0.44 tons of scrap, 0.27 tons
of coke and coke breeze, 0.1 tons of coal and briquettes, and 0.3 tons of lime per ton
of steel in total (German production mix). As these huge material amounts are directly related to operational and investment costs, too, it is easy to understand that the
steel industry from itself is always aiming to reduce its use of natural resources to an
economically balanced minimum. Energy and material intensity are for many years
and will be the most effective drivers to reduce the factors of consumption to be as
economical (=efficient) as possible.
9
Steel making also has impacts on the environment, so on ecology, but above mentioned effects also lead to reduce them. Nevertheless, the steel industry is also aiming to reduce the environmental impacts as far as possible. During all these efforts it
should not be forgotten, that sustainability goes far beyond optimizing environmental
effects and reducing use of resources. The impacts on the other pillars of sustainability, economy and society, have to be examined and balanced in the same manner. A
solid industry is the base of modern societies, but without economic feasibility and
competitive framework conditions, no industrial sector will be able to invest in research and development, new production processes or facilities for environmental
protection etc.
3. Resource efficiency is not only limited to input materials
Resource efficiency does not end with reduced input materials. It is also related to
the production processes themselves, to recirculation of materials, within production
or at the end of use phases, and to by-products and wastes. The use of scrap as
(secondary) raw material in steel making is one of the most important efficiency
measures developed by humankind in history.
Recycling of steel and goods made of steel keeps it in the material cycle endlessly. A
new and convincing LCA study (life cycle assessment) has been elaborated by
Prof. Finkbeiner from the Technical University of Berlin, Institute for Sustainable Engineering, together with the steel industry in Germany [6]. Its final outcome, CO2
emissions from steel making are lower than hitherto estimated and the inherent properties of steel are preserved during recycling, shows the enormous relationship between resource efficiency and recycling. By a new integrated method, called “Multiple
Recycling Approach” (MRA), it is possible to account for the multiple life cycles possible for steel. Steel scrap from used goods can be recycled to new products after a
product (made of steel) has reached the end of its use-phase and has been collected
and returned to the steel shops. Prof. Finkbeiner shows that the ecological footprint
of steel is reduced with each recycling step. Thereby the CO2 emissions from steel
production are about 40 per cent lower than calculated from primary production only.
Thus, a realistic CO2 emission per ton of steel is calculated on less than 1 ton CO2
10
per ton of steel, figure 1. The study also proves the 100 per cent recyclability of steel
and the preservation of its inherent properties during recycling.
During the different steps of steel making, production is also aiming for by-products
and other materials that can be used internally or in other industrial sectors. Examples are manifold from tar to iron oxides. Most efficient internal circuits have been
installed to maximise material usage and to avoid wastes as far as possible, figure 2.
Concerning this matter, figure 3 shows an example for using dusts and sludges and
reducing dumping to 20 % (Germany, 2010).
Slags are the most important example for by-product use from metallurgical processes. In 2012, steel works in Germany produced over 13 Mio tons of slag as byproducts. The total amount of blast furnace slag, 7 Mio tons, is used as granulated
slag in cement making (85 %), as mineral mixtures (14 %) or for other purposes.
Steelworks slag (~ 4 Mio tons) is mainly used as building material (54 %), for internal
recirculation (12 %) or as fertilizer (8 %). Such usage rates require a perfect material
management, quality management and professional sales departments that also develop new markets. A continuously research work for developing new methods for
slag processing is essential. Only small amounts of special slag fractions are dumped
on landfills. This success story shows the high impact of product and waste legislation for resource efficiency of the steel industry. Nevertheless, rates for using the byproduct slag are extremely sensitive to legislation.
4. Present and planned legislation
Legislation on the resource efficient use of by-products and secondary raw materials
is manifold. On European level the present framework is given by the waste framework directive (WFD) [7]. Its most important innovations are, that by-products are exempted from its scope and that waste may cease to be waste. Being a by-product
means, these substances will be out of the scope of the WFD from the beginning.
They never enter the waste regime. The WFD gives some minimum criteria, which
aim at a certain and lawful use of the substance without any waste typical processing, while its production is an integral part of the production of the main product.
Its use should not lead to adverse environmental or human health impacts. The criteria should be worked out in detail by a comitology process, which is still pending. The
11
WFD also gives some guidance on end-of-waste (EOW) criteria. Here the European
Commission (COM) has put much more efforts in developing such criteria in comitology, e.g. on scrap material.
The introduction of the so called 5 steps hierarchy as a priority order in waste prevention and management was important, too. Prevention ranks before recovery. Recovery has been divided into preparation for re-use, recycling and other recovery, like
energy recovery. The last ranking is given to disposal.
Whether the WFD with its exclusion of by-products and EOW criteria caused progress regarding resource efficiency must be denied. This can simply be shown by the
example of scrap. Steel scrap is recycled 100 % into the value chain and nearly without losses. Steel scrap according to the European scrap specification list normally
needs no further treatment for direct use it in steel making. If treatment is necessary,
this is done by scrap suppliers. Negative environmental or human health impacts can
be foreclosed in most cases. EOW criteria for steel have been defined by comitology.
The steel industry herein requested practicable criteria that could be checked easily
and within normal operation of scrap delivery. But the final “Council regulation establishing criteria determining when certain types of scrap metal cease to be waste under Directive 2008/98/EC of the European Parliament and of the Council” sets the
limit for foreign materials (steriles) on ≤ 2 % by weight [8]. This strict limit destroyed
all ambitions of the scrap collectors and dealers and steel industry to move scrap
flows from the waste to the non-waste regime – all this without any advantage for the
environment. Applying this regulation means additional bureaucracy and financial
burden. In front of this experience, we will expect no progress when the COM ever
starts with by-product criteria for slag, although the COM has developed some years
ago a communication on the interpretative communication on waste and by-products
that led into the right direction [9].
WFD was implemented in most EU member states until December 12, 2010. The
German Kreislaufwirtschaftsgesetz (Closed Substance Cycle Waste Management
Act) took over the regulations on by-products, EOW and the 5 step hierarchy [10].
According to the environmental ministry, product responsibility is at the heart of waste
management policy in Germany. It should put the idea into practice that waste avoidance is best achieved by holding the generator of waste responsible. Products shall
12
be designed to reduce waste occurrence and to allow environmentally sound recovery and disposal of the residual substances, both in the production of the goods and
in their subsequent use.
Regarding future legislation, the “Roadmap to a Resource Efficient Europe” describes
challenges and opportunities for Europe and the idea how to make Europe more resource efficient. In order to transform the economy onto a resource-efficient path the
COM sees the need for policies that recognise the interdependencies between the
economy, wellbeing and natural capital and seeks to remove barriers to improved
resource efficiency, whilst providing a fair, flexible, predictable and coherent basis for
business to operate. In order to promote further sustainable consumption and production, it gives a colourful bouquet of measures such as

Strengthening the requirements on Green Public Procurement;

Establishing a common methodological approach to benchmark the environmental performance of products, services and companies;

Addressing the environmental footprint of products and setting requirements under the eco-design directive and expanding it to non-energy related products;

Extending producer responsibility to the full life-cycle of the products.
To improve recycling and to turn waste into a resource, the COM wants to:

Stimulate secondary materials markets and demands for recycled materials
through economic incentives and developing of EOW criteria;

Review existing prevention, re-use, recycling, recovery and landfill diversion targets, asses/introduce minimum recycled material rates, durability and reusability
criteria and extent producer responsibility;

Eradicate illegal waste shipments (esp. hazardous waste);

Improve public and private investment into R&D for resource;

Promote regular exchange on best practices and market based instruments.
The sword of Damocles is given by the planning on taxes on pollution and resources
in combination with the aim of shifting taxation away from labor to environmental impacts and the review of fiscal policies and instruments with a view to support resource efficiency more effectively. This is to be rejected but shows the relevance and
13
the highly endangering potential of the presently running discussions on European
level. There are manifold measures listed for all different resources. Minerals and
metals are pointed out as essential for resource efficiency. Additionally interactions
with the raw materials initiative and the climate and energy policies are outlined.
Above mentioned German ProgRess is not to be seen as hurrying ahead of the German authorities, but as an addition, although – or even – because German Environmental ministry and COM compared their notes. ProgRess includes basic programmatic statements, strategic approaches along the entire value chain and specific examples. It is shaped by four guiding principles:
1. Joining ecological necessities with economic opportunities, innovation support and
social responsibility;
2. Viewing global responsibility as a key focus of German resource policy;
3. Making economic and production practices less dependent on primary resources,
developing and expanding closed cycle management;
4. Securing sustainable resource use by guiding society towards quality growth.
Examples for measures are to strengthen efficiency advice for enterprises, support
environmental management systems, take greater account of resource aspects in
standardization, place greater focus on the use of resource-efficient products and
services in public procurement, strengthen voluntary product labelling and certification schemes and enhance closed cycle management.
The third part of the program on specific examples includes sections on bulk metals,
like steel, but also on sustainable construction. The chapter on bulk metals mentions
the importance of steel but also the rising prices of primary and secondary raw materials. Although nearly all steel scrap is recycled in Germany, it cannot meet the need
for steel. In total, the German government is aiming to increase material-efficient and
energy-efficient production, processing and recycling methods by
•
Giving greater publicity to and increasing expansion of the programs for identifying and improving material and energy efficiency in companies;
•
Improving the market conditions for material and energy efficient products, for
example through market incentive programs and eco-labelling;
•
Ensuring integration of and focusing on material and energy efficiency in training;
14
•
Promoting cross-sectorial knowledge exchange;
•
Revitalizing recycling markets by development of EOW criteria.
Although proposed by the steel industry in Germany, ProgRess does not mention the
possibilities of using the by-product slag as constructional material and its possibilities to reduce the use of natural resources in a sufficient manner. All in all, the resource efficiency strategies/programs on European and national level are more or
less in line with the original idea of the WFD to increase recycling and to develop Europe to a recycling society. But last mentioned example makes clear: we are far away
from applying a real integrated approach in policies.
Resource efficiency is not only limited by environmental, technical and economic effects. Resource efficiency in reality depends on legal framework conditions and policies. Here we face some developments, especially on national level in Germany, that
are able to jeopardize the aims of European strategy and national program.
Everyone should be aware that we have substance related policies, like REACH
(2006/121/EC) that seems to be completely disregarded by authorities and politicians
that are dealing with waste or recycling topics [11]. Reach gives all necessary legal
tools to handle any kind of material that is not waste. So, if a material ceases to be
waste or is a by-product, it automatically falls in the scope of REACH. That is why the
European steelmakers decided to register slags within REACH.
For some years, Germany is planning regulations on the usage of alternative constructional materials, on groundwater and soil protection and on handling substances
endangering a body of water. All of them are mainly focussing on the protection of
groundwater, although having different headlines. They are connected with an initiation of so called insignificance threshold values, e.g. the maximum value of a parameter that can be tolerated without expecting negative effects on the groundwater
quality. Depending on the place of evaluation, limit values basing on insignificance
threshold values will enormously affect the usage of all kinds of material.
Regarding the regulation on alternative constructional materials, especially the usage
of slags will be dramatically influenced. We are afraid of more than 2.5 Mio tons of
slags per year – mainly steelworks slag – that could not be used for constructional
purposes any longer with the danger of being landfilled with costs estimated to more
15
than 150 Mio Euro and loss of valuable space on landfills or need for additional landfill respectively. It is important, that this regulation will only be applied on artificial materials like industrial by-products or recycling materials but not on natural materials
like stones, gravel or sand, although we know that these materials may have higher
leaching values than slags. In connection with WFD and KrWG it is also interesting
that German policy makers presently plan a differentiation between waste and nonwaste within that regulation, as the German ministry of justice criticised the legal
basis of former drafts. With regard to above mentioned insignificance threshold/limit
values, it might only be possible to classify granulated slag as by-product (hopefully)
while all other slag types stay in the waste regime. This is a typical example for national solo attempts.
Another danger is popping up regarding to the handling of substances that may be
dangerous to water. A new federal regulation shall displace 16 different regulations in
the Federal States of Germany by one. Instead of focusing on unregulated substances only, the scope of the draft covers nearly any kind of substance or mixture. Each
substance or mixture is threatened to be automatically classified as generally endangering a body of water, except it is finally accepted as harmless by a specially build
commission on the basis of stringent rules. This means that for example slag processing, handling and storage might be classified as potentially endangering water
and in worst case that plants and areas have to be protected by roofs and doublewalled constructions. Billions of Euros might be invested to meet such requirements.
With the experience of history in European legislature, there is much fear that the
German developments and ideas will soon be brought into the European Community
and once more will complicate the process of increasing resource efficiency and becoming a sustainable society.
5. Summary and way forward
The steel industry supports the aims of sustainability and resource efficiency without
any restraints. Making processes more efficient and using resource sustainable is our
daily business. It is necessary to think not only in one direction but to take into account all possible interactions and effects, from technical point of view but especially
regarding policy making and legislature.
16
There is a need to accept industrial activities and efforts for an environmental, social
and economic sound industrial development. Legislation should base on a holistic
approach that takes all interactions between different policies into account and balances the 3 pillars of sustainability. Therefore it is also necessary to pay regard on
the multiple recycling approaches that has been elaborated for life cycle assessment
purposes. Instead of an increasing number of regulations and more complexity of
legislation, we need a better and unified implementation of EU legislation in member
states (with regard to regional differences) and a better application of existing legislation instead of increasing the legal jungle. R&D and pilot projects have to be supported instead of increasing the financial burden to the industry and revoking its
economic basis.
References
[1]
http://en.wikipedia.org/wiki/Resource_efficiency, 9.8.2013, 18:00
[2]
COM (2011) 571 final of 20.9.2011
[3]
COM (2011) 21 of 26.1.2011
[4]
http://ec.europa.eu/environment/resource_efficiency/about/index_en.htm,
9.8.2013, 18:19
[5]
http://www.bmu.de/fileadmin/bmu-import/files/pdfs/allgemein/application/
pdf/progress_bf.pdf, 9.8.2013, 19:29
[6]
http://www.stahl-online.de/Deutsch/Linke_Navigation/MedienLounge/_
Dokumente/120621_Finkbeiner_Multi-Recycling_von_Stahl.pdf
[7]
Directive 2008/98/EC of 19.11.2008
[8]
Council regulation (EU) No 333/2011 of 31 March 2011
[9]
COM (2007) 59 final of 21.2.2007
[10] http://www.gesetze-im-internet.de/bundesrecht/krwg/gesamt.pdf; 12.08.2013
14:04
[11] 2006/121/EC of 18.12.2006
17
Figure 1:
Global Warming Potential for the production of steel calculated by
MRA (Germany, TU Berlin 2012)
Figure 2:
Simplified material circuits in the steel industry
18
Figure 3:
Usage ratios of dusts and sludges from steelmaking in %
(Germany, 2010)
Figure 4:
Production and utilization shares of blast furnace and
steel works slag in % (Germany, 2012)
19
Yu-Chen Lee
Study of Volume Stability and Recycling of BOF Slag at China Steel
Ceramic Materials Section, New Materials R & D Dept., China Steel Corp., Taiwan
Abstract
The use of BOF slag is always limited by its inherent property of disintegration after
hydration. In this study, XRD, optical microscopy, scanning electronic microscopy
and expansion measurements were employed to examine microstructural variation
and expansion behavior of different slags after steam pyrolysis, steam aging, waterquenching and BSSF (Baosteel Short Slag Flow) treatment. The results indicated
that expansion ratio of all stabilizing treated slags were still higher than 14 to 35
Vol.% due to free lime and periclase were tightly coated and surrounded by low melting point liquid slag, and hydration reaction was definitely obstructed. On the contrary, hot slag treatment developed by China Steel has demonstrated superior volume
stability with less than 0.6 vol. % expansion based upon twenty-six pilot trials. In addition, a hot treatment station with annual capacity of six hundred thousand tons of
slag was built-up in July, 2012. Meanwhile, the study of valorization and engineering
application for stabilized slag such as dike, building materials and functional products
is ongoing.
1. Introduction
In China Steel Corp.(herby abbreviated by CSC) , there are about 3.07 million tons of
iron-making slags and 1.23 million tons of steel-making slags outputed in 2010. The
last includes blast furnace slag and de-sulfurization slag with each quantity of 2.76
and 0.31 million tons, respectively. The latter are composed of converter or BOF
(Basic Oxygen Furnnace) slag, a by-product of metallurgical reaction during steelmaking processs at near 1700℃, and secondary refining slag with each quantity of
1.09 and 0.14 million tons, respectively. More than ninty percents of blast furnace
20
slag are water-queenching and reused for cement industry. The rest is air-cooling
slag and recycled for the basement of road construction.
Steel-making slags are followed by the procedure of crushing, magnetic separation
and sieving process to obtain different product for recycling and utilization. Chemical
analysis indicated that the dioxin content in the converter slag was about 0.0001 ngTEQ/g. This value was much lower than the EPA controlling limitaion of 1.0 ngTEQ/g (dry-based). In addition, TCLP (Toxicity Characteristic Leaching Procedure)
analysis, shown as Table 1, also revealed that heavy metal elements such as Hg, As,
Cu, Cd, Cr, Cr+6, Pb and Se, etc was significantly less than EPA controlling decree.
Obviously, there is not any problem in dioxin pollution and leaching of heavy metal
elements in BOF slag.
Table 1: TCLP analysis results of BOF slag at China Steel Corp.
Compositions (ppm) (mg/l)
BOF Slag (as received)
Control standard (EPA, Taiwan)
Hg
As
Cu
Cd
Cr
Cr
+6
Pb
Se
<0.003 <0.1 <0.1 <0.03 <0.1 <0.05 <0.3 <0.03
0.2
5.0 15.0
1.0
5.0
2.5
5.0
1.0
Normally, converter slags are widely used in different field of application, e.g. in the
building industry, road construction and engineering filling or as concrete additive. It
was known as concrete can crack and can even disintegrate mainly by hydration
reaction of the reminded free lime and periclase in slag.
Therefore,, in Europe,
Japan and United State, steel-making slags must be placed in open yard to allow
natural weathering for more than half to two years before used for road
construction.(1,2)
The destinations of steel-making slag in CSC were recycled and reused as scrap with
a proportion of 3.6 wt% (one of third is recycled in CSC’s BOF), raw material for
sinter ore (12.5 wt%), aggregate for asphalt concrete (3 wt%). Majority of steelmaking slag, 71.4 wt%, were used for civial engineering application including
pavement of sidewalk. Excluding the hydraulic application such as dike etc., the
destinations of steel-making slag in CSC are quite the same as Japanese, European
and North American steel plants(1,2). Recently, steel-making slag utilization became a
severe problem due to large scale reclamation sites being increasingly hard to find in
21
Taiwan. Moreover, a significant decree of total chromium with an unbelievable value
of 250 ppm limiation is still adopted in EPA, Taiwan. As a result, there existed a
severe restriction for the utilization of steel-making slag.
As well known, volume expansion resulting from slaking or hydration of free lime and
periclase components in BOF slag is the main reason to restrict the recycling
direction of steel-making slag(3,4).
Therefore, this study is focused on the
investigation of residual volume expansion of slag and to evaluate different so–called
volume stabilization treatment methods for converter slag and to develop a fully
stabilized process for CSC’s BOF slag, and to make a effort on any potential high
value-added application technology to accelerate the utilization of steel-making slag.
2. Experimental procedures
(1).Materials
Some specimens were from steel plant of China Steel. On the contrary, samples
treated by different commercialized processes were taken from several Chinese steel
plants surveyed by a CSC’s task force in 2006 and 2011.
(2). Sample Preparation
BOF slag was grinded to powder with a grain size below 104m. Two kinds of cylinder
specimen
were
prepared:
(1).Rectangular
bar
with
a
dimension
of
20x20x120mm, (2).Cylinder bar with a 50mm diameter and 100mm height. After
hardening, specimens were set for 24 hours before properties test.
(3). Property analysis
Two methods were applied for volume expansion/residual expansion test: (1).Test in
autoclave: in the atmosphere of 132℃ and 2 kg/cm2 steam pressure for 8h per each
time, and 10 times repetition. (2).Test in hot water bath: specimen was immerged in
80 to 80℃ hot water in normal pressure. All testing specimen were checked every
day and recorded the situation of crack, collapse or disintegration.
22
3. Results and Discussion
(1). Comparison on volume stability of six slag treatment methods
Nowadays, there are six different treatment methods for converter slag, including
steam aging (normal and higher pressure), HK water-quenching, steam pyrolysis,
BSSF (Baosteel Short Slag Flow) and hot slag treatment, shown in Table 2.
Table 2: Comparison of six commercialized slag stabilization methods
Items
High pres-
Normal
sure steam
pressure
aging
steam aging
Major Users SMI Wakayama
Choice of
Cold slag
SMI
Kashima
HK water- Steam pyquenching
Hot slag Modification/
rolysis
Treatment
Liuzhou Major steel
Baosteel, ThyssenKrupp,
Steel,
plants in
POSCO,
A.M. Gent,
China
China
Dragon
Rouge Steel
Steel, etc.
CSC etc.
Cold slag Slag with All kind of
slag property
BSSF
low vis-
hot slag
All kind of Basicity lower
hot slag
than 5.5
equipped
cosity
with skimmer
Operation
50 tons/
1300
Characteristics batch, within tons/batch,
3 hours
Utilization
pot,
within 6 days 30min.
Road con- Road construction,
< 20 ton/ ~250 ton/
struction,
Raw
batch, with- tons/pot,
in 12h
Road con-
30min./pot
cement
gate
12000 T/M 30000 T/M
--
--
within 2040min.
tion, AC ag-
cement
gregate,
additive
Dike, concrete
additive
Capacity
20-40 tons/pot,
Road con- Road construc-
material struction, AC struction,
AC aggre- AC aggre- for sinter aggregate,
gate
20-40
etc.
--
25000-45000
T/M
Apparently, there exists a regional distribution for these slag treatment methods, i. e.,
(1) Normal or high temperature(~180℃)/ pressure (1MPa) steam aging method was
almost adopted in Japan, (2) HK water-quenching treatment was only adopted in Liuzhou Steel, (3) Steam pyrolysis developed by China Metallurgical Group Corp., is
23
more popular but only in China; (4) BSSF treatment was adopted by Baosteel,
POSCO, Dragon Steel (a subsidiary company of CSC) and Indian steel plants; (5)
Hot slag modification/ treatment method was initiated in TKS and promoted to other
three steel plants in Germany and United States. ArcelorMittal Gent plant also established the treatment station alone in Belgium. In 2012, CSC designed independently
and built up a two-stands injection station with annual capacity of 0.6 million tons in
Taiwan.
(2) Evaluation on the residual expansion of BOF slag treated by different methods
A. Normal pressure steam aging and high pressure steam aging
In order to find out a fully stabilized method for slag to avoid volume expansion, CSC
set up two sets of normal pressure steam aging equipments for trial from 2001 to
2003.
Unfortunately, the volume expansion was always higher than 6 to 20%.
Therefore, in 2005, CSC delivered 300kg slag to SMI and were treated with high
pressure steam aging equipment for 4 to 8 hours trial test. The residual expansion
was evaluated in CSC laboratory and shown as Figure 1. The results indicated that
residual expansion of original slag with a grain size great or small than 5mm, and
after 4 and 8 hours steam aging were 39.74%, 26.72%, 36.97%, 18.07%, 36.96%
and 15.36%, respectively. All expansion values were much higher than below 0.5%,
a limit value for road construction according to Taiwan national standard.
B. BSSF treatment
As shown in Figure 2, the residual expansion of BOF slag with a grain size great than
5mm after BSSF treatment is 14.02%, comparing with a 30.16% expansion of original
slag. Besides, the free lime content is in the range of 2.25 to 4.21 wt%, still higher
than a limit value of 0.2 wt%. Moreover, the specimens were all collapsed after cement soundness test by autoclave, as shown in Figure 3.
C. Steam pyrolysis
Sample from Anshan Steel was analyzed by chemical composition and shown the
free lime content was about 10.25 wt%. As a result, the specimens were all collapsed after cement soundness test, also shown in Figure 3.
24
45
Residual Expansion (%)
40
35
30
25
20
15
Original slag(>5mm) before steam aging
SMI-Steam Aging 4h(>5mm)
SMI-Steam Aging 8h(>5mm)
Original slag(<5mm) before steam aging
SMI-Steam Aging 4h(<5mm)
SMI-Steam Aging 4h(<5mm)
10
5
0
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
Duration Time (Days)(in 80℃ Hot Water)
Fig.1. Residual expansion curves after high steam aging in SMI for different grain
size of BOF slag
35.00
30.16%
Residual Expansion (%)
30.00
Original fresh slag from Baosteel
25.00
Baosteel slag after BSSF(> 5mm)
20.00
14.02%
15.00
10.00
5.00
0.00
0
25
50
75
100
125
150
175
200
225
Time (Days)
Fig.2. Residual expansion of BOF slag treated by Baosteel Short Slag Flowchart
(BSSF) method
25
Fig.3 Cement soundness test samples composed of BSSF, Steam pyrolysis and Water-quenching methods treated slags were disintegrated.
D. HK water-quenching treatment
Sample from Liuzhou Steel was analyzed by chemical composition and shown the
free lime content was about 7.39 wt%. Therefore, the specimens were all collapsed
after cement soundness test, also shown in Figure 3.
E. Hot slag Modification/ Treatment in laboratory
Figure 4 shows the residual expansion of CSC BOF slag before and after different
silica sand addition from 10, 30 and 50wt% and melted in 1500℃ for 10 or 20
minutes. The testing result revealed that the residual expansion can be reduced from
original 20% to 0.01~0.46%. Obviously, hot slag modification is the most effective
method to stabilize BOF slag. Therefore, in 2010 and 2011, a serious field tests,
more than twenty-six trials, were executed in De-sulfurization station. A positive result was proven that the residual expansion of BOF slag was decreased from
37~42% of original slag to 0.2~0.6%, nearly to meet the requirement of a limit value
of 0.5%. In addition, the content of free lime and periclase was also diminished from
14~18% to 0~0.13%. The pH value is reduced from 12.6 of original slag to 11.7 after
silica sand injection.
26
Finally, a hot treatment station with annual capacity of six hundred thousand tons of
slag was built-up in July, 2012. Meanwhile, the study of valorization and engineering
application for stabilized slag such as dike, building materials and functional products
Residual Expansion (%)
is ongoing.
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Original BOF slag
BOF Slag after water bath treatment
10S1510
10S1520
30S1510
30S1520
50S1510
50S1520
10/30/50wt%Silica sand were
added, then after 1500℃x10/20min.
0
5
10
15
20
25
30
35
40
45
50
Time(Days)( in 80℃hot water bath)
Fig.4. Residual expansion of BOF slag after different silica sand addition and melted
in 1500℃ for 10 or 20 minutes
4. Conclusion
After a serious of volume stability evaluation of specimens from different treatment
methods, it is demonstrated that only hot slag modification/ treatment can fully stabilize BOF slag. The residual expansion of silica sand injection treated slag is reduced
to less than 0.6%. Free lime and periclase content is also diminished to 0~0.13%.
pH value is less than 11.7. Therefore, in 2012, a two[stand silica sand injection station with annual capacity of six hundred thousand tons of slag was established and
hot trial from then. Besides, a study of valorization and engineering application for
stabilized slag such as dike, building materials and functional products is also ongoing.
27
Reference
1. Schriftenreihe der Forschungsgemeinschaft Eisenhuttenschlacken, "Iron
and Steel-making slags－ Properties and Utilisation", ISSN 0948-4787,
2000.
2. G. Maltyas, "Utilization of steelmaking slag", Iron and Steel Engineer, pp2930, Aug. 1978.
3. Kiichi NARITA, et al., "On the Weathering Mechanism of LD Converter
Slag", Tetsu-to-Hagane, Vol. 64, No.10, pp1558-1567, 1978.
4. Frank Wachsmuth, et al., "Contribution to the Structure of BOF-Salgs and
its Influence on their Volume Stability", Can. Met. Quart., Vol. 20, No. 3,
pp279-284, 1981.
28
Efficiency of quartz addition on EAF slag stability
D. Mombelli*, C. Mapelli, S. Barella, A. Gruttadauria
Politecnico di Milano, Dipartimento di Meccanica, Via La Masa 1, 20156 Milano, Italy
G. Le Saout, E. Garcia-Diaz
École des mines d’Alès, Centre des matériaux des mines d’Alès (C2MA), Avenue de
Clavières 6, 30100 Alès cedex, France
* corresponding author: davide.mombelli@mail.polimi.it
Abstract
EAF slag could be re-utilized as alternative stone material only if they are chemically
stable in contact with water. The presence of hydraulic phases as larnite (2CaO.SiO2)
could imply the releasing of dangerous elements into environment. Chemical
treatment seems to be the only way to assure completely stable structure, especially
for long-term application. Efficiency of silica addition during the deslagging period is
presented in this study. Microstructural characterization of modified slag is performed
by SEM and XRD analysis. Elution tests have been performed according to EN
12457-2 standard and the obtained results have been compared with and without
silica addition. The obtained results demonstrate the reliability of the proposed
process.
Introduction
EAF slag are today considered effectively equivalent to common stone material
employed in civil engineering applications. Their physical and mechanical
characteristics often are higher than the traditional raw material [1] and, in the recent
decades, their use is increased exponentially [2].
Road construction and concrete production are common applications in which EAF
slag could be mixed to or completely replaced the traditional raw material [3,4,5,6,7].
Granulated slag are mixed with other substances, i.e. bituminous binders or cement,
thus they are not directly exposed to the environment. Generally for such
applications, elution tests are not required by the standards because the slag is not in
contact with water. However, several studies [5,7,8] investigated the leaching
behavior of slag aggregates, highlighting that also such aggregates in contact with
29
water could release dangerous chemical elements (especially Ba, V and Cr) as well
as unbound slag.
Other investigation on the leaching behavior of EAF slag [9,10,11,12,13] tried to
identify the mineralogical phases major responsible of dangerous species release.
Experimental test identify the calcium silicates, especially larnite (2CaO.SiO2 or
C2S1), hartrurite (3CaO.SiO2 or C3S) and bredigite (7CaO.MgO.4SiO2), as the
phases reacting with water and responsible of elements releasing, as barium [10,11].
Other phases could contribute to increase the elements amount in elute:
brownmillerite
(4CaO.Al2O3.Fe2O3
or
C4AF),
calcium
aluminate
and
non-
stoichiometric spinels could be associated to Cr release. Barium oxide is a
substitutional chemical compound of CaO in calcium silicates whereas chromium (in
both trivalent or hexavalent form) could replace iron and aluminum oxide in
brownmillerite-type phase [14] and in calcium aluminates [15]. Hydration process
facilitates the migration into water of species as Ca, Mg or Ba contributing to increase
the pH of the solution. In basic environment chromium oxide could be easily
dissolved in hexavalent form, that is the most reactive and dangerous (it is classified
as carcinogenic by the World Health Organization).
Hydration reaction needs some days to be completed. Thus, this phenomenon is very
difficult to appreciate in EAF slag during the 24 hours of standard leaching test (EN
12457-2 standard), because only a fraction of slag, lower than 1%, reacts with water.
However this amount is enough to solubilize a quantity of Ba, Cr and V that
overcomes the limits imposed by environmental regulations. As the polluting
elements are into the silicates, one way to overcome the problem is to reduce their
amount or to reduce their solubility. It could be achieved by promoting the
transformation of the phases in which the aluminum is in the tetra-coordinated form
and increasing the number of SiO4 tetrahedra [16]. Thus, a thermochemical
inertization process was developed in order to transform EAF slag in absolute safe
by-product useful both for aggregates and unbound applications. The treatment
consists in pure quartz addition into the fluid slag during deslagging operation. This
process is based on the same principle adopted by Drissen et al. [17] to stabilize
BOF slag. The admixed slag is collected in the slag-pot that contributes to stir it,
1
Cement chemist notation is adopted : C: CaO, S: SiO2, A: Al2O3, F: Fe2O3. For example, larnite
2CaO.SiO2 is written C2S
30
homogenizing the microstructure.
In this work the benefits conferred by silica treatment on slag release behavior are
presented. The efficiency of silica addition is experimentally demonstrated by XRD,
SEM and elution tests analyses, with regards on liquid-on-solid ratio effects.
Experimental procedure
Two batch of carbon steel slag are investigated and modified by the above
mentioned inertization process. Their chemical composition are measured by XRF
and checked by SEM-EDS and are reported in Table 1 together with sample
identification and optical basicity ( ). The analyzed slags are located on phase
ternary diagram shown in Figure 1.
Table 1. XRF chemical composition (weight %) and samples identification.
Sample ID
SiO2
Al2O3
CaO
MgO
FeO
S1
12.36
8.96
26.10
3.21
37.53
0.800
S2
10.36
7.26
28.67
3.98
39.69
0.820
MS1
18.34
9.25
23.26
2.85
35.93
0.759
MS2
22.39
8.20
19.18
2.24
37.29
0.739
Figure 1. Samples identification on (a) CaO-SiO2-FeO and (b) CaO-Al2O3-SiO2
phase diagram[18].
The comparison between as-received (S1 and S2) and modified slag (MS1 and MS2)
was performed from morphological and microstructural point of view with XRD and
31
SEM analyses.
X-ray diffraction (XRD) data were collected using a Bruker D8 Advance diffractometer
in a
-
configuration employing the Cu K
radiation ( =1.54Å) with a fixed
divergence slit size 0.5° and a rotating sample stage. The samples in form of powder
obtained by ring mill grinding (average diameter 15
m) were scanned between 10°
and 80° with the Vantec detector. The qualitative analysis was performed with EVA
software.
Morphological and microstructural characterization was performed by Zeiss EVO50
Scanning Electron Microscopy (SEM) equipped with Oxford Inca EDS probe. Slags
were moulded in araldite-based resin, grinded and polished. SEM analyses were
carried out in backscattered electrons mode (BSE) in order to identify and check the
different phases pointed out by XRD. General and local chemical compositions were
measured exploiting EDS probe.
Slag leaching behaviour was investigated performing standard leaching test
according to EN 12457-2 (24 h in 10 l/kg deionized water stirred by rotatory mixer at
10 rpm) on 4 mm granulated slag. The slag dissolution rate will depend on the
surface-on-volume (S/V) ratio and on the liquid-on-solid (L/S) ratio. The
hydrodynamic conditions under varying conditions of flow and stirring also influence
the dissolution rate. For these reasons, three different experimental conditions were
employed:
-
powdered slags leaching behaviour were investigated varying the L/S ratio
(10, 100, 1000 l/kg). Weight loss, water pH and conductivity were measured
after each tests and the dried powders were investigated by XRD;
-
crushed slag particles (average dimension 4 mm) were leached in the same
condition as standard leaching test varying the L/S ratio (10, 100, 1000 l/kg).
Slags were characterized by SEM before and after elution tests in the same
areas to detect morphological alterations;
-
polished section of massive grains were leached in the same conditions as the
standard test and followed by SEM investigation in order to detect which
phases were dissolved.
Water analyses to determine chemical species concentration were performed by ICPOES.
32
Results and discussion
Microstructural characterization
XRD analysis performed on powdered slag allows us to identify the mineralogical
phases featuring the different samples (Figure 2).
As-received slags are characterized by large presence of wustite ((Fe, Mg,
Mn)O), larnite and small amount of Mg-Cr spinels. Traces of gehlenite and
brownmillerite are also detected in sample S1 and S2 respectively (Figure 2a).
SEM analyses confirm the results of XRD and allow to characterize morphology
and distribution of each structural constituent (Figure 3). As-received samples
are featured by high fraction of C2S, respectively about 37% (S1 sample) and
31% (sample S2), estimated by selective dissolution in methanol-salicylic acid
solution [19]. In samples S1, Mg-Cr spinels (phase C in Figure 3a) and gehlenite
islands (D) are finely dispersed in larnite matrix (B). Wustite (A) has the typical
dendritic form. In samples S2, wustite (A in figure 3c) appears thinner and more
dispersed than in slag S1 and surrounded by brownmillerite-type phases (E). In
both batch, chromium is mainly bound in spinel-like phases even if important
amounts are also detected in brownmillerite, larnite and wustite (Table 2). Vanadium and barium are mainly present in larnite and brownmillerite and in gehlenite in the sample S1. The presence of such dangerous elements in hydraulic
phases classifies the as-received slags as potential dangerous wastes.
The thermochemical treatment modifies the slags microstructure. As pointed
out in XRD pattern (Figure 2b), the admixed silica leads to another phase assemblage. In particular, larnite and brownmillerite react with SiO2 to form gehlenite. In fact, diffraction patterns pointed out very intense peaks typical of gehlenite-akermanite phases, whereas no more reflection of belite and brownmillerite are present. In MS2 samples secondary peaks are also detected, probably
associated to kirschsteinite that is present in solid solution with gehlenite.
33
Figure 2. XRD patterns of (a) as-received and (b) modified slag.
Morphologically, the microstructures appear homogeneous and constituted by
gehlenite matrix (D in Figure 3b-d) where wustite (A in Figure 3b-3d) and Mg-Crspinels (C) are dispersed. Probably, the stirring effect of slag-pot contributes to
homogenize the microstructure and to complete the diffusive reactions between the different phases. Wustite appears finer and thinner with pronounced
dendritic structures, whereas spinels are now aggregated in larger structures.
Chromium is completely fixed in these oxides whereas barium is totally migrated into gehlenite matrix. In this form the slag should be completely stable
and hind Ba, V and Cr releasing.
34
Figure 3. SEM-BSE micrographies of as-received (a-c) and modified (b-d) slags.
Table 2.Local chemical composition of phases pointed out by SEM-BSE analysis.
SEM-EDS Analysis (atomic*100)
S1 A
Mg
Al
0.2
0.0
9
1
S1 B
S1 C
S1 D
Si
Ca
Ti
V
Cr
Mn
Fe
Sb
Ba
0.0
0.0
0.1
1.4
0.00
4
1
9
1
2
0.9
1.7
0.00
0.0
0.0
0.0
0.00
0
3
1
0
2
1
1
0.3
0.2
0.0
0.00
0.4
0.1
0.1
0.00
9
1
3
1
9
0
9
2
0.0
0.4
0.4
0.8
0.0
0.0
0.4
0.00
7
4
7
2
0
6
3
4
Phas
e
W
L
Sp
G
35
S2 A
0.3
0.0
0.0
0.0
0.1
1.3
3
1
7
3
4
9
S2 B
S2 E
S2 F
W
0.0
0.8
1.7
0.0
0.00
0.0
0.0
0.0
0.00
5
3
0
1
1
0
5
1
1
0.2
0.1
1.2
0.1
0.00
0.0
0.0
0.2
0.00
5
1
1
0
4
1
1
4
6
0.5
0.2
1.4
0.0
0.0
0.0
0.00
4
0
0
1
5
1
3
0.1
0.0
0.0
0.0
0.00
0.0
0.1
1.5
1
1
2
1
1
1
6
4
0.2
0.2
0.0
0.00
0.4
0.0
0.2
0.00
2
7
2
1
2
9
8
1
0.0
0.2
0.5
0.8
0.0
0.5
0.00
7
8
9
8
6
0
1
0.0
0.3
0.8
1.1
0.00
0.0
0.1
0.0
0.00
8
5
2
4
1
1
1
1
3
0.1
0.0
0.0
0.00
0.0
0.1
1.5
2
2
2
1
1
3
6
0.2
0.2
0.0
0.00
0.4
0.0
0.2
0
3
1
2
3
9
9
MS2
0.2
0.2
0.0
0.0
0.00
0.4
0.0
0.2
C’
1
5
2
1
1
3
9
8
0.0
0.1
0.8
0.9
0.0
0.0
0.3
0.00
6
8
4
5
1
8
9
4
MS1 A
MS1 C
MS1 D
MS1 D’
MS2 A
MS2 C
MS2 D
L
B
CA
W
Sp
G
G
W
Sp
Sp
G
W: wustite; L: larnite; B: brownmillerite; CA: calcium aluminate; G: gehlenite; Sp: MgCr-spinel
Standard elution tests
Standard elution tests (Table 3), performed on 4 mm granulated slag, confirm the
slag behavior predicted through microstructural analyses. Focusing the attention only
on the more problematic environmental factors (Ba, V, Cr, pH), it is possible to note
that as-received slag have barium and vanadium release problems. Moreover the
water pH after test is very high, with value close to upper limits. The environment is
36
quite caustic and could produce several damage to vegetables, rocks and animals. In
these conditions slag use is hazardous without appropriate treatment, also for such
applications where slags are bound. In these slags no Cr release is detected.
Probably in this case, the spinel-like phases are stable and limit chromium leaching.
After the quartz treatment, pH sensibly decreases and Ba and V concentrations in
water are below the limits. The results obtained after standard tests are the evidence
of the inertization technique efficiency. However, to better understand which phases
are responsible to elements leaching and to confirm the complete stability of modified
slag, further investigation were performed on slag powder varying the water/slag
ratio.
Table 3.Standard elution test results of as-received and modified slag.
Leaching test (EN 12457-2) on 4 mm slag: 24h deionized water 10 l/kg
Element
Units
S1
MS1
S2
MS2
Limits
Ba
mg/l
0.80
0.68
4.97
0.14
1
V
µg/l
136
48
n.d.
33
250
Cr
µg/l
n.d.
n.d.
n.d.
n.d.
50
11.9
10.7
11.4
10.6
5.5-12
487
315
1.800
206
pH
Conductibility
µS/cm
Leaching test with different liquid-to-solid ratio
Liquid-to-solid ratio is one of the fundamental factors that rules the interaction
between slag and environment. The concentration of elute substances depends on
solution pH, on their solubility and on the kinetics of dissolution. The increase of L/S
ratio should allow to maintain pH more acid and contemporary to work far enough
from saturation limit [20]. The analyses were focused not only on dangerous heavy
metals but also on the main elements constituting the different phases (Al, Ca, Mg,
Si) in order to correlate them with the slag leaching behavior.
Table 4 summarizes the obtained results expressed in mg/kg in order to compare the
different L/S ratios. The results for S1 sample are not presented in this study because
will be discussed in another paper, but comply with S2 sample.
Effects of liquid/solid ratio are quite evident for all the samples: increasing the water
volume the dissolved slag amount increase too. However, the effect on modified slag
37
is sensibly less than on as-received ones. In fact, the weight loss of modified slag is
of an order of magnitude lower than non-modified samples as well as the calcium and
aluminum release.
Table 4. Elution test results on powdered slag as a function of water/slag ratio.
Sample
S2
L/S Weigh
[l/kg t loss
]
%
10
0.83
100
7.73
1000 12.69
MS1
10
0.87
100
3.05
1000 7.78
MS2
10
0.87
100
2.30
1000 3.30
Cond.
Mg
Al
Si
Ca
Ba
Cr
pH [ S/cm [mg/kg [mg/kg [mg/kg [mg/kg [mg/kg [mg/kg
11.5
8
11.5
0
11.2
6
10.2
0
10.0
0
9.20
10.5
1
10.4
8
9.75
]
]
]
]
]
]
]
1847
1
3733
19
4154
145
1
1063
29
3400
50
2500
120
2
232
315
10800
800
33500
112
7
310
18
20
330
300
3
0.1
126
67
140
800
2200
10
1
37
370
1600
4000
10000
40
10
260
39
43
170
320
2
0.3
99.3
66
100
400
1500
7
1
54.1
123
1800
2000
8100
27
1
38
Figure 4. XRD patterns of (a) S2 and (b) MS2 leached slag at different L/S ratio
In silica treated slag, calcium is totally bound in gehlenite and its low content in elute
is index of C2AS poor hydraulic phase as observed in cement [21]. Aluminum, in
modified slag, is present both in Mg-Cr-spinel and in gehlenite. Its low amount in
water is synonymous of higher spinel stability (and this behavior is confirmed by the
very low concentration of Cr in water).
Another interesting aspect is the correlation between magnesium and chromium in
the leachate. High concentration of leachate Mg seems to be accompanied by higher
content of Cr. This behavior could be attributed to low stability of spinel-like phases.
The hypothesis, stressed on the basis of elution tests results, are confirmed by XRD
analysis performed on leached powdered slag (Figure 4). In order to compare the
different spectra, normalization on more intense wustite peak is performed. Wustite is
selected because it does not react with water [222]. In as-received samples (in
39
particular S2) the C2S and C4AF peaks decrease their intensity increasing the L/S
ratio (Figure 4a). Also spinel peaks undergone intensity reduction. The different XRD
spectrum show that, using standard L/S ratio (10 l/kg), the interaction between slag
and water is very limited and probably interest only few material portions. This
conjecture is confirmed by SEM observation on fractured slag surface (Figure 5a).
Increasing the L/S ratio, the fraction of C2S dissolved rapidly increase whereas
Brownmillerite dissolution is clear evident only at higher water ratios (Figure 4a and
Figure 5b-c). On the modified slag, diffraction patterns demonstrate the slag stability.
Although the solid-to-liquid ratio increases, no modification in peaks intensity is
pointed out.
SEM analyses are in agreement with diffraction analyses and confirm the absolute
integrity of gehlenite matrix (Figure 5d-e-f).
40
Figure 5. SEM microgrpahies of leached slag at different L/S ratios: sample S1
at (a) 10, (b) 100, (c) 1000 l/kg and sample MS1 at (d) 10, (e) 100, (f) 1000 l/kg.
Elution test on polished section
In order to identify more precisely which phases are interested by water attack,
leaching test simulation is also performed on polished section (Figure 5d-h).This
procedure is needed because BSE probe contrast was not enough to unambiguously
41
identify the different phases on massive slag grains.
Polished slags were immerse in 350 ml of deionized water for 24 h and then
observed by SEM in the same areas investigated during the morphological
characterization (Figure 3). In this case the L/S ratio is extremely high (tends to
10000 l/kg) probably allowing to maintain water under saturated with respect to
hydraulic phases and avoiding pH modification.
These analyses clearly indicated that larnite (C2S) is the most soluble phase in the
water. However, high amount of water is required to dissolve it completely (Figure
6a). Brownmillerite is only partially eroded by water even if its hydration seems to be
enough to dissolve amounts of chromium and barium that contribute to increase the
total leachate content in water.
On the other hand, modified slags remain completely unaltered despite the very high
liquid ratio, demonstrating the inertization process reliability (Figure 6b).
Figure 6. SEM microgrpahies of leached slag on polished section: (a) S1 and
(b) MS1 samples.
Conclusions
In conclusion, efficiency of inertization treatment performed by quartz addition on
EAF slag is experimentally demonstrated.
- XRD and SEM analysis allow to identify which phases are more soluble in water,
ruling the release of dangerous chemical species. Larnite is the most soluble phase
followed by brownmillerite. Gehlenite-type phases have proved to be complete
insoluble;
- slag use without any stabilization treatment could lead to dangerous environmental
42
consequences, even in the case that the leaching limits are observed. In fact, on long
term scenario, the pH alteration due to cyclic carbonation phenomena could enhance
the leaching of dangerous heavy metals;
- the reaction between added SiO2 and other silicates characterizing the slag
microstructures promotes the formation of Gehlenite-type phase;
- Gehlenite is completely stable and assure safe and inert behavior to modified slag;
- modified slag stability is demonstrated through leaching test featured by different
L/S ratios;
- modified slag could be used both for unbound applications and as aggregates
without polluting humans and environments.
Acknowledgments
Authors want to acknowledge Simone and Umberto Di Landro (DILAB labs, Crema
(CR), Italy) and Thierry Vincent (École des mines d’Ales, C2MA, Alès, France) for
ICP-OES analyses.
References
[1] Sorlini S., Sanzeni A., Rondi L. Reuse of steel slag in bituminous paving mixtures.
Journal of Hazardous Materials 2012; 209–210:84–91
[2] Internet reference http://www.euroslag.com/products/statistics/
[3] Zheng G-H. and Kozinski J.A. Solid Waste Remediation in the Metallurgical
Industry: Application and Environmental Impact – McGill Metals Processing Center.
Department of Mining and Metallurgical Engineering. McGill University
[4] Moltz H. and Geiseler J. Products of steel slags an opportunity to save natural
resources. Waste Management 2001; 21:258-293
[5] Chaurand P., Rose J., Briois V., Olivi L., Hazemann J-L., Domas J., Bottero J-Y.
and Proux O. Environmental impacts of steel slag reused in road construction: A
crystallographic and molecular (XANES) approach. Journal of Hazardous Materials
2007; B139:537–542
[6] Del Fabbro M., Stefanutti M., Ceschia C. Impiego di derivati delle scorie di forno
ad arco elettrico come Materiale eco-compatibile nella sovrastruttura stradale. Proc.
Of XI S.I.I.V Conference. Verona (Italy). 28–30 November 2001
[7] Lind B.B., Fallman A-M. and Larsson L.B. Environmental impact of ferrochrome
slag in road construction. Waste Management 2001; 21:255-264
43
[8] Van Zomeren A., van der Laan S.R., Kobesen H.B.A., Huijgen W.J.J. and Comans R.N.J. Changes in mineralogical and leaching properties of converter steel slag
resulting from accelerated carbonation at low CO2 pressure. Waste Management
2011; 31:2236–2244
[9] Barella S., Gruttadauria A., Magni F., Mapelli C. and Mombelli D. Survey about
Safe and Reliable Use of EAF Slag. ISIJ International 2012; 52(12):2295-2302.
[10] Mombelli D., Mapelli C., Gruttadauria A., Baldizzone C., Magni F., Levrangi P.L.
and Simone P. Analisys of Electric Arc Furnace Slag. Steel Research International
2012; 83(11):1012-1019. [11] Gelfi M., Cornacchia G. and Roberti R. Investigations
on leaching behavior of EAF steel slags. 6th European Slag Conference EUROSLAG
2010, Madrid (Spain).
[12] De Windt L., Chaurand P. and Rose J. Kinetics of steel slag leaching: Batch
tests and modeling. Waste Management 2011; 31:225–235
[13] Faellman A-M. Leaching of chromium and barium from steel slag in laboratory
and field tests - a solubility controlled process?. Waste Management 2000; 20:149154
[14] Forschungszentrum Karlsruhe Umweltforschungszentrum Contaminated Soil
2000 Vol.2. Proceedings of Seventh International FZK/TNO Conference on Contaminated Soil Leipzig (Germany) September 2000
[15] Indipendet Evironmental Technical Evaluation Group (IETEG). Chromium (VI)
Handbook. CRC Press. Baton Rouge (LA). 2004
[16] Internet reference:
http://www.sdasr.unict.it/materiale/8Alterazione%20di%20minerali%20e%20rocce.pdf.
[17] Drissen P., Ehrenberg A., Kühn M. and Mudersbach D. Recent Development in
Slag Treatment and Dust Recycling. Steel research international 2009; 80(10):737745
[18] Levin E.M., Robbins C.E. and McMurdie H.F. Phase Diagrams for Ceramist, Vol.
1, The American Ceramic Society, Columbus, OH, USA, (1964).
[19] Klemn W.A. and Skalny J. Selective dissolution of clinker minerals and its applications. Martin Marietta Technical Rep 1977;77:26.
[20] Nicoleau L., Nonat A. and Perrey D. The di- and tricalcium silicate dissolutions.
Cement and Concrete Research 2013; 47:14–30.
44
[21] Pöllmann H. Mineralogy and crystal chemistry of calcium aluminate cement. Proceedings of the International Conference on Calcium Aluminate Cement, Edinburgh,
July 2001. IOM Communications Ltd, London, 2001
[22] Belhadj E., Diliberto C. and Lecomte A. Characterization and activation of Basic
Oxygen Furnace slag. Cement & Concrete Composites 2012; 34:34–40
Dr.-Ing. Horst Kappes, Head of B.A. Energy & By-Products
MSc ETH Daniel Michels, Process Engineer, B.A. Energy & By-Products
45
Theme 2
Metallurgy and Processing
46
H.
Schliephake,
B.
Dettmer,
K.
Schulbert,
T.
Zehn,
T.
Rekersdrees
1)
P. Drissen, D. Mudersbach 2)
Sustainable stabilisation and reuse of ladle furnace slag from electric steelmaking
1) GMH GmbH, Georgsmarienhuette, Germany, 2) FEhS ‐ Institut für Baustoff‐Forschung e.V., Duisburg, Germany Abstract
In 2012 the total production of slags from electric steelmaking in Germany was more
than 2 M tonnes for both quality and high alloyed steel production. Worldwide the
production of electric arc furnace (EAF) and ladle furnace (LF) slag has been estimated for 2012 together at 40-50 M tonnes. Today EAF slag in Europe is well characterised and used as construction material. There is vast experience in using EAF
slag as aggregates for different applications, but the utilisation rate of LF slag could
be improved. The target treatment of the liquid LF slag to stabilise the dicalciumsilicate disintegration can be a better internal recycling or an increased external utilisation.
This paper summarises the efforts of the European steel works concerning stabilisation of naturally disintegrating LF slag and the different ways of internal recycling in
the electric arc furnace, as liquid, disintegrated or as stabilised material. In focus are
especially different test trails and research work at the steel work of Georgsmarienhütte. Investigations to stabilise the LF slag by treatment of the liquid slag with
different slag conditioners or by fast solidification will be presented. On the other
hand the stabilised LF slag has been successfully reused in the EAF as lime substitute.
Introduction
Iron and steel making can be divided into primary processes, like crude iron production in blast furnace process or production of crude steel in BOF- and EAF-
47
steelmaking, and secondary metallurgical operations, aiming at the final adjustment
of steel qualities and temperature for casting.
Slag qualities generated in primary iron and steel making are almost entirely used.
Slag qualities generated in secondary metallurgy, referred to as ladle furnace slag,
often reveal specific properties that restrict processing and utilisation:
 Ladle furnace slag tends to disintegrate during cooling of the already solid slag
and the resulting fine grained material causes dust pollution during processing,
handling and transport.
 During processing they tend to build-up deposits and cause clogging of the sieving screens.
 As fine grained construction material they are competing with numerous inexpensive products of other origin, including natural materials.
 Occasionally due to large specific surface area trace elements with environmental
relevance are released, which might cause further limitations in utilisation.
Reason for the disintegration of the ladle furnace slag already during cooling is a
change in modification of the mineral dicalciumsilicate (C2S). Dicalciumslicate is a
common constituent in ladle furnace slag, due to the adjustment of slag composition
for metallurgical reasons. During cooling down the -modification (larnite) transforms
to the -modification (calcio-olivine) at temperatures below 500 °C. The change in
modification is accompanied by an increase in volume of about 12 %. As -C2S crystals are uniformly spread in the slag the texture of slag aggregates is disturbed and
the entire slag decomposes into fine grained material already in the slag pot or in the
slag yard [2,3,4].
48
To overcome this detrimental phase transformation there are two options for slag
stabilisation:
-
conditioning of the -modification of C2S by additions
-
accelerated cooling of the slag within a distinct temperature range
Conditioning
Ladle furnace slag qualities that are prone to disintegration can be stabilised by introducing distinct elements, e.g. boron or phosphor, in the crystal structure of C2S. In
this case the -modification of C2S remains stable down to ambient temperature
without phase transformation. Stabilisation usually is achieved with concentrations of
less than 1 wt.-% B2O3 or P2O5 in the slag.
Prevention of slag disintegration due to phase transformation of C2S by addition of
boron or phosphor is a well known and an efficient method [4]. A successful stabilisation requires a homogeneous liquid slag of low viscosity. Otherwise no uniform and
homogeneous distribution of boron or phosphor in the slag is achieved and the slag
will disintegrate during cooling, at least partly.
In daily operation of steelmaking this approved method often fails. For example a
high viscosity of the slag, an insufficient temperature range for treatment or a partly
solidified slag because of some delay in operation can cause an insufficient treatment. The method has to be adapted specifically to given operational routines in
steelmaking.
Accelerated cooling
A rapid cooling of slag results in the formation of small crystals or ideally in a glassy
structure of the solidified slag. In both cases the disintegration of slag is blocked.
However, slag qualities with a high ratio of CaO/SiO2 as steel slags hardly solidify in
a glassy state. Due to the bad thermal conductivity of lime-silicate in slag qualities the
heat removal is very bad. Consequently it is just the outer surface of the slag that is
49
stabilised in operational practice. The total amount of fines is reduced, but disintegration is not totally prevented and dust formation is still possible.
Stabilisation of ladle furnace slag in carbon steel making has been investigated by
the FEhS-Institute in collaboration with the steel shop of Georgsmarienhuette GmbH
aiming at minimisation of dust formation and use of coarse ladle furnace slag as aggregates in construction works as well as the internal recycling of ladle furnace slag
as flux material in the electric arc furnace.
Stabilisation of ladle furnace slag in carbon steel making
Stabilisation was tested in laboratory and under operational conditions by conditioning of the slag with additions as well as by accelerated cooling of the slag.
Conditioning of liquid ladle furnace slag Conditioning of slag by additions aimed at the prevention of dust formation and the
use of coarse aggregates as construction material. As additions some commercially
available materials have been tested (see Table 1).
Conditioner
B2O3
P2O5
[wt.-%]
[wt.-%]
1
Vitribore, Stollberg/DAMET
30
-
2
Glassrock, Stollberg/DAMET
20
10
3
Vitroc, Rio Tinto
68
-
4
BoSi, Schirmbeck
10
-
5
BOOR, AFFILIPS
46
-
6
CPS, AFFILIPS
-
24
Table 1:
B2O3- and P2O5-containing additions used for stabilisation
These conditioners are siliceous materials with different concentrations of B2O3 or
P2O5. There is no risk of explosion when added to the hot liquid ladle furnace slag,
50
because chemically bound H2O and CO2 are not present. All materials have grain
size well below a few millimetres, which should enhance the dissolution in the liquid
slag.
Prior to plant trials a series of lab-trials was performed to check the amounts necessary for stabilisation and to identify potential problems with dissolution.
Under lab-conditions there was no problem with dissolution of any of the materials
tested. Partial stabilisation was achieved already with addition of about 0.2 wt.-%
B2O3. Complete stabilisation was reached by addition of at least 0.5 wt.-% B2O3. Figure 1 shows examples of re-melted ladle furnace slag. On the left a re-melted slag
without any addition is shown. Due to controlled slow cooling this untreated slag is
disintegrated by the phase transformation of C2S. By addition of 0.5 wt.-% of conditioner 1 (Figure 1, centre) and 1 wt.-% of conditioner 2 (Figure 1, right) disintegration
was blocked at the same boundary conditions. Some fines of the later samples are
due to the mechanical processing of the solidified slag.
Figure 1:
Disintegrated untreated slag (left) and stabilised slag (centre and right)
X-ray diffraction confirmed, that disintegration was due to the formation of -C2S (calcio-olivine) whereas in stabilised slag -C2S (larnite) crystallised during cooling (Figure 2).
51
Impulse
800
G:\Roentgen\XRD 2008\9759.xrdml
G:\Roentgen\XRD 2008\9760.xrdml
disintegrated,
with calcio-olivine
600
stabilised with boron,
with larnite
400
200
0
31
31,50
32
32,50
33
Position [°2Theta] (Kupfer (Cu))
Reflexliste
01-077-0388; Ca2 Si O4; Larnite
01-077-1113; Ca2 Al ( Al Si O7 ); Gehlenite, syn
01-070-2450; Ca2 Si O4; Calcio-olivine, syn
Figure 2:
Cut-out of superposed X-ray diagrams, showing -C2S (calcio-olivine) of
disintegrated (black) and -C2S (larnite) for stabilised ladle furnace slag (blue)
Based on the results of the lab-experiments at FEhS-Institute operational trials started in the steel shop of Georgsmarienhuette GmbH. The ladle furnace slag at
Georgmarienhuette GmbH is still liquid after tapping of steel into the continuous caster and homogeneous addition of conditioners is possible during slag transfer from the
steel ladle to the slag pot.
First series were performed by manual addition of conditioners 1, 2 and 3 (see Table 1). Assuming an average amount of ladle furnace slag of 2 t per heat an addition
of about 30 kg of conditioner 1, corresponding to 0.5 wt.-% B2O3, should be sufficient
for stabilisation. This addition already reduced the visual detectable amount of disintegrated slag and dust formation. In a second step the amount of conditioners was
increased to 60 kg to further prevent the disintegration of the LF slag. The addition of
60 kg of conditioner corresponds to concentrations of 1 to 2 wt.-% of B2O3 and no
disintegration or dust formation was observed in the slag yard.
Despite this success the manual conditioning comprises some disadvantages in operational practice like the required manpower for handling and refilling into easy to
handle bags. For this reason a couple of series with pneumatic injection of conditioners have been tested. Conditioners were injected with pressurised air via a lance into
the flowing stream of slag during slag transfer from the steel ladle to the slag pot
52
(cold test: Figure 3). Another advantage of this technique is the uniform distribution of
conditioners in the ladle furnace slag compared to manual addition.
But in both cases there is a discrepancy between the need of slow tapping of liquid
slag for sufficient treatment and the need of fast tapping of the slag to avoid incrustations of solidifying slag at the walls of the transfer ladle.
Despite the better method of slag treatment by injection the complete handling of this
treatment was still tricky. The method has to be adapted specifically to given operational routines in different steel works.
53
Figure 3: Injection of conditioner into the slag pot
(picture for technical reason without slag),
during the tests liquid slag came from left side
In general all conditioners in the tests could be injected pneumatically, even in trials
with high injection rates. With the addition of up to 120 kg of conditioners, corresponding to approximately 9 wt.-% of the mass of the slag, no negative effects on
slag viscosity or dissolution of conditioners have been observed.
Table 2 exemplarily shows results of a test campaign with conditioner 5 (see Table 1). Shown are the amount of addition, concentrations of B2O3 and P2O5 after stabilisation and, the presence of -C2S (larnite) or -C2S (calcio-olivine) and a remark
concerning the observed disintegration.
Amount of
B2O3
P2O5
conditioner 5
[wt.-%]
[wt.-%]
5 (referene)
0
< 0,01
0,02
5.1
30 kg/heat
0,03
< 0,01
5.2
30 kg/heat
0,07
5.3
40 kg/heat
5.4
series:
-C2S
-C2S
(larnite)
(calcioolivine)
Disintegration
X
yes
(X)
X
yes
0,01
(X)
X
yes
0,28
< 0,01
(X)
(X)
partly
40 kg/heat
0,33
< 0,01
(X)
(X)
partly
5.5
40 kg/heat
1,57
< 0,01
X
no
5.6
50 kg/heat
1,21
0,03
X
no
Table 2:
Characteristics of injection trials with conditioner 5
(X := main mineral constituent, (X) := minor mineral constituent)
The analytical results of all operational trials confirmed the findings of the lab-tests, a
successful stabilisation is achieved by adjusting a concentration of min. 0.5 wt.-%
B2O3 in the ladle furnace slag. The pneumatic injection is a suitable technique for
safe and precise addition of conditioners.
54
Accelerated cooling of liquid ladle furnace slag As a second option to prevent the disintegration of ladle furnace slag the accelerated
cooling has been tested in operational practice in the steel shop of Georgsmarienhuette GmbH, too. Accelerated cooling of slag is achieved by tapping slag into
large flat layer with correspondingly high surface area to improve the heat transfer to
air and floor under the slag. First trials with thin layers in the slag yard were not successfully due to insufficient heat transfer in to the ground (solidified slag). To improve
the heat transfer in a second step the liquid ladle furnace slag was tapped on steel
plates (Figure 4). Tapping without any delay and adjustment of thin slag layers turned
out to be the key parameters for successful stabilisation and prevention of dust formation. Slag layers of more than 10 cm in height are detrimental with respect to stabilisation. But the maximum width of the slag layer depends from the composition of
the ladle furnace slag.
Figure 4:
Pre-tests for accelerated cooling of ladle furnace slag in the steel shop
Internal recycling of ladle furnace slag
There are different ways of internal recycling of secondary metallurgical slag in the
electric arc furnace, as liquid, disintegrated or as stabilised material.
55
Recycling of stabilised ladle furnace slag
The test trials at Georgsmarienhuette GmbH have shown that in contrast to stabilisation with conditioners, accelerated cooling has economical advantages as there is no
need to purchase conditioners, no additional handling or technical equipment needed
for injection. After accelerated cooling the solidified, stabilised ladle furnace slag can
be recycled internally into the electric arc furnace as flux because it has no boron or
phosphor, but is still rich in CaO and MgO. This way it will serve as a substitute for
virgin lime or dolomite, creating another economic benefit. But accelerated cooling
will require additional space for tapping in thin but large layers at the steelworks with
comparably low temperature of the steel plates throughout operation. Stabilised slag
by accelerated cooling has been successfully reused in EAF at Georgsmarienhuette
GmbH, without negative effects to metallurgy and economy.
Recycling of disintegrated ladle furnace slag
Another option is the recycling of the disintegrated slag by injection directly into the
EAF without any treatment of the liquid slag. For this a complex device for processing
is necessary (Figure 5).
Figure 5: Processing and recycling of ladle furnace slag at PITTINI Ferriere Nord
Since many years this method has been in operational practice at the steel work of
PITTINI Ferriere Nord SPA. First step is the tapping of liquid slag in an open slag
yard with the hot slag transported by trucks in closed boxes. The disintegrated slag is
56
separated from metal and is sieved; the fine powder is collected in a silo and finally
injected into the EAF. The complete device needs an expensive dedusting system.
Another option is the external manufacturing of the disintegrated ladle furnace slag
with a binder or under high pressure into stones and charging these stones with the
scrap basket into the EAF. This method is the objective of an ongoing research project of FEhS-Institute and Georgsmarienhuette GmbH.
Recycling of liquid ladle furnace slag
Examples of European steel works which recycle liquid slag from secondary metallurgy into the EAF in order to utilise the melting heat of the ladle furnace slag are RIVA Acciaio (Verona Works) in Italy or Metal Ravne d.o.o. in Slovenia. Metal Ravne
manufactures long steel products including high speed steel, special steel, construction steel and tool steel products. Its UHP 40 t EAF is mainly charged with ferrous
and Ni containing scrap. The EAF is situated directly besides the ladle furnace decreasing problems associated with the transportation of the liquid ladle furnace slag
to the EAF. Recycling of liquid slag requires a special infrastructure and excellent
logistical organisation in the steel shop. The Verona Works of the RIVA Group has
the necessary infrastructure to perform liquid slag recycling.
After slag free tapping at RIVA Verona works, the EAF top slag is put into the steel
ladle. In the ladle furnace the slag is adjusted. At the end of casting on the strand
caster, the ladle is emptied into the EAF if possible, otherwise emptied into a slag
yard, in order to fit the ladle for new steel filling.
Preferably ladle slag with high concentrations of CaO and MgO but low concentrations of SiO2 should be recycled. Calcium oxide and MgO are welcome as slag formers. However, during recycling of the liquid ladle slag, the slag has to be recycled
immediately to avoid a drop of temperature under the liquidus temperature. Making
the separation of different slag qualities not possible.
Ladle furnace slag is mainly composed of CaO, MgO, Al2O3 and SiO2. During cooling
in the slag yard the ladle furnace slag at Verona Works usually disintegrates due to
phase transition of dicalciumsilicate. Handling of solidified ladle slag might cause
problems during handling and storage caused by dust emissions. Further disintegration is caused by hydration of free lime and free MgO in contact with moisture of air
later in the slag yard.
57
Recycling of liquid ladle furnace slag requires special infrastructure and good logistical managment in the steel shop. Under the specific conditions given at RAVNE the
recycling rate is 100 % and at RIVA the recycling of at least 80 % of ladle slag is possible without negatively influencing the steel quality and steelmaking. Since the ladle
furnace slag is still liquid no problems were seen with the dissolution in the EAF process. Changes in chemical and mineral composition of the EAF slag are not detectable. The additional energy demand for the recycling is low compared to the total energy consumption in EAF steelmaking. Most of this surplus in energy is necessary to
compensate for the energy loss of heat radiation during roof opening. On the other
hand the recycling of liquid ladle furnace slag in the EAF means the direct recycling
of the remaining steel in the ladle furnace slag into the EAF, leading to better economical benefits of the complete steel making process.
Conclusion
Avoiding the disintegration of ladle furnace slag can be achieved by addition of conditioners with boron or phosphor or by accelerated cooling. The solidified slag can be
processed into aggregates with comparable technical and environmental characteristics found in EAF slag that is used in road making. Some restrictions with respect to
volume stability exist due to the presence of free MgO in the ladle furnace slag that is
not affected by conditioning or accelerated cooling. In case of higher additions of boron some restrictions might be seen due to forthcoming regulations on the leaching of
boron, which could be a limiting factor for the utilisation of the treated slag as construction material.
As no conditioners are required in stabilisation by accelerated cooling the internal
recycling as flux in EAF steelmaking is another option. Benefits are seen in the substitution of virgin lime or dolomite by ladle furnace slag. A limited substitution of lime
and dolomite by ladle furnace slag in a ratio 1 to 2 revealed no metallurgical disadvantages as was shown in a former research project [5] and during the current investigations at Georgsmarienhuette GmbH.
Other options for internal recycling of ladle slag in the EAF are the direct addition of
liquid slag or slag stones by scrap basket and the injection of solidified slag powder.
The best technique for the internal recycling of ladle furnace slag depends on the
58
special boundary conditions in the steel works. If these requirements are overcome,
economical and ecological benefits could be achieved with this procedure.
References
[1]
Elstner, I., Leers, K.-J., Niesel, K.: Untersuchungen zum Dicalciumsilikat-
Zerfall von Hüttenschlacken, Tonind.-Ztg. 94 (1970), Nr. 8, S.317 - 331
[1]
Drissen,
P.,
Arlt,
K.-J.:
Entstehung
feinkörniger
Pfannenschlacken
Report des Forschungsinstituts (2000) 7. Jahrgang, Nr. 2, S. 12
[1]
Mudersbach, D. et al.: Stabilisierung von zerfallsverdächtigen Edelstahlschla-
cken
Report des Forschungsinstituts (2006) Nr. 2, S.4/5
[1]
Drissen, P. et al.: Efficient Utilisation of Raw Materials Used in Secondary Metallurgy
as Flux in Steelmaking Furnaces
European Commission - Executive Committee C1, Contract No. 7210-PR-203, Final report
2003
59
I. J. McDonald and A. Werner
Dry Slag Granulation – The Environmentally Friendly Way to Making Cement
Siemens VAI Metals Technologies, 7 Fudan Way, Thornaby, TS17 6ER, UK
Siemens VAI Metals Technologies, Turmstraße 44, Linz, 4031, Austria
Abstract
Each year approximately 400 million tons of blast furnace slag are produced worldwide. The slag, which has a tapping temperature of around 1,500°C, is normally used
as a substitute for cement clinker or as an aggregate material in road construction.
The current state-of-the-art practice is to granulate molten blast furnace slag in wetgranulation plants using large volumes of water. However, up until now it has not
been possible to utilize the remnant heat energy of the molten slag, which amounts
to approximately 1.5 GJ of energy per ton of slag.
In an R&D project currently underway by a consortium of companies comprised of
Siemens Metals Technologies, voestalpine Stahl GmbH (Austria), ThyssenKrupp
Steel Europe AG (Germany), the FEhS Building Materials Institute (Germany) and
the University of Leoben, Austria, a new technology based on dry-slag-granulation is
being investigated to use air to cool molten slag and to recover the remnant heat energy for heating applications or for the generation of electrical energy. At the same
time, the slag product should also fulfill the same criteria as wet-granulated slag for
use in the cement industry.
The project was officially launched on September 1, 2011. A technical plant was set
up at the testing facilities of the University of Leoben, and in the summer and autumn
of 2012, a series of dry-slag granulation campaigns were carried out using remelted
blast furnace slag. The slag product quality and the elevated offgas temperatures
proved very promising and the successes of these campaigns will form the cornerstone of the next site based activities.
60
The development of this plant is now underway and is scheduled for installation in
the middle of 2014.
Substitution of Cement Clinker with Slag Sand
Traditional manufacturing of cement clinker from limestone, sand, clay and other
components requires a high-temperature process (around 1450°C). It is also associated with high demand for raw materials, high input of primary energy and high specific CO2 emissions (roughly 1 t of CO2 per ton of clinker). The substitution of cement
clinker by blast furnace slag sand is an attractive economic alternative for the cement
industry, because it reduces high energy costs and considerably improves the company’s CO2 balance. Approximately 1 ton of CO2 can be saved for each tonne of
clinker substituted by slag sand because not only primary energy is saved, but also
the release of the carbon dioxide chemically bound in the limestone is avoided.
Conventional Granulation Technique for the Production of Slag Sand
In this case the slag is quickly "quenched" in granulation plants using large quantities
of water, producing a fine-grained, amorphous but also wet product, known as slag
sand. Due to the "frozen" crystallization energy, the slag sand when ground to cement fines, form hydration products in conjunction with water (latent hydraulic behaviour). These products essentially correspond to the hydration products of Portland
cement clinker, the main component of Portland cement. The key prerequisite for the
use of slag sand as a binding agent in the building material industry is thus satisfied.
Therefore approximately 80% of blast furnace slag sand is used as cement additive
and realises valuable revenue rather than being disposed of as land-fill.
The wet granulation process operates with a high water to slag ratio of about 8:1.
This wet process is not susceptible to any fluctuations in the quantity and properties
of the slag. Furthermore the wet process has the following drawbacks:
 Despite mechanical dewatering in drums, silos or heaps, a residual moisture of
about 10 - 12 % moisture remains in the slag sand. For the manufacturing of cement, the product therefore first has to be re-dried, with high energy expenditure.
Assuming 10 % residual moisture, the required drying energy amounts to around
132 kWh/t.
61
 For granulation with open water circuits, vapour containing sulphur can be released, and a correspondingly large amount of fresh water (about 1 m³/t) has to be
fed into the system. Granulation plants with closed water circuits and condensation
systems prevent the emission of water vapour containing sulphur.
 When slag is quenched with water, the high energy potential of liquid slag is wasted to heat and evaporated water. For granulation, cold water is normally used, the
circulated water has to be cooled in cooling towers, which are in some cases
equipped with electrically operated fans. Finally the heat is released to the environment at a low temperature level without being used.
Alternative Technique for Producing Vitreous Blast Furnace Slag
Huge amounts of water and of drying energy can be avoided by dry dispersion and
quick cooling of the liquid slag. The essential prerequisite for the introduction of an
alternative dry technique is that the obtained product needs identical or even better
properties compared to the slag sand produced conventionally using wet granulation.
This applies in particular to the glass content (target > 95%), which is a key parameter for the reactivity and hence the quality of the slag sand.
The glass content has a direct impact on the strength of the cements and concretes.
However, the required glass content can only be achieved by sudden cooling below
the transformation temperature of approximately 900°C. Due to the less efficient
cooling mechanism of water-free quenching, the dry process is technically more challenging than conventional water based granulation.
Obviously “dry” granulation requires no subsequent drying of the product. This leads
to a CO2 reduction of roughly 30 kg/t in comparison with wet process. Given global
production of approximately 210 million t of slag sand (2007), this is equivalent to a
potential CO2 reduction of over 6.3 million t per year.
There are two methods in which slag can be fed to any slag granulation plant. The
first involves a granulator which is located close to the furnace and enables the slag
to be delivered to the plant direct from the slag runner. The second is a system remote from the furnace involving the transfer of slag via slag pots and pouring the slag
into a granulator via a ladle tilter.
62
Granulation of slag running directly from a blast furnace is technically more challenging than slag pot delivery of slag where, in principle, the slag flow rate can be regulated. Consequently, there are two basic granulator formats, one for blast furnace
slag with direct runner slag delivery and the other, a less complex design for any slag
delivered via slag pots.
The first module of the DSG process can offer a very simple low cost option for slag
disposal with only the spinning cup and slag runner being present. The granules can
be roughly strewn by the cup and gathered using a loader after each tapping. There
would be no possibility for heat recovery using this simple method, but a cement
grade granulate can still be produced.
The second and more practical solution bearing in mind the normal lack of space
available around a blast furnace is to confine the atomising granulate inside a container. Circulating air is required at this point which leads to the opportunity of waste
heat recovery through the increased off gas air temperature.
The Concept
Hot Air to Chimney
Slag Runner
Spinning cup
Static water jacket
Modified Fluidised Bed
Cooling air in
Granulated Slag
Main drive shaft & bearings
Figure 1. Dry Granulation Concept
Dry slag granulation is based on molten slag atomisation using a variable speed rotating cup or dish (see Fig.1). The slag is delivered on to the centre of the cup from a
slag runner via a vertical refractory lined pipe. The rotation of the cup forces the slag
63
outwards to the cup lip where it is atomised (see Fig 2). The resulting slag droplets
cool in their flight towards the water jacketed chamber wall. On impact with the wall,
the droplets are sufficiently solid to ensure they do not stick to the wall. This characteristic is further enhanced by the presence of the water jacket.
The solidifying granules fall into a mobile bed of granules that is designed to ensure
that there is no agglomeration. The bed is kept in motion by the design of the cooling
air distributor that imparts a circumferential motion to particles.
The cooling particles fall into a discharge trough that forms an inner annulus. Some
are recycled to intercept the solidifying particles in flight from the cup to assist in their
cooling. The remainder are further cooled as they are blown towards discharge ports
and thence on to conveyors for transport to storage.
Some particles are lifted out of the cooling bed and scour the chamber walls, further
reducing the possibility of solidifying droplets, in flight from the cup, sticking on the
walls. Any carryover of particles in the cooling air is minimised by flow straighteners
in the upper levels of the chamber that are designed to reduce velocities in this region. The air is finally discharged via a stack or stacks. For blast furnace slag applications, slag wool arresters and collectors are included.
Figure 2. Atomising Slag
64
The Current R&D Focus by SVAI on the Subject of DSG
With environmental and energy saving considerations becoming ever more important
and even becoming enshrined in legislation, there is clearly a need for a major improvement in slag handling.
Our past experience of the dry granulation process is being further enhanced with
heat recovery trials to satisfy this requirement and is now a major R&D project at
SVAI in conjunction with industrial partners.
The process of atomising slag using a spinning cup has been suitably proven by
SVAI (formerly Davy and Kvaerner) over the years and it is known the granulate produced by this method at over 95% glassy is suitable for use in the cement industry.
The challenge and focus for us now is to produce the same granulate at air temperatures which are high enough to make the process suitable for waste heat recovery.
Trials are now being conducted on a mini granulation plant located at the University
of Leoben in Austria (see Fig. 3). The results gained so far have been very encouraging and have matched the CFD calculations for product size and Offtake temperatures made at the beginning of the project (see Fig 4).
Figure 3. Mini Granulator at Leoben
Figure 4. CFD Model Showing Granular
Flight and Temperature
65
Results from Leoben Trials
Chemical and Mechanical properties
Blast furnace slag is considered unfriendly when fresh because it gives off sulphur
dioxide, and in the presence of water Hydrogen Sulphide (rotten egg smell) and
Sulphuric acid are generated. These are at least a nuisance and at worst potentially
dangerous. Fortunately the material stabilises rapidly when cooled, and the potential
for obnoxious leachate diminishes very rapidly after the ‘first flush’. However, the
generation of sulphuric acid causes considerable corrosion damage in the vicinity of
Blast Furnaces.
The dry granulation process eliminates H2S and significantly
reduces sulphur emissions, furthermore the leachability of sulphur and other
compounds is also reduced due to the glassy nature of the product.
The product quality is as follows:+95% Glass content across full size spectrum (1-6mm) with very low porosity (see
Figs. 5a and b).
Figure 5a. Microscopic (x25) view of
Figure 5b. Slag Analysis Showing Glassy
Granule <3mm Showing Low Porosity
Structure when Ground to 40-60 Micron
of Grain
Loss on Ignition < 0.1%Average particle size is 1 to 3 mm. This is dependent on cup
speed and slag properties.
(see Table 1. below showing typical sieve sizing analysis)
66
particle size
[mm]
mass
[%]
accumulated
[%]
0-1
7,02
7,02
1-2
39,64
46,65
2-3
33,36
80.02
3-4
16,11
96,13
4-6, 3
3,87
100,00
sum
100
Table 1. Typical Sieve Sizing Analysis
A comparison can be seen between granulate produced using the dry granulating
method shown in Figure 6 and the wet method shown in Figure 7.
Figure 6. Dry Granulated Slag Recently
Figure 7. A Typical Wet Granulated Slag
Produced At Leoben
In addition, since slag granulate is to be used as feedstock for the cement industry,
the following parameters are also important:
Grinding energy required to reduce granules from 3mm to reach 4000cm2/g (blaine)
= 70Kwh/t
It can be seen here, the relative block crushing strength of 100% ordinary Portland
cement when 50% of the OPC is substituted with ground material from blast furnace
slag that has been made using the dry granulation method.
67
Block Crushing Strength (N/mm2)
Curing Time
(Days)
100% OPC*
50% OPC / 50% DSG
2
34.3
15.6
7
49.5
34.1
28
59.6
57.2
91
66.1
69.2
Notes
OPC – Ordinary Portland Cement
DSG – Dry Granulated Slag
Table 2. Showing Relative Block Crushing Strength of OPC And OPC/DSG Mixture.
The temperatures achieved in the mini plant have so far indicated that the target
temperature of between 400 and 650 degrees centrigrade is certainly achievable on
the small scale (see Fig. 8).
Calculations for a full scale plant based upon a CFD model which has been calibrated using data from the mini scale plant show figures toward the maximum target figure are also achievable. The potential energy harvest based on a nominal one Tonne
of slag per minute which is typical for a blast furnace output of
around
3300TeHM/day would be in the region of 6MWel.
Figure 8. Typical Leoben Data Collection
68
Heat Recovery Developments
Several systems capable of utilising the energy in hot air delivered from the granulator have been considered. The major complication is the intermittent availability of
molten slag.
The temperature of air leaving the granulator is estimated at 400° C. By tuning the
cooling air distribution this could be increased significantly, perhaps to 650° C. The
hot air could be used for direct heating or drying or for steam raising, in which case
an accumulator would be necessary to even out the steam flow. Recovery systems
are applicable to both blast furnace slag granulators and to slag pot systems.
Depending on the plant setup the energy can be used directly for preheating or heating purposes (see Fig.9), or for the production of process steam and/or electricity
(see Fig.10).
An energy potential of more than 20 MWth or alternatively a power generation of
about 6 MWel was calculated for a slag mass flow rate of 1 t/min - which is the average slag flow for a blast furnace with an annual production of 1,7 Million tonnes and a
slag rate of 30%.
69
Figure 9. Dry slag granulation with pre-heating – for a slag mass flow rate of 1t/min a
pre heating energy potential of > 20MWtherm was calculated
Figure 10. Dry slag granulation with steam / power generation – In case of power
generation a potential of ~ 6MWel was calculated for a slag mass flow rate of 1t/min
70
Conclusion
Liquid Blast Furnace slag represents one of the largest high temperature reserves in
the steel industry that is still not utilised.
Through research and ongoing progress with our industrial partners, Siemens Metals
Technology is dedicated to reaching the first viable solution for heat recovery from
Blast Furnace slag on an industrial scale.
The prospect of waste heat recovery and a cleaner ecological footprint are clear benefits for the Ironmaker.
This doubles with the advantages to the cement manufacturer of receiving a first
class product which is dry, glassy and easy to handle making this a very good fit with
our overall customers needs.
Dry Slag Granulation with Heat Recovery stands as one of our very top Research
and Development programmes to deliver a customer focused, value added solution.
The summary advantages of the Dry Slag Granulation are: Heat recovery is possible due to the prolonged higher process temperature
 The unit can handle the full slag flow rate direct from a Blast Furnace
 High grade granulate that is suitable for use in the cement industry
 Potentially lower capital cost than an equivalent wet system
 Potentially lower operating and maintenance costs than an equivalent wet system
 Elimination of water systems
 No ground water contamination
 Handle able product
 No downstream drying costs
 No steam emissions and associated visibility, environmental and corrosion problems
 Significantly lower sulphur emissions
71
Mixing method for cooling and full vitrification of BFS
H.Kappes, Paul Wurth S.A., 32, rue d’Alsace, L‐1122 Luxembourg Abstract Paul Wurth has approached the topic of dry slag granulation with the simple but effective mixing method. The aim is to achieve fast cooling and thus full vitrification of
the blast furnace slag while creating a product with highest possible exergy content.
The basic principle of the mixing method is straight forward: liquid slag is poured into
moulds of a slag caster and steel spheres are added evenly over the surface of the
liquid slag in the mould and penetrate into the liquid slag. The steel spheres act as
cooling elements by providing a large contact surface for heat transfer, thus enabling
rapid cooling of the blast furnace slag. Within thirty seconds, heat transfer between
the slag and the steel spheres has taken place, resulting in a solidified cake consisting of vitrified blast furnace slag and enclosed steel spheres at a temperature between 600°C and 800°C. The cake breaks easily apart upon impact on a rigid surface
and the resulting loose mixture is then conveyed into a refractory lined heat exchanger where it is cooled down to ambient temperature by a counter current air flow.
Due to the counter current arrangement, high air temperatures of up to 500-700°C
may be achieved, satisfying the requirements for further use in subsequent Rankine
processes for instance. Finally, the steel spheres are extracted from the cooled mixture by means of magnetic separators and are fed back to the slag caster.
After successful accomplishment of different test series and convincing product qualities realized in these trials, Paul Wurth decided to build a full scale pilot plant. An
agreement was found with Dillinger Hütte in Saarland/Germany to build a dry slag
granulation plant to be operated at blast furnace 4 of Dillingen works and able to treat
up to 6 t/min of liquid slag. Civil works started in November 2012, followed by mechanical erection works in May 2013. The plant is currently being cold commissioned
with the aim of receiving the first hot slag in mid-October 2013.
72
Traditional Slag Handling
Blast Furnace slag is either cooled down in slag pits or it is quenched in water. Cooling in slag pits with no or little water spraying causes the slag to solidify relatively
slowly, providing sufficient time for the molecules to organize themselves in a crystalline structure. Crystalline Blast Furnace slag is a rather coarse material which can be
used for rail, road and other construction work. If the slag is quenched with water, the
slag is cooled down too fast for the molecules to organize themselves in a crystalline
structure so that they are solidified in amorphous or vitreous form, in which the product is called Granulated Blast Furnace slag (GBFS). GFBS is a valuable raw material
for the cement industry. It replaces the clinker in the cement to a large percentage
and gives the cement attractive properties. Replacement of the clinker means: The
raw material does not need to be mined and does not need to undergo calcination,
so that the use of GBFS in the cement industry is a huge contribution to save energy
and to reduce CO2 emissions. This all being valid even for water granulated slag, the
benefits can be topped by adopting a dry slag granulation technology with energy
recovery, in order to recover the sensible heat from the slag, which is leaving the
blast furnace at approximately 1500 °C.
73
Dry Slag Granulation with Energy Recovery
Such a dry slag granulation process accumulates many advantages in comparison to
the process of water granulation:










No water consumption
0.7 m3 water / t of slag saved compared to wet granulation
Heat recovery possible
1800 MJ contained in one t of slag
30-40 MW thermal energy for average size blast furnace
Reduced sulphurous emissions
Reduced transportation costs due to dry product
No drying costs in cement plants
No freezing in of slag in winter
Higher bulk density
However there are basically three preconditions to be fulfilled, which are fundamental
to the success of such technology:
1. Guarantee the vitreous solidification form by quenching the slag, because the
product value of the GBFS surpasses the value of the recovered energy.
2. Cope with the cyclic operation of the blast furnace
3. Provide the recovered energy at high temperature and at high and constant
flux rate.
The physical properties of liquid BF slag with



Low thermal conductivity
High thermal capacity
High and strongly temperature dependant viscosity
are not helping the cause.
74
The answer for the fulfilment of these preconditions and overcoming of the challenging properties of the slag is the mixing method for cooling and full vitrification of BFS.
Mixing method for cooling and full vitrification of BFS
Paul Wurth has approached the subject of dry slag granulation with the simple but
effective mixing method. The aim is to achieve fast cooling and thus full vitrification of
the blast furnace slag while creating a product with highest possible exergy content.
The basic principle of the mixing method is straight forward: liquid slag is poured into
moulds of a slag caster and steel spheres are added evenly over the surface of the
liquid slag in the mould and penetrate into the liquid slag. The steel spheres act as
cooling elements by providing a large contact surface for heat transfer, thus enabling
rapid cooling of the blast furnace slag. In the following we will look into the fulfilment
of the preconditions mentioned above.
Precondition 1), full vitrification:
75
High cooling speed is the key to vitreous solidification of the slag. To compensate for
the low thermal conductivity and the high thermal capacity of the slag, at given temperature levels, physics only offers a large heat transfer area as compensation. This
same principle applies to water quenching where the high pressure water jets disintegrate the slag stream for creating a large heat transfer contact surface
The steel balls, dropped into a mould filled with liquid slag, offer a solid and robust
solution for providing the required heat transfer area. The density difference between
liquid slag and steel balls is sufficiently high so that even high and varying slag viscosity does not pose any problem. The steel balls penetrate evenly and homogeneously at all temperature levels above 1330 °C.
Precondition 2), Coping with the cyclic operation of the blast furnace.
The answer is to take the slag as it leaves the furnace, regardless of flow and temperature. Buffering is shifted behind solidification, when there is a coarse, solid product, which can be stored and handled without problems. Again the mixing method is
robust enough to deal with varying slag flows and temperature conditions. Any process technology, which is dependent on flow rate limitations and/or temperature limitations, will enforce buffering of the liquid slag. Buffering of liquid slag will cost energy
or loss of vitrified product, both strongly influencing the economy.
Precondition 3), Provide the recovered energy at high temperature and at high and
constant flux rates.
The mixing temperature between BF slag and cold steel balls lies at approximately
650 °C, which allows energy recovery in form of hot air at a temperature of 600 °C,
sufficient for operating a steam turbine, or direct use as preheated combustion air.
The ingot consisting of steel balls and slag is dropped onto an impact plate where the
ingot disintegrates into slag particles and steel balls with a large surface area to allow
76
Heat flux rates of 30 to 40 MW transferred from slag and steel balls into the hot air as
energy carrier.
Additionally the method has the advantage, that the processes of solidification and
energy recovery are separated by the buffer so to let them happen at their specific
optimum conditions.
Project History
The first ideas were collected in 2009 followed by a first test series where the first set
of operating parameters were established and the confidence was gained that industrial flow rates will be achievable.
During the second test series in 2011 the optimum size of the cooling bodies was
established, the casting capacity and slag quality was confirmed and the disintegration pattern of the ingot was tested in a mould with the actual dimensions of future
plants.
77
The process has also been successfully tested for BOF slag, EAF slag and FeNi slag
Product Quality.
The product quality for blast furnace slag was tested in cooperation with FEhS – Institut für Baustoff - Forschung, with the following results:
The glass content meets the requirements for the cement industry as in general a
glass content of 95% and more was measured.
78
The grindability is comparable to wet granulated slags and most importantly, the
compression strength and reactivity of the cements produced with dry granulated
slag do not differ from cements produced with water granulated slags.
79
Pilot Plant
Following the favourable outcome of the second test series, Paul Wurth found an
agreement with ROGESA for implementation of a full scale dry slag solidification pilot
plant to be built on their blast furnace BF4. The pilot plant will be built in two phases.
The first phase consists of 2 areas, namely the slag caster and the material handling
area. The slag caster is connected to the slag runner of BF 4 and is located next to
the slag pits. It has full capacity and can receive slag at a rate of 6t/min from the furnace. The material handling area is an offline facility, which is built to operate with
reduced capacity.
80
The slag caster operates at full BF capacity, will however receive only one BF tap per
day. The plant consists of slag runner, caster, steel ball holding bin, dosing system,
and the impact plate. The facility shall prove all capacity and quality parameters on
full production level.
The GFBS produced will be tested by partners in the cement industry under industrial
conditions.
81
The material handling area in phase 1 is an offline solution with reduced capacity.
The hot material from the cast is transferred into the cooling bins where it is cooled
overnight. The next morning the material is sieved, crushed if necessary, and steel
balls and slag are separated by magnet.
82
Current erection and commissioning status
All major erection works on the pilot plant had been successfully finished by beginning of October 2013. The necessary civil works as well as the steel works were
achieved in a total of 8 months. The following figure shows the surroundings of the
blast furnace and the dry slag granulation pilot plant:
Blast furnace 4 of Dillinger Hütte steelworks may be depicted in the background of
the image. A slag runner connects the blast furnace cast house to the slag caster.
The steel sphere bin and the slag moulds can be clearly seen in the picture.
83
The previous figure shows the installation from the other side. The connection from
the slag runner to the pilot plant can be seen.
Liquid slag is poured into cast iron moulds (length 2m) before the steel spheres are
added evenly over the whole cross surface of the slag filled moulds.
84
The following figure shows the fully constructed material treatment area which is designed for handling one tap per day of granulated material. Hot mixture consisting of
vitrified blast furnace slag and steel spheres is discharged to five cooling bins in
which cooling takes place over night. The following morning cold material is extracted
from the bins and conveyed to a combined sieving and crushing unit before being
further conveyed to a magnetic separator. The steel spheres may then be used in a
following trial.
All major mechanical and electrical erection works are successfully accomplished.
Some minor works are still on-going. The slag caster is currently being cold commissioned. All electric motors are running successfully, the cooling circuits and all related
components are working and sequence testing is on-going.
The material treatment area has been successfully commissioned with cold material.
Hot commissioning of the complete pilot plant with liquid slag from the blast furnace
is currently foreseen for mid of October 2013.
Phase 2
Phase 2 of the pilot plant will be tackled once the first phase is running successfully.
85
In phase 2 the heat recovery and steam generation part will be added in order to
make the whole system capable of continuous operation.
From the existing slag caster the hot ingot will drop on the impact plate, will disintegrate and will be transferred by hot conveyor into the buffer.
The buffer is designed to compensate for the normal cyclic operation of the BF. Additionally the buffer serves to provide for sufficient energy to the steam generator to run
at reduced capacity for one more missed out BF tap.
Below the buffer the counter current heat exchanger is installed in which the energy
from the mixture is transferred to hot air at a temperature of approximately 600 °C.
The mixture is withdrawn from the heat exchange at a temperature of approximately
50 °C at constant mass flow to provide at the same time a constant energy flow to
the steam generator.
The steam is superheated at 15 bar and 320 °C to be used in the existing steam turbines operating at Rogesa.
86
The cold material is extracted from the heat exchanger. Most of the steel balls will
have been liberated already from the slag so that those particles can be sieved off.
The balance of material will run through an impact crusher where final liberation will
occur. Thereafter a magnet will separate steel balls from slag.
The steel balls will then be recirculated to the steel ball holding bin, closing the circuit
and allowing continuous operation of the plant.
Conclusion
 The mixing method as process for dry slag granulation with energy recovery
produces cement quality vitrified slag.
 It is not affected by varying slag flows, varying temperatures and varying
chemistries of the slag.
 It provides steady energy output from the heat recovery due to a buffer, which
is compensating the cyclic BF production by storing the hot but solid mixture of
slag and steel balls.
 The process has been successfully tested for BF slag, BOF slag, EAF slag
and FeNi slag
 The mixing method is the most robust process in the market.
87
D. Poirier1*, M. Gotelip Barbosa1, W.Xuan1, J. Poirier2, G. Thevenin2, D. Bulteel3,4
Controlled cooling of BOF slag to enhance Fe-recovery
1. ArcelorMittal Maizières, Research and Development, BP 30320, 57283 Maizières‐lès‐Metz Cedex, France 2. CEMHTI CNRS UPR3079, Site Haute Température, 1D avenue de la Recherche Scientifique, 45071 Orléans Cedex 2, France 3. Univ Lille Nord de France, F‐59000 Lille, France 4. Mines Douai, LGCgE MPE‐GCE, F‐59508 Douai, France * Corresponding author: +33 (0) 387 70 43 09 / delphine.poirier@arcelormittal.com Abstract BOF slag recycling in the steelmaking process is limited by its phosphorus content
(1-3 wt-% P2O5). Fortunately phosphorus exhibits a remarkable segregation in the
slag micro-structure and is present only in the 3CaO.P2O5-2CaO.SiO2 solid solution.
The iron containing phases are left free of phosphorus. The objective of the research
is to develop a methodology to separate iron and phosphorus by mineral processing
technologies, especially magnetic separation.
Magnetic separation tests done at 0.3 T on ground standard industrial slag has
shown a bad efficiency: in the magnetic fraction, only 66% of the initial Fe mass was
recovered, and 46% of the initial P was eliminated. However, testing on synthetic
slag mineral phases demonstrated the non magnetic behaviour of dicalcium silicate
and the anti-ferromagnetic behaviour of dicalcium ferrite. Thus, magnetic separation
is possible, but the slag micro-structure must be improved. The chosen route is to
optimize slag structure by slow cooling. At lab scale the size of the dicalcium silicate
crystals is increased by a factor 2.5 by programming a cooling speed of 1 K/min with
88
an intermediate plateau of crystallisation at 1573 K during 5 h, reinforced by an addition of 5 wt-% SiO2 to the slag. Magnetic separation tests done on the lab scale slowly cooled slag, at 0.3 T, allowed recovery of 74 wt-% of initial Fe mass and elimination of 61 wt-% of initial P content. Even though the performance needs still to be improved, the positive impact of slag slow cooling is demonstrated as both the iron recovery and P separation rates increased.
1. Introduction
Basic oxygen furnace (BOF) slag is one of the main by-products of an integrated
steel plant with rates amounting to around 110 kg per ton of steel. The slag contains
some valuable components like CaO (up to 50%), FeOx (up to 25%), and MgO (610%) in addition to non valuable ones: SiO2 (10-15%) and 1 to 3% of P2O5.
Internal recycling of BOF slag to the sinter plant or blast furnace to substitute fluxes
and iron ore allows savings in raw materials and energy consumption and relieves
the pressure on natural resources and on landfill.
On the other hand, the production of high quality steels requires extremely low phosphorus contents as it will affect the mechanical properties of the steel. Thus BOF slag
internal recycling is limited by its phosphorus content that will revert back to the hot
metal. Removal of phosphorus would, therefore, allow the use of more steelmaking
slag without any reverse effects on steel quality.
BOF slag is mainly composed of four major mineral phases: dicalcium silicate, calcium alumino-ferrite, free lime and wustite with Fe substituted by Mg and Mn. Due to a
remarkable segregation of phosphorus in the dicalcium silicate phase, the iron containing phases are left free of phosphorus [1]. Thus mineral processing technologies
such as grinding plus magnetic separation should allow selective recovery of the iron
bearing phases.
Magnetic separation is investigated in this paper. Optimisation of the separation selectivity is reached by studying of the slag mineral structure. A similar approach has
already been studied in the past [2, 3, 4] but with slags showing differences in terms
of chemical composition and mineral structure.
89
2. Materials and method
Standard slag
The BOF slag received for this investigation was prepared from a selection of 12
heats in a French steel plant. After scrap removal and slag crushing with a jaw
crusher, the slag was ground with a roller mill to reach the following specifications
96 wt-% < 630 µm and 75 wt-% < 160 µm. The chemical composition of the product
is shown in Table 5.
Fetot
SiO2
CaO
Al2O3
TiO2
MgO
P2O5
Cr2O3
MnO
LOI
wt.-% 19.80 10.48 43.54
2.98
0.49
7.16
1.46
0.17
2.91
5.63
Table 5: Chemical composition of the steelmaking slag sample (wt.-%)
Synthetic phases
One batch of 1 kg of dicalcium ferrite was synthesised by heating a mixture of lime
and ferric oxide in a magnesia crucible. Dicalcium ferrite forms an eutectic system
with magnesia at 1703 K, it has a solidus temperature of 1501 K and a liquidus temperature of 1721 K. Considering these temperatures it was chosen to work at a temperature just above the dicalcium ferrite solidus, to avoid any breakthrough or impregnation of the magnesia crucible. The mixture was heated for 3 hours at 1573 K.
Thus dicalcium ferrite is produced rather by sintering than by melting.
Two batches of 1 kg of synthetic dicalcium silicates with two different P2O5 contents
(3 wt.-% and 8 wt.-%) were synthesised by heating a mixture of lime, calcium silicate
(wollastonite, CaO.SiO2) and phosphoric anhydride (P2O5) in a magnesia crucible.
Oxide mixture was calculated to meet the following stoechiometry Ca(2-(x/2))PxSi(1-x)O4.
The choice of the heating route is made by considering the following temperatures:
-
Pure dicalcium silicate has a solidus temperature of 1737 K and a liquidus
temperature of 2403 K,
-
Calcium silicate has a solidus temperature of 1709 K and a liquidus temperature of 1817 K,
-
Calcium silicate forms an eutectic system with magnesia at 1727 K.
90
The mixture was heated for 2 hours at 1673 K and then at 1873 K for 1 hour.
The synthetic phases were ground to minus 1 mm with a ring mill. The chemical
compositions of the different synthetic phases are shown in Table 6.
wt.-%
Fetot
Dicalcium ferrite
CaO
Al2O3
MgO
P2O5
MnO
41.15 < 0.02 40.81
0.06
0.12
n.a
0.19
0.25
33.14 61.28
0.47
0.55
3.03
< 0.03
0.31
29.16 61.64
0.46
0.44
8.30
< 0.03
Dicalcium silicate
+ 3 wt.-% P2O5
Dicalcium silicate
+ 8 wt.-% P2O5
SiO2
n.a: not analysed
Table 6: Chemical composition of the synthetic phases (wt.-%)
Slow-cooled slag
Slow-cooled slag samples were obtained by heating 8 g of as-received slag enriched
with 5 % SiO2 in a Pt-Rh crucible at a temperature of 1873 K for 5 hours, cooling at a
rate of 1 K/min to reach 1573 K for 5 hours and finally cooling to room temperature at
a rate of 1 K/min in air atmosphere. The slow-cooled slag was ground to minus 100
µm with a ring mill. Its chemical composition is shown in Table 7.
Fetot
SiO2
CaO
Al2O3
TiO2
MgO
P2O5
MnO
wt.-% 21.65 15.48 41.26
2.38
0.44
6.30
1.40
2.67
Table 7: Chemical composition of the slow-cooled slag sample (wt.-%)
Analysis of microstructures
The microstructures of the slags were studied with a scanning electron microscope
(SEM) and the major phases were examined by energy dispersive X-ray spectroscopy (EDS).
91
3. Results
Standard slag
The microstructure of the standard slag is shown in Figure 7. The three main mineral
phases identified by EDS are:
-
wustite phase containing variable amounts of magnesium and manganese oxides,
-
Dicalcium ferrite containing variable amounts of alumina,
-
Dicalcium silicate containing variable amounts of phosphorus.
The stoichiometry of the different mineral phases is not regular, for example, the stoichiometry of dicalcium silicate can be closer to the one of tricalcium silicate. Other
minor mineral phases are found such as free lime and can be identified by XRD analysis.
The crystal sizes are variable: ranging from a few µm to over 100 µm. As shown in
Figure 7, grinding did not allow full liberation of the different mineral phases: thus
many grains are multi-phased with a mix of wustite, dicalcium ferrite and dicalcium
silicate.

Light grey: wustite containing a few wt-% MgO and MnO

Grey: Dicalcium ferrite

Dark grey: Dicalcium silicate
Figure 7: SEM image of ground standard slag
Successive magnetic separation test were performed by a Davis Tube tester at four
different magnetic fields (0.1 T, 0.2 T, 0.3 T and 0.4 T) and each test lasted 5
92
minutes. The testing procedure is presented in Figure 8 and the results are presented
in Table 8.
Magnetic I
Ground slag
Davis Tube
0.1 T
Non mag. VI
Magnetic II
Davis Tube
0.2 T
Non mag. I
Davis Tube
0.4 T
Non mag. III
Non mag. II
Davis Tube
0.3 T
Magnetic III
Magnetic VI
Figure 8: Flowchart of Davis Tube testing procedure
Product
Mass
(wt.-%)
Magnetic I
0
Magnetic II
2.00
Magnetic III
4.60
Magnetic VI
7.65
Non mag. VI
85.75
Table 8: Davis tube testing results on standard slag
Due to the very small amount of magnetic material recovered we may conclude that
the separation is unsuccessful. Davis Tube testing is done on 20 g of material. Mass
recovery was not sufficient to allow chemical analysis. However XRD analysis
showed differences in-between peak intensities of the non magnetic and magnetic
fractions. The magnetic fractions are enriched in iron containing phases.
Further testing was done with a Wet High Intensity Magnetic Separator (WHIMS).
Four tests were done at different magnetic fields: the results are seen in Figure 9. By
increasing the magnetic field, the amount of iron recovered increases but, at the
same time, the P2O5 elimination decreases. Magnetic separation seems possible, but
selectivity must be improved by modifying the slag mineral structure to allow the liberation of the mineral phases.
93
100
Magnetic fraction (wt.-%)
90
80
70
60
50
40
Fe recovery
P2O5 elimination
Mass
30
20
10
0
0.2
0.3
0.4
0.5
Magnetic field intesity (T)
Figure 9: WHIMS tests done on ground standard slag
Synthetic phases
Due to the poor results obtained with standard slag, it was decided to first analyse
the magnetic behaviour of dicalcium silicate and dicalcium ferrite. The magnetic behaviour of wustite is not to question.
The microstructures of two of the three synthetic mineral phases are shown in Figure 10. The preponderance of the expected mineral phases is confirmed by EDS
analysis (Figure 10) but also by XRD. Due to the fact that the synthesis could not be
done above the liquidus point of the different oxides, the presence of small amounts
of partially reacted minerals (mono-calcium ferrite, mono-calcium silicate) is seen on
the SEM images and confirmed by XRD.
94
CaO Fe2O3
44,3 55,7
CaO Fe2O3
29,9 70,1
MgO SiO2 P2O5 CaO
1,5 33,7 4,3 60,5
MgOAl2O3 SiO2 P2O5 CaO
0,8 3,5 46,2 0,7 48,8
Dicalcium silicate and 3 wt.-% P2O5
Dicalcium ferrite
Figure 10: SEM image of synthetic phases and EDS analysis
The magnetic behaviour of the synthetic phases was studied. A first series of tests
was done with a Davis Tube tester; the magnetic intensity was not sufficient to recover any dicalcium ferrite, and all the more no dicalcium silicate. Further testing was
done by using a High Intensity Magnetic Separator (HIMS) on a dry basis; the results
Magnetic fraction, recovery (wt.-%)
are given in Figure 11.
100
90
80
Ca2Fe2O5
70
Ca2SiO4 + 3% P2O5
60
Ca2SiO4 + 8% P2O5
50
40
30
20
10
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
Magnetic Field Intensity (T)
Figure 11: Magnetic recovery of synthetic dicalcium silicates and dicalcium
ferrite by HIMS
Theses results allow us to conclude that:
-
Dicalcium silicates do not respond to magnetic solicitation even at high intensities of over 1.2 T.
-
Dicalcium ferrite is sensitive to magnetisation even at low intensities (0.1 T)
but it is necessary to reach a magnetic field of 0.6 T to recover more than
95
90 wt.-% of the product. Such characteristics are distinctive of an antiferromagnetic material as described in literature [5].
Slow-cooled samples
The previous results have proven:
- The feasibility to separate the slag phases containing iron from the ones containing
phosphorus by magnetic separation,
- That the liberation of the mineral phases on as received slag is insufficient to reach
a satisfactory result.
The as-received ground slag has a mass median diameter of 75 µm. The mineral
structure presented in Figure 7 shows that all particles over 100 µm are multi-phased
with the presence of mineral phases smaller than 50 µm. Thus further grinding to less
than 50 µm would be needed to allow full liberation of mineral phases but magnetic
separation would become difficult due to the small particle size. A possible solution is
to increase the size of the mineral phases by decreasing the slag cooling speed. The
microstructure of slow-cooled slag at a speed of 1 K/min and enriched with 5 wt.-%
SiO2 is compared to standard slag in Figure 12.
b
a
a) Standard slag

Light grey: wustite containing a
b) Slow-cooled slag

Black: MgO with a few wt.-%
FeO
few wt-% MgO and MnO

Grey: Dicalcium ferrite

Grey: Dicalcium silicate

Black: Dicalcium silicate

White: Dicalcium ferrite
Figure 12: SEM image of ground standard slag and slow-cooled slag
96
SiO2 is added to the slag to promote the formation of dicalcium silicate. The grain
size of the slow-cooled slag is much larger and more uniform. EDS analysis (not presented here) shows that the chemical analysis of each phase is regular.
The wustite phase, in slow-cooled slag, is depleted in FeO and forms periclase. Table 9 presents the iron oxidation state and free lime content of standard and slowcooled slag. Only iron (III) is found in the slow-cooled slag: thus oxidation occurred
during the heating and cooling cycle. Iron (III) probably combines with free lime to
form dicalcium ferrite. This hypothesis is consistent with both the Fe2+ and free lime
content of the standard compared to the slow-cooled slag. XRD analysis confirms the
previous results with the disappearance of the peaks corresponding to wustite and
CaO and the appearance of peaks corresponding to Fe2O3 and periclase.
In the slow cooled slag dicalcium silicate and magnesia particles are trapped in a
dicalcium ferrite matrix. Indeed dicalcium ferrite is the last phase to be solidified.
wt.-%
Standard slag
Slow-cooled slag
Fe0
Fe2+
Fetot
Free lime
1.9
8.4
19.8
8
< 0.2
< 0.2
21.6
0.6
Table 9: Iron oxidation state and free lime content of standard and slow-cooled
slag (wt.-%)
We can conclude from the previous observations that controlling slag cooling conditions allows great changes in terms of slag mineral structure: both the quality and the
size of the mineral phases are improved. As the size of the dicalcium silicates is in
the range of 100 µm, slag grinding, within the specifications needed for magnetic
separation, should allow a good liberation of the iron- and phosphorus-containing
phases. The structure of ground slow-cooled slag after grinding is presented in Figure 13. The finest grains are made of a single phase of either dicalcium silicate or
dicalcium ferrite while the coarser grains, over 100 µm, are mainly composed of
dicalcium ferrite. Dicalcium ferrite seems to be harder to grind than dicalcium silicate.
The liberation of the mineral phases is largely improved compared to standard slag.
97
200 µm

White: Dicalcium ferrite

Brown: Dicalcium silicate
Figure 13: Microscope view (x10) of the ground slow-cooled slag
Magnetic separation tests were done with a Davis tube tester under the same conditions than standard slag. The results are presented in Table 10 and Figure 14.
Product
(wt.-%)
Distribution over the frac-
Composition (wt.-%)
Mass
Fe
tions (wt.-%)
P2O5 MgO CaO SiO2
Magnetic I
3.1
31.3
0.6
10.9 27.1
Magnetic II
4.7
26.2
1.0
Magnetic III
50.3
25.1
Magnetic VI
8.7
Non mag. VI
33.2
Fe
P2O5 MgO CaO SiO2
8.3
4.9
1.3
6.0
2.1
1.9
8.4
34.7 10.5
6.1
3.2
7.0
4.0
3.6
1.0
8.3
35.7 10.6 62.7 34.9 73.1 43.9 38.6
15.4
2.0
4.6
47.5 16.6
11.4
7.0
10.1 10.4
11.9
2.2
1.2
49.1 19.0 19.6 49.3
6.9
39.9 45.5
6.6
Table 10: Davis tube testing results on slow-cooled slag
Low magnetic fields of 0.1 T and 0.2 T do not allow a sufficient mass recovery. While
increasing the intensity up to 0.4 T the phosphorus content in the magnetic fraction
increases while its iron content decreases. Such observations are consistent with the
recovery of multi-phased grains: Finally, the optimum magnetic field intensity is found
to be 0.3 T with the highest mass recovery, a total iron recovery of 73.7 wt.-% and
phosphorus elimination of 60.7 wt.-% as seen in Figure 14.
The comparison of the XRD diffractogram of the non magnetic and the magnetic fractions shows large differences which confirm the success of the separation.
98
No testing was done with the WHIMS due to an insufficient amount of slow-cooled
slag available.
100
Magnetic fraction (wt.-%)
90
80
70
Fe recovery
P2O5 elimination
Mass
60
50
40
30
20
10
0
0.1
0.2
0.3
0.4
Magnetic field intesity (T)
Figure 14: Successive Davis tube tests on slow-cooled slag: cumulative curves
These results confirm the interest of slow-cooling to improve the quality of the slag
mineral structure as both the iron recovery and the phosphorus elimination are improved.
4. Conclusion
Efficiency of magnetic separation for BOF slag enrichment has been investigated.
The objective is to achieve sufficient iron recovery while eliminating phosphorus.
Analysis of iron and phosphorus speciation in the slag mineral phases confirmed the
segregation of phosphorus in dicalcium silicate. The feasibility of magnetic separation
was confirmed by studying separately the magnetic behaviour of dicalcium ferrite and
dicalcium silicate which were found to be respectively anti-ferromagnetic and non
magnetic. However a first limitation was met with standard ground slag due to insufficient mineral liberation. To overcome that difficulty, slag mineral structure was improved by SiO2 enrichment and slow-cooling at a speed of 1 K/min. Magnetic separation testing on the ground slow-cooled slag allowed the recovery of 73.7 wt.-% of the
initial iron content and the elimination of 60.7 wt.-% of the initial phosphorus content.
99
Such results demonstrate the efficiency of monitoring the slag mineral structure to
reach the initial objective. Investigations are ongoing to achieve greater separation by
comparing different magnetic separation systems and monitoring the testing parameters.
Finally, a necessary step is to work on the grinding step. Indeed mineral liberation
was improved by slag slow-cooling but the presence of multi-phased grains leads to
a limited elimination of initial phosphorus content as soon as a higher iron recovery
rate is targeted. Finding the right grinding technology and parameters will allow further improvements.
5. Acknowledgements
This research has been supported by the French Ministry of Economy and Finances
and administered by the General Directorate for Competitiveness, Industry and Services (DGCIS).
6. References
[1] Bodénan F., Gautier M., Rafai N., Poirier J., Piantone P., Franceschini G., Moulin
I., Chaurand P., Rose J. Phosphorus speciation in dicalcium silicate polymorphs of
basic oxygen furnace (BOF) slag – Preliminary results. WASCON Proceedings,
(2009)
[2] Fujita, Iwasaki, Phosphorus removal by High-gradient magnetic separation from
steelmaking slags slow-cooled in air atmosphere, Process mineralogy (1988), VIII, p
293-308
[3] Fujita, Iwasaki, Phosphorus removal from steelmaking slags slow-cooled in a nonoxidizing atmosphere by magnetic separation / flotation, Iron and steelmaker (1989),
16, p 47-55
[4] Fregeau-Wu, Iwasaki, Fujita, Removal of phosphorus from steelmaking slags – a
Literature Survey, Mineral processing and extractive metallurgy review (1993), 12, p
19-36
[5] M. Eibschütz, V. Ganiel, S. Shtrikman; Mössbauer and Magnetic Studies of Dicalcium Ferrite (Ca2Fe2O5). J. Mat. Sci., 4 (1969), p. 574
100
Theme 3
Research and Applications
101
Ayşe Ece Yıldızçelik, Aslan Ünal, Onuralp Yücel
Industrial Utilization of EAF slag as Aggregate
Department of Metallurgy and Materials Engineering- Istanbul Technical University,
Colakoglu Metalurji A. S., Turkey
Abstract
Unlike the common steel production method adopted in the world, two-third of production in Turkey is made by Electric Arc Furnace (EAF) . Related to the increase in
the production of iron-steel, slags and utilization of slag have become more remarkable subject. EAF slags are used as aggregates in asphalt and concrete, fertilizer in
agricultural applications and land filling materials. Use of natural aggregates as raw
material for asphalt industry causes serious environmental degradations. The purpose of this study is to probe the use of EAF slags as asphalt aggregate, finding the
optimum rate of hot mix asphalt mixtures and to examine the properties of asphalt
based on these mixtures.
Introduction
Nowadays, steel production increases day by day in all over the world. In 2011,
World steel production is 1.49Gt [1]. According to manufacturing methods, 35 % of
steel production in around all world is BOF process, whereas in Turkey this rate is 75
% [2]. As a consequence of steel production, it is estimated that 5 Mt of slag occurs
every year in Turkey [3].
The base material used in highway construction is aggregate. Aggregate is defined
by the British Standard Glossary of Highway as ‘’ Aggregate is the mineral component which make up main structure of mixtures such as asphalt, asphalt-macadam
and concrete’’ [4]. Asphalt, which is used in road, airfields and upper layers of other
areas, is prepared by mixing mineral aggregates, bitumen and bituminous binders
102
[5]. Asphalt structure is composed of three layers: wearing layer ,binder layer and
bituminous base layer [5]. Physical properties of EAF slag and natural aggregate is
very similar. Physical properties of EAF slag and natural aggregate are given comparatively in Table 1 [5].
PROPERTIES
EAF SLAG
NATURAL
AGGRE-
GATE
Los Angeles Abrasion coeffi- 13
15-29
cient
Wearing Resistance
Frost
8
Re- 1.0
8-11
0.0-1.7
sistance(MgSO4,%weight)
Granule Gradation (%weight)
0.5
0.5
Water Absorption(%weight)
>1
<1
Bulk Density(mg/m3)
3.4
2.8
Volume Stability
2.9
-
Table 1: Physical properties of EAF slag and natural aggregate
Experimental Studies
The experimental studies divided into two separate groups: chemical analysis and
physical tests. The experiments were performed according to ASTM standards.
Chemical Analysis
Chemical analysis were carried out to determine chemical composition of EAF slags
and the rate of free CaO. The amount of free CaO is one of the most significant component which affect asphalt quality severely.
103
Physical Tests
Experiments that are performed to determine having desired properties on use of
EAF slag as asphalt aggregate are carried out three main groups: slag experiments,
design study and wearing surface experiments.
slag experiments
In this group of experiments, EAF slags were compared with natural aggregate. In
accordance with this purpose these experiments were performed: Los Angeles abrasion test, determination of resistance to freezing and thawing, measurement of specific gravity and water absorption test, filler density, flakiness index, peeling strength
and methylene blue test.
design study
In this step of experimental study, different slag and limestone mixtures were prepared for three different asphalt layers. After adding bitumen to these mixtures, cylindrical specimens called Marshall briquettes were formed. In design of bituminous
base and binder layer, 20-37mm, 12- 20mm, 5mm - 12mm and 0-5 size of limestone
and EAF slag and also B 50/70 penetration bitumen were used. The briquettes were
prepared at 135 oC with 2x75 dash..
In design of wearing layer, 12mm - 20mm, 5mm - 12mm and 0 - 5mm size of limestone and EAF slag and also B 50/70 penetration bitumen were used. The briquettes
were prepared at 135 oC with 2x75 dash. Percentages of slag in the mixture are:
45wt% for bituminous base layer, 35 wt% for binder layer and 15 wt% for wearing
layer.
experiments of wearing surface
Water sensitivity of asphalt mixtures test are performed to measure resistance of mixture to damage after they have been in contact with water. The dry and wet specific
weights, average diameters, average heights and average indirect tensile strength of
Marshall briquettes (6 briquettes for wearing layer, 6 briquettes for binder layer) were
measured. This test was performed to wearing and binder layers
104
Wheel track testing was performed according to standard of TS EN 12697-22. Hamburg Test Device is used for measurement the sensitivity to permanent deformation
of bituminous mixtures under force. This test was performed to wearing and binder
layers. Test parameters for wearing layer and binder layer are shown in Table 2 and
Table 3, respectively.
Definition of specimen
Wearing type 1
Specimen density before the test
2.496gr/cm3
Test temperature
60 °C
Average thickness of test specimen
65mm
Environmental conditions
Temperature: 19.8°C, Moisture % 63.7
Table 2: Parameters of wheel track testing for wearing type-1
Definition of specimen
Binder Course
Specimen density before the test
2.583gr/cm3
Test temperature
60 °C
Average thickness of test specimen
65.1mm
Environmental conditions
Temperature: 21.5°C, Moisture: % 53.2
Table 3: Parameters of wheel track testing for binder layer
105
Results and Discussions
Results of the chemical analysis are shown in Table 2 and Table 3.
Fe
Ca
Si
Al
Mg
C
Ca- free
0-5mm
28.32
18.18
7.84
4.00
3.47
1.27
0.26
5-9mm
29.20
17.62
8.03
4.01
3.00
0.26
0.24
9-12mm
27.90
17.68
8.11
4.13
2.90
0.26
0.25
12-25mm
31.01
16.97
7.57
4.10
2.96
0.39
0.22
Table 2: Chemical compositions of elements which belong to EAF(%)
Fe2O3
CaO
SiO2
MgO
Al2O3
0-5 mm
39.26
25.45
16.82
5.79
7.53
5-9 mm
39.95
24.68
17.20
5.01
7.57
9-12 mm
38.62
24.75
17.38
4.84
7.79
12-25 mm 43.44
23.77
16.23
4.94
7.73
Table 3: Chemical compositions of oxides which belong to EAF (%)
Results of Los Angeles abrasion test, resistance to freezing and thawing test, specific
gravity and water absorption test, filler density, flakiness index and methylene blue
absorption test were complied with technical specification for highways published by
Republic of Turkey General Directorate of Highways (GDH). The peeling strength of
EAF slag was determined 15-20%, but this value is slightly different according to
technical specification for highways of GDH.
According to Marshall Method the amount of optimum bitumen was calculated for all
layers of asphalt structure and results are shown in Table 4, Table 5 and Table 6.
106
3.97 ± 0.5
Optimum Bitumen %
Practical specific gravity
gr/cm3
Stability,
Void,
2.609
kg
1510
%
5.20
Filled voids with asphalt(bitumen), %
60.0
V.M.A., (voids between aggregates) %
13.70
Yield,
mm.
3.50
Max. Theoretical specific
gravity (DT)
2.757
Table 4: Results of Marshall experiment at bituminous base layer
4.37 ± 0.3
Optimum Bitumen %
Practical specific gravity
Stability,
Void,
,
gr/cm3
kg
%
2.583
1640
4.80
Filled voids with asphalt(bitumen),
%
65.0
V.M.A., (voids between aggregates)
%
14.30
Yield,
mm.
3.60
Max. Theoretical specific gravity
(DT)
2.717
Rate of Filler/Bitumen
1.06
Table 5: Results of Marshall experiment at binder layer
107
Optimum Bitumen %
Practical specific gravity
Stability,
Void,
4.87 ± 0.3
,
gr/cm3
kg
1670
%
4.25
Filled voids with asphalt(bitumen),
V.M.A., (voids between
%
Yield,
2.469
%
71.0
aggregates) 14.90
mm.
4.20
Max. Theoretical specific gravity (DT)
2.609
Rate of Filler/Bitumen
1.16
Table 6: Results of Marshall experiment at wearing layer
The results of the water sensitivity test are shown in Table 7 and Table 8. Also Marshall briquettes were tested by compressive testing machine and some cracks were
seen partly. There was no crack at aggregates.
Definition of specimen and mixture
Wearing Type 1
type
Method
TS EN 12697-12 A method
Average density of dry specimen
2.421 gr/cm3
Average density of wet specimen
2.410 gr/cm3
Average indirect tensile strength of
1060.2 kPa
dry specimen
(ITSd )
Average indirect tensile strength of
832.8 kPa
wet specimen (ITSw )
Indirect tensile strength rate (ITSR)
% 78.6
Table 7: Results of water sensitivity experiment for wearing type 1
108
Definition of specimen and mixture
Binder Layer
type
Method
TS EN 12697-12 A method
Average density of dry specimen
2.548 gr/cm3
Average density of wet specimen
2.540 gr/cm3
Average indirect tensile strength of
1012.9 kPa
dry specimen
(ITSd )
Average indirect tensile strength of
663.6 kPa
wet specimen (ITSw )
Indirect tensile strength rate (ITSR)
% 65.5
Table 8: Results of water sensitivity experiment for binder layer
The results of Wheel track test was found 5.95% and 5.50%reductionin thickness for
the wearing surface and binder layer respectively.
Conclusion
Highway industry has significant importance all over the world, especially in Turkey,
and EAF slag can be used as aggregate for road construction properly. Disposal of
waste by utilizing in industry has became a remarkable issue day by day. EAF slag
which is the waste of great importance for Turkey and all over the world should be
used many industrial area.
References
[1] http://www.worldsteel.org/statistics/statistics-archive/2011-steel-production.html,
access date 09/01/2013
[2]
http://www.recyclingdergisi.com/HaberlerDetay.aspx?ID=34,
20/03/2013
access
date
[3]
http://www.karyapsan.com.tr/asfalt-ansiklopedisi.aspx, access date 24/05/2012
[4]
M. Ilıcalı , Asphalt and Applications, ISFALT, Turkey, 2001
[5] T. Sofilic, A.Mladonavic, U. Sofilic, Characterization of EAF steel slag as aggregate
for
use
in
road
construction,
http://www.aidic.it/CISAP4/webpapers/17Sofilic.pdf,access date 20/02/2012
109
6
H. Epstein , R. I. Iacobescu
B. Blanpain7
7,8
, Y. Pontikes7,8, A. Malfliet7, L. Machiels7,8, P.T. Jones7,
Stabilization of CaO-SiO2-MgO (CSM) Slags by Recycled Alumina
RVA, Les Islettes 55120, France
Abstract
Stainless steel slags generated in melting and refining operations are CaO-SiO2-MgO
rich with Cr2O3, Al2O3 and F- in minor quantities. A major issue is collapse of the slag
structure on cooling as the high dicalcium silicate (C2S) content undergoes a phase
transformation from β-C2S (monoclinic) to γ-C2S (orthorhombic). This phenomenon is
accompanied by a volume increase of around 12%. Consequently, slag handling and
storage are problematic. Furthermore, structural collapse prevents the realization of
commercial value for the slag. β to γ conversion of only 4% slag by weight is sufficient to cause the dusting phenomenon. CSM slags are often treated with borates to
prevent β to γ transformation of C2S. However, is borates are expensive and health
concerns may limit their use in the future.
Valoxy®, an alumina-rich material derived from the recycling of aluminium salt slags,
offers an alternative route to CSM slag stabilization in which the formation of C2S is
prevented altogether. In lab trials the stabilized slag demonstrated improved microstructure, less porosity and superior micro-hardness compared to slags stabilized by
borates. Furthermore, the high level of spinel in slag stabilized by Valoxy suggested
superior entrapment of Cr-bearing compounds and by implication reduced Crleaching. The combination of a low cost stabilization route, improved slag properties
and environmental benefits should make Valoxy attractive to stainless steel producers and re-processors of CSM slags.
6
RVA, Les Islettes 55120, France
7
Centre for High Temperature Processes and Sustainable Materials Management, Department of Metallurgy and
Materials Engineering, KU Leuven, Kasteelpark Arenberg 44 bus 2450, 3001 Heverlee, Belgium
8
Secondary Resources for Building Materials, Consortium in Sustainable Inorganic Materials Management, SIM2,
KU Leuven, Belgium
110
1.
Introduction
1.1 Argon Oxygen Decarburization
Argon Oxygen Decarburization (AOD) is a process used in the production of stainless
steel and other high grade alloys containing oxidizable elements such as chromium
and aluminum. AOD is part of a duplex process in which scrap or virgin raw materials
are first melted in an electric arc furnace (EAF) or induction furnace. The molten metal is then decarburized and refined, in an AOD vessel, to less than 0.05% carbon.
1.2 AOD Slag
The AOD process produces a slag containing CaO, SiO2 and MgO as major oxides;
and Cr2O3, Al2O3 and F- in minor quantities. The CaO/SiO2 ratio, generally 1.5-2.0,
varies between different steel plants and between different AOD heats in the same
plant. The major issue arising during AOD slag cooling is the high content of dicalcium silicate (C2S). C2S undergoes several phase transformations (Figure 15) [i], the
most important of which is the inversion from β-C2S (monoclinic) to γ-C2S (orthorhombic). This transformation is accompanied by a volume increase of around 12%
[ii]. β to γ transformation causes disintegration of the slag, the so called “dusting effect,” making slag handling and storage problematic and the achievement of economic value for the slag virtually impossible. Formation of only 4 wt.% of γ- C2S is enough
to cause slag disintegration.
111
Figure 15. Polymorphic transformations of C2S
1.3 AOD Slag Stabilization – State of the Art
Three methods have proved effective in preventing formation of γ-C2S in AOD slag.
(a) Stabilization of β-C2S through addition of stabilizing elements such as boron
(Figure 16) [iii].
(b) Avoid formation of C2S by migrating from the stable region of the CaO-SiO2-Al2O3
phase diagram (Figure 17).
(c) Combination of (a) and (b)
112
Figure 16. The influence of foreign ions in stabilization of β-C2S
Volumetric stability is related to the presence of C2S in the slag. In order to achieve a
commercial value for their AOD slag, steel companies are compelled to incorporate
an additive that prevents β to γ transformation of C2S on cooling. The additive commonly used is sodium borate, acting as a dopant, which is both expensive and has
associated health risks.
An alternative stabilization mechanism to boron is the addition of alumina which
causes the formation of the stable C2AS (dicalcium aluminosilicate) complex. Whereas boron chemically stabilizes the β form of C2S to prevent transformation into γ-C2S,
the addition of alumina changes the slag chemistry considerably and the formation of
C2S is avoided. The presence of alumina reduces slag basicity and moves the system out of the C2S stable region of the CaO-SiO2--Al2O3 phase diagram.
113
C2S
Figure 17. CaO-SiO2-Al2O3 phase diagram: pathways to prevention of C2S formation
2.
Valoxy as Stabilizer of AOD Slag
The concept of using pure alumina and alumina/borate mixtures to stabilize C2S in
AOD slag has been verified at the University of Leuven at lab scale [iv]. A 4-8% addition of alumina was found to reduce the borate requirement by 50%. Actual alumina
requirement depended on the calcium oxide/silica ratio in the slag.
Given the known potential for stabilization of AOD slags by pure alumina, RVA commissioned a study at the University of Leuven to evaluate the potential for using Valoxy as the alumina source [v]. In particular, the research set out to investigate
whether:
(a) the alumina content of Valoxy could be applied in the stabilization role
(b) the residual oxides in Valoxy could contribute positively to stabilizing the C2S
system and improving overall slag properties.
114
The work was performed on Valoxy and AOD slag dried to constant weight. The results of the Leuven study on Valoxy may be summarized as follows (all percentage
additions are by weight):
1. Stabilization of AOD slag was possible with 15% Valoxy additions (Figure 18).
Figure 18. AOD slag sample treated with 15% Valoxy by weight
2. Valoxy addition of 15% to AOD slag delivered comparable microstructure and
micro-hardness results (Vickers) with 15% alumina addition. Whereas Al2O3
addition to the slag promoted the formation of gehlenite, a mineral with hardness
between 5 and 6, Valoxy addition promoted the formation of Ca,Mg-silicates
(hardness 6) and higher spinel content (hardness 8).
3. AOD slags with 10% and 5% Valoxy additions were successfully stabilized by
Dehybor, a commercial boron product containing 53% B2O3 by weight, with additions of 0.5%, 0.3% and 0.2%.
4. The stable slag products achieved with 5% Valoxy and 0.5%, 0.3% and 0.2%
Dehybor delivered comparable Vickers micro-hardness results with 15 % Valoxy
AOD slag. Hardness results for all combinations added to the slag are shown below (Figure 19).
115
Figure 19. Vickers micro-hardness tests results of all produced stable samples with alumina, Valoxy and
Boron additions, in wt.% (Val = Valoxy; B= B2O3).
5. In samples with Valoxy, spinel formation was favoured over gehlenite due to an
extra source of MgO available in the system, provided by Valoxy. Elemental
maps indicated that Cr was associated with Fe, Al, Mg, Mn and some Ti, probably in a spinel phase. According to other studies, Cr entrapment in the spinel
phase minimizes leaching. This suggested an additional benefit of applying Valoxy to AOD slags, namely improved Cr entrapment.
6. AOD slag with Dehybor at 0.5%, 0.3% and 0.2% additions alone delivered slag
products with a more porous structure and slightly lower Vickers micro-hardness
compared to AOD with the same Dehybor additions and Valoxy at 5%.
7. With additions of 5% Valoxy and small amounts of Dehybor, the slag was more
compact compared with samples of AOD with Dehybor only.
8. A summary of Valoxy/boron combinations and their effect on slag stability is
shown below (Figure 6).
116
Figure 6. Volumetric stability of AOD slag as a function of equivalent B2O3 and Valoxy additions
The Leuven study thus demonstrated that the addition of Valoxy can stabilize AOD
slag without addition of borates. Moreover, there are strong indications that the final
slag is better in terms of:
 microstructure - lower porosity
 more spinel crystals incorporating Cr - thus minimizing leaching.
The weight of Valoxy required, >10 %, implies a slag processing step involving the
supply of extra heat, most cost-effectively by the addition of trace quantities (around
0.6%) of 25/75 Fe-Si alloy.
3.
Environmental Benefits
AOD slag remediation needs to take account of the potential for chromium leaching.
In view of its effect on both humans and the environment, Cr is subject to rigorous
legislative controls. For example, an EU directive regarding Cr(VI) came into effect in
2003 prohibiting the use or supply of cements containing more than 2 ppm watersoluble chromium by mass of cement [vi].
A number of adverse health effects, have been associated with Cr(VI) exposure. According to NIOSH [vii], all Cr(VI) compounds are considered potential occupational
117
carcinogens. Nevertheless, Cr is rarely oxidized and in a number of cases is entrapped in slag in the spinel phase. This implies that the slag has the appropriate
chemistry, i.e. sufficient Al is present in the composition. This is not generally the
case for AOD slags and is thus an additional driver for Valoxy addition.
Mudersbach et al. reported additions of bauxite, Al2O3-containing residues and aluminium metal as methods to increase the stability of stainless steel EAF slags and
stabilize Cr [viii]. The aim of the additions is to decrease the basicity of the slags and
favour the formation of spinel-type phases during solidification. In such a case, even
if the Cr content of the slag is high, leaching can be suppressed. The Mudersbach
study seemed to confirm that spinel behaves, in practice, as a stable phase with respect to chromium leaching. More specifically, the authors proposed that additions of
unspecified Al2O3-containing residues could mitigate chromium leaching from EAF
slags. The alumina content of Valoxy could therefore play a favourable role in the
prevention of chromium leaching from AOD slags.
4.
Conclusions
4.1 General
Volumetric stabilization of AOD slag cooled at room temperature is possible in three
ways:
(a) addition of a boron-bearing compound
(b) addition of pure Al2O3
(c) addition of Valoxy
Volumetric stability of AOD slag is related to the presence of C2S. Addition of boron,
acting as a dopant, prevents the thermodynamically expected β to γ transformation.
Indeed, formation of 4 wt.% of γ-C2S is enough to cause slag disintegration. The addition of Al-bearing compounds such as Valoxy, changes the slag chemistry considerably and C2S formation is prevented.
118
4.2
Valoxy Addition
In the Leuven study:

Stabilization of AOD slag was possible by 15 wt.% Valoxy addition.

Valoxy addition of 15wt.% to AOD slag delivered comparable microstructure and
Vickers micro- hardness results with 15wt.% alumina addition to AOD slag.

AOD slags with 10 wt.% and 5 wt.% Valoxy additions were successfully stabilized
by Dehybor (53 wt.% B2O3) additions at 0.5, 0.3 and 0.2 wt.%.

The stable products, with Valoxy addition at 5 wt.% and 0.5, 0.3 and 0.2 wt.%
Dehybor, delivered comparable Vickers micro-hardness results with 15wt.% Valoxy AOD.

In samples with Valoxy, spinel formation was favoured instead of gehlenite due to
an extra source of MgO available in the system (provided by Valoxy).

Elemental maps indicated that Cr was associated with Fe, Al, Mg and Mn, some
Ti also, most probably in a spinel phase. This suggested an additional benefit of
using Valoxy, namely minimized chromium leaching.
4.3

Valoxy and Dehybor Additions
AOD with Dehybor in 0.5, 0.3 and 0.2 wt.% delivered products with a more porous structure and a slight lower Vickers micro-hardness results compared to
AOD with Valoxy in 5 wt.% and same Dehybor additions.

C2S was successfully stabilized as β- polymorphic form by Dehybor additions in
0.5, 0.3 and 0.2wt.%.

Valoxy addition of 5 wt.% and of small amounts of Dehybor to AOD slag created
a slag with a more compact morphology compared with samples of AOD with
Dehybor only.
The Leuven study demonstrated that the addition of Valoxy can stabilize AOD slag
without addition of borates. Moreover, there are strong indications that Valoxy-treated
AOD slag was better in terms of micro-structure with less porosity and more spinel
119
crystals incorporating Cr, thus minimizing leaching). The amount of Valoxy required
(>10 wt%) implies a slag processing step wherein extra heat is supplied.
References
[1]
Kim, Y. Y., I. Nettleship, et al. (1992). Phase Transformations in Dicalcium Silicate: II, TEM Studies of Crystallography, Microstructure, and Mechanism.
Journal of American Ceramic Society. 75: 2407-2419.
[1]
Bridge, T. E. (1966). "Bredigite, larnite and γ dicalcium silicates from marble
canyon." The American Mineralogist 51: 1766-1774.
[1]
Taylor, H. F. W., Ed. (1990). Cement Chemistry. London, Academic Press.
[1]
“Additions of Industrial Residues for Hot Stage Engineering of Stainless Steel
Slags.” Pontikes et al., Proceedings of the 2nd International Slag Valorization
Symposium, April 2011, Leuven, p314
[1]
Iacobescu, R.I., Pontikes, P., Malfliet, A., Machiels, L., Epstein, H., Jones,
P.T., and Blanpain, B., “A Secondary Alumina Source for the Stabilization of
CaO-SiO-MgO Slags.” Proceedings of the 3rd International Slag Valorization
Symposium, KU Leuven, Belgium (March 2013): 311-314.
[1]
European Parliament Directive (2003). "2003/53/EC of the European Parliament and of the Council."
[1]
U.S. Department of Health and Human Services, Centers for Disease Control
and Prevention, et al. (2008)."Criteria document update, Occupational Exposure to Hexavalent Chromium." External Review Draft.
[1]
Mudersbach, D., M. Kühn, et al. (2009). Chrome immobilization in EAF-slags
from high-alloy steelmaking: tests at FEhS institute and development of an
operational slag treatment process. First International Slag Valorization Symposium, Leuven, Belgium.
The technical contribution of Inspyro Ltd. to this work is acknowledged.
Address for correspondence: hepstein@013net.net
120
I. Unamuno1) and A. Morillon2)
Recent and former European RFCS slag research projects
Slag treatment and utilisation
1) GERDAU I+D Europa, Basauri, Spain, 2) AMEtech, Lyon, France Abstract Over more than twenty years, the European Commission (EC) has supported research on iron and steelmaking slags. Since 2002, the Research Fund for Coal and
Steel (RFCS) Programme has been set up and is used by EC to support further research. Through RFCS projects industry, research organisations, public and private
sectors and/or universities within the EU Member States come together to work on
common problems. Cooperation between European partners is necessary not only to
share knowledge and best practice, but to help the European authorities set up reasonable common regulations. This can be achieved through research work that convinces the authorities and customers that slags are valuable products, with sometimes better properties than natural stones, used normally for such applications. Not
only has slag been tested as a metallurgical tool, but also the mechanical-technical
properties and especially the environmental compatibility of the solidified slag are
being intensively investigated. To demonstrate that no harmful impacts occur to the
environment by using slag products is one of the most urgent problems to work on in
the future and to modify them when any requirement is not fully met.
This report summarises all the past and current RFCS projects that investigate slag.
The research projects deal with state-of-the-art of slag applications in Europe, optimisation of slag practice during primary and secondary carbon or stainless steelmaking,
treatment of slag after steelmaking, utilisation of fine-grained or coarse-grained steel
slag and use of slag in agriculture.
121
Introduction
Since 2002, the Research Fund for Coal and Steel (RFCS) Programme has been set
up and is used by the European Commission (EC) to support research on iron and
steelmaking slags. The RFCS is a continuation of funding provided for research on
steel and coal by the European Steel and Coal Community (ECSC). Through RFCS
projects industry, research organisations, public and private sectors and/or universities within the EU Member States come together to work on common problems. The
projects supported by RFCS cover every aspect of coal and steel production: production processes, application, utilisation and conversion of resources, safety at work,
environmental protection and reduction of CO2 emissions.
During the production of iron and steel, considerable amounts of slags are produced.
The use of iron and steel slags as high quality products influences the cost of ironand steelmaking. The steel industry puts great emphasis on treating slag as a byproduct and limiting the amount that is deposited. Internal and external recycling of
slag also results in ecological and economical advantages for iron and steel production. Some of the reasons that slag is not reused include: small quantity of a given
type of slag produced, environmental limitations, mechanical-technical properties
limitations and nonuniform regulations.
Not only is slag being tested as metallurgical tool throughout EU, but also the mechanical-technical properties and especially the environmental compatibility of the
solidified slag are being intensely investigated. To demonstrate that no harmful impacts occur to the environment by using slag products is one of the most urgent
problems to work on in the future and to modify them when any requirement is not
fully met. Only use of slag that is environmentally safe can take place and give confidence to the public.
Cooperation between European partners is necessary not only to share knowledge
and best practice, but to help the European authorities set up reasonable common
regulations. Today, common regulations for slags do not exist or are not uniformly
applied in the EU. Safety and usability of any slag is determined based on local regulations, even within the same country different regulations are applied by local gov122
ernments to evaluate the slag. This means that the environmental safety and utilisation of any given slag does not only depend on the properties of the slag, but on the
regulation limits that apply in the area where the slag is produced or used. Common
regulations can be achieved through research work that convinces the authorities
and customers that slags are valuable products, with the same or sometimes better
properties than natural stones, used normally for such applications. As slag is produced in different plants and countries it is important to show that solutions found in
one country can work for others.
This paper gives an overview of the type of slag research done or is currently ongoing that is sponsored by RFCS or formally by ECSC. This is not a limitation into the
type of research that can be sponsored with the help of RFCS, but a way to see what
has been done to help clarify new research areas. The research projects in this paper
have been divided into seven areas: overview of slag applications in Europe, optimisation of slag practice during primary and secondary carbon or stainless steelmaking,
treatment of slag after steelmaking, utilisation of fine-grained or coarse-grained steel
slag. Some of the projects deal with different areas. Since 1994 there have been
more than 60 research projects (both sponsored by ECSC or RFCS) that deal completely or partly with slag, out of these 20 deal directly with improving slag quality to
create a value added product.
Overview
One research project supported by ECSC, gave an overview of the properties and
utilisation of slags from 1980 to 1998, establishing the state of the art of slag applications in Europe at that time. It summarises the success of development of different
fields of applications for slag utilisation, after 20 years of research, which led to increase in slag use throughout Europe. The conclusion was that new areas of slag
utilisation need to be identified to further reduce the amount of slag deposited, while
due to new environmental legislations slag optimisation has to be constantly researched and improved [1]. Since the research project was conducted new elements
are in focus and new legislation requirements have been passed. This requires techniques for internal slag recycling or outside use, which have not been previously investigated.
123
Optimisation of slag practice during primary steelmaking
To optimise the slag practice during primary steelmaking (BOF and EAF) several different areas were investigated. Optimisation of slag foaming in EAF was addressed
in four projects with additional differences: treatment of high-chrome steel [2], recycling of stainless steel dusts [3], control of EAF burners and injectors for oxygen and
carbon [4] and production of ultra high strength steel grades [5]. Foamy slag influences not only the productivity of EAF, but also after tapping different properties with
respect to technical and environmental parameters of the solidified slag. Slag splashing in BOF was investigated and challenged the assumption that improved vessel
shape improves the process consistency [6]. An optical camera monitoring system
was developed to observe the scrap-melting process, which allowed for excellent
quality images of foaming slag behaviour [7]. Mathematical models were improved for
smelting reduction processes which includes a dynamic slag droplet model of converter [8]. Improving of BOF blowing by controlling the foaming slag has been investigated and measurements of process-critical components in steelmaking slags
[9/10/11]. Detection systems for slag composition in BOF, EAF, converter and ladle
have also been developed [12/13/14/15/16/17]. As part of a project to design new
generic steel grades physical properties of slag were collected [18]. Image sensor
analysis of deslagging was developed to minimise the metallic loss [19].
Metallurgy in the processes influences the quality of iron and steel slags after solidification relating to the later utilisation. Seven projects deal with optimisation of slag
[20/21/22/24/25/26/27/28/29], but only two with the main goal of producing value
added slag product [28/29]. Alkali, chlorine and sulphur were the main concern in
three projects. Optimisation of alkali removal and desulphurisation of the final slag
have been demonstrated by laboratory and industrial investigations [20]. Recycling of
slag [24/25/26] and change in slag composition for better steel production [27] were
also investigated. Liquid BF slag was treated with oxidising agents (ore or mill scale)
to de-gas the slag in the slag pot which produced a less porous slag, resulting in slag
with better properties [28].
SLACON project (ongoing) deals directly with producing a good quality end product
that can be used by the construction industry. To decrease/eliminate critical elements
124
(F-, Ba, Cr, Mo, Se, V) two different techniques will be investigated: immobilisation of
these leachable substances during hot slag liquid stage and/or recycling of washing/cooling water from slag treatment with elimination of leachable components. Due
to different regional regulations slag from different European steelworks have different problems, not just because of the chemical composition of the slag, but also because of different requirements that are put on slag that is utilised [29].
Optimisation of slag practice during secondary steelmaking
The main aim of research projects during secondary steelmaking is the interaction
between refractory material, slag and steel [30/31/32] and cleanness of the steel,
which has been investigated by analysis or change in the slag [33/34/35/36/37/38].
The quality of secondary metallurgical slag (ladle furnace slag) depends on the metallurgy in the vessel, but no investigations concerning the later use of this slag designed for optimised steel cleanliness have been done in these European research
projects.
Other projects dealt with fast vacuum slag analysis [39], bubble bursting [40], slag
foaming [41] and ladle stirring processes [42].
Optimisation of slag from stainless steelmaking
Two research projects were done on scorification of chrome during high alloy
steelmaking to improve the environmental behaviour of the EAF-slag from stainless
steelmaking. The prevention of chrome scorification was achieved by optimisation of
the furnace atmosphere, slag forming and reduction agents during melting. This created chrome spinel phases, which have low solubility. Additionally the liquid EAF-slag
has been treated successfully with different agents during tapping, i.e. with bauxite,
Al2O3-containing residues or mill scale [43/44].
EPOSS project dealt with energy efficiency by conditioning techniques for slag foaming [45]. As part of a resource-saving operation and control of stainless steel refining
project, extensive slag analyses regarding chemical composition and phase structure
were evaluated [46].
125
Treatment of slag after steelmaking
The objective of one pilot and demonstration project was to develop a flexible process outside the main metallurgical line allowing total transformation of both steel
slag and in plant by-products into value added products. The most suitable reactor
according to previous experience was a DC furnace with a hollow electrode for simultaneous treatment of slags and fine-grained materials. By charging through the electrode all materials are going into the DC-plasma and are efficiently treated [47].
Utilisation of fine-grained steel slags
During solidification of dicalcium silicate containing steel slags and by hydration of
CaOfree- or MgOfree-containing steel slags the material can disintegrate. Special applications for fine-grained steel slags have to be developed; as sealing material
[48/49/50/51], cement [52] or as fertiliser [53/54/55].
Three projects deal with using slag as fertiliser, two are finished [53/54] and one is
ongoing [55]. The long term effects of using slag as fertiliser are very important to
show that slag can be as good as other industrial or natural fertilisers and to give
confidence to the public. Not only the effect of slag fertilisers on soil, but also the effect on the plant health and accumulation of metals are investigated [55]. The PSPBOF project (started 2013) aims to separate P-rich slag fraction from P-poor slag
fraction produced in the BOF. The P-rich slag can be used as a fertiliser and P-poor
and Fe-rich slag can be recycled internally in sinter plant or directly in BF [56].
The ULTRAFINE project dealt with the environmental issues associated with fine
fraction pollution due to processing of fine material (slag included). When dealing with
fine fractions of slag, emissions of fine particles have to be kept in mind [57].
Utilisation of coarse-grained steel slags
Some projects dealing with the use of coarse slag in different applications have been
investigated (but are limited). Solutions for technical and environmental problems
were investigated for using slag as aggregate to improve the subsoil. In this research
project no negative impact on the environment was found due to slag [58]. The slag
characteristics have been investigated to develop a technical guide for using slag in
road construction with respect to the quality of groundwater [59]. A special treatment
126
process was developed to solve the problem of volume instability in slag due to free
lime. After tapping the BOF slag into slag pot, the BOF slag is treated by injection of
quartz sand and oxygen. Free lime interacts with the SiO2 and iron oxides forming
stable calcium silicate and calcium ferrite phases [60]. Promotion and valorisation of
BOF slag was investigated to determine the potential use in civil engineering and agriculture [61]. Slag was also used as filter material to remove P from waste water
treatment plants in pilot scale tests. Slag showed good P sorption capacities and
possibility to be reused as P-rich fertiliser [62].
SLACON project (mentioned in “optimisation of slag practice during primary
steelmaking” section) deals also with the use of slag after treatment in the construction industry [29].
Conclusion
About 30 % of slag projects sponsored by European Commission deal directly with
slag as value added product. And most of the projects were sponsored during the
time of ECSC (16 out of the 20 projects dealing with slag utilisation were sponsored
before creation of RFCS). Only 4 projects sponsored by RFCS deal directly with the
use of slag as a value added product. This is modest, because the use of slag is not
limited to a given region or country, but is of a European interest. Helping to solve
current problems and working on more uniform European regulations is an important
aspect of future work in Europe. As regulations change sharing knowledge throughout Europe regarding how to solve problems that might arise in the future will reduce
the costs of slag production. In addition, providing the regulatory agencies with sound
research as to the benefits of utilisation of slag can create benefits for all iron and
steel producers.
127
Acknowledgments
These projects have been supported by European Commission (EC) through different programs like ECSC or RFCS.
References
[1]
7210-ZZ/585: Summary report on RTD in iron and steel slags, 1998
[2]
7210-CB/406: Development of foamy slag method in the electric arc furnace
melting and treatment of high-chrome steel, 2000
[3]
7215-PP/026: Foaming of the slag and recycling of stainless steel dusts by
injection into the electric arc furnace for stainless steels, 2002
[4]
RFSR-CT-2003-00031 EAFDYNCON: Dynamic control of EAF burners and injectors for oxygen and carbon for improved and reproducible furnace operation
and slag foaming, 2007
[5]
RFSR-CT-2006-00005 EAF-PROMS: Cost efficient metallurgy for the production of novel ultra high strength deep drawable steel grades with high Mn contents from 10 to 25 wt.-% by using EAF steel making route, 2008
[6]
7210-PR/133: Consistent BOS performance, 2002
[7]
RFSR-CT-2004-00008 EAFCAMERA: Control by camera of the EAF operations in airtight conditions, 2007
[8]
7210-AA/419: Development of a combined postcombustion model (CPM) for
smelting reduction processes, 1997
[9]
7210-CB/206: Improvement of the EAF performances through optimization of
foaming slag practice, 1998
[10]
7210-CB/903: Improvement of the EAF performances through an optimization
of the foaming slag practice - slag level measurement using radio wave techniques, 1998
[11]
7210-CB/817/904/905: Radio wave interferometer technique for BOF slag control, 2000
128
[12]
7210-PR/271: In situ, quick sensing system for measurements of processcritical components in steelmaking slags (INQUISSS), 2004
[13]
RFSR-CT-2003-00042 BOFDYN: Dynamic end point in BOF through a fast
and simultaneous determination of the steel/slag composition, 2006
[14]
7210-PR/230: New developments for the quantification of non conductive materials (slags, inclusions) in steel industry by optical emission spectrometries
(laser, spark), 2003
[15]
7210-PR/231: Fast analysis of production control samples without preparation,
2003
[16]
7210-PR/302: On-line slag analysis utilizing contact free microwave technology,
2004
[17]
7210-PR/296: Development of techniques for the production of glassy standard materials for the analysis of slags by spectroscopic methods, 2004
[18]
RFSR-CT-2004-00027 ISA-PESR: Integrative simulation of advanced protective gas electro-slag-remelting for the production of high-quality steels, 2007
[19]
RFSR-CT-2010-00005 OPTDESLAG: Increased yield and enhanced steel
quality by improved deslagging and slag conditioning, 2013
[20]
7210-PR/068: Investigations of chlorine and alkali behaviour in the blast furnace and optimization of blast furnace slag with respect to alkali retention capacity, 2001
[21]
7210-PR/071: Injection of slag correction components into the blast furnace,
2001
[22]
7210-PR/074: Optimized blast furnace slag water quenching with sulphur
compounds control, 2001
[23]
RFSR-CT-2008-00004 BATHFOAM: Control of slag and refining conditions in
the BOF, 2008
[24]
7210-PR/203: Efficient utilisation of waste products from secondary steelmaking as flux materials for electric arc furnace, 2003
[25]
RFSR-CT-2007-00010 URIOM: Upgrading and utilisation of residual iron oxide
materials for hot metal production, 2007
129
[26]
RFSR-CT-2012-00039 REFFIPLANT: Efficient use of resources in steel plants
through process integration, 2015
[27]
RFSR-CT-2005-00004 HINIST: Mastering of P-ESR technology for high nitrogen steel grades for high value applications, 2008
[28]
7210-AA/126: Investigations to decrease the porosity of blast furnace slags,
1994
[29]
RFSR-CT-2012-00006 SLACON: Control of slag quality for utilisation in the
construction industry, 2015
[30]
7210-CC/302: Development of techniques to minimise ladle/slag interactions
and prevent uncontrolled inclusion modification, 1999
[31]
RFSR-CT-2007-00011 STEELCLEANCONTROL: Development of steel grade
related slag systems with low reoxidation potential in ladle and optimised ladle
glaze technique for improving steel cleanliness, 2010
[32]
RFSR-CT-2009-00003 LADLIFE: Enhanced steel ladle life by improving the
resistance of lining to thermal, thermomechanical and thermochemical alteration, 2012
[33]
7210-CC/118: Development and control of suitable slag systems for improving
steel cleanliness in ladle treatment and tundish metallurgy, 1998
[34]
7210-CC/808: Improving deoxidation practices for ultra clean steel production,
1998
[35]
7210-PR/080: Desulphurisation of liquid steel with refining top slags, 2001
[36]
7210-PR/270: Improvement of inclusion flotation during RH treatment, 2004
[37]
7210-PR/275: Definition of ladle change strategy to avoid slag entrapment and
to control inclusion population, 2004
[38]
7210-PR/329: De-oxidation practice and slag ability to trap non-metallic inclusions and their influence on the castability and steel cleanliness, 2005
[39]
RFSR-CT-2003-00043 AVAS: Feasibility of a fast vacuum slag analysis by
laser OES in secondary steelmaking, 2006
[40]
7210-CC/123: Control of ejections caused by bubble bursting in secondary
steelmaking processes, 1999
130
[41]
7210-PR/079: Control of inclusion, slag foaming and temperature in vacuum
degassing, 2001
[42]
RFSR-CT-2007-00009 StImprove: Improvement of ladle stirring to minimise
slag emulsification and reoxidation during alloying and rinsing, 2010
[43]
7210-CB/124: Decreasing the scorification of chrome, 1998
[44]
7215-PP/044: Chrome immobilisation in EAF slags from high alloy steelmaking - Development of a slag treatment process, 2004
[45]
RFSR-CT-2007-00006 EPOSS: Energy and productivity optimised EAF stainless steel making by adjusted slag foaming and chemical energy supply, 2010
[46]
RFSR-CT-2007-00007 OPCONSTAINLESS: Resource-saving operation and
control of stainless steel refining in VOD and AOD process, 2010
[47]
7215-AA/903: The in-plant by-product melting (IPBM) process, 1998
[48]
7210-XA/104: Untersuchungen über Möglichkeiten zur Vermeidung von Rißbildungen bei der Verwendung von Gemischen aus kristallinen und glasigen
Hochofenschlacken als Tragschichtmaterial, 1979
[49]
7210-CB/115: Utilisation of fine-grained steel slags for mineral sealing products, 1995
[50]
7210-CB/605: Utilisation of fine-grained steel slags for mineral sealing products - Laboratory and practical investigation, 1995
[51]
7210-PP/028: Innovative use of iron- and steelmaking by-products for the
sealing and securing of steel industry deposits, 2005
[52]
7210-XA/108: Verbesserte Verwertung von Hochofen- und Stahlwerkschlacken, 1986
[53]
7210-CB/935: Production of NPK fertilizers from steel manufacturing byproducts & improved fertilization through computerized simulation techniques,
1996
[54]
7210-PR/267: Sustainable agriculture using blast furnace and steel slags as
liming agents, 2004
131
[55]
RFSR-CT-2011-00037 SLAGFERTILISER: Impact of long-term application of
blast furnace and steel slags as liming materials on soil fertility, crop yields
and plant Health, 2014
[56]
RFSR-CT-2013-00032 PSP-BOF: Impact of long-term application of blast furnace and steel slags as liming materials on soil fertility, crop yields and plant
health, 2016
[57]
RFSR-CT-2004-00049 ULTRAFINE: Characterisation of emission and impact
of ultrafine particulate, 2007
[58]
7210-XA/105: Research on steelworks slag, especially its use in road construction, 1983
[59]
7210-PR/195: Characterization, modelling & validation of the impact of iron
and steelmaking slags used in road construction on groundwater, 2003
[60]
7210-CB/112/113 Investigation of the production of volume-stable construction
material from steel slags,1995
[61]
7210-CB/203: The valorization of BOF slag in combination with urban waste,
1994
[62]
RFSP-CT-2009-00028 SLASORB: Using slag as sorbent to remove phosphorus from wastewater, 2012
132
J.S. Chen1), S.F. Chen1), M.C. Liao1), W.C. Chen 2), T.L. Tao 3), B.L. Hsu 4), T.K. Hsu
4)
Construction of Test Sections to Evaluate Performance of Basic Oxygen Furnace (BOF) Steel Slag as Aggregate in Stone Mastic Asphalt
1)
Department of Civil Engineering, National Cheng Kung University, Tainan 701,
Taiwan
2)
Department of Civil Engineering, Kao Yuan University, Kaohsiung 821, Taiwan
3)
China Steel Corporation, Kaohsiung 812, Taiwan
4)
CHC Resources Corporation, Kaohsiung 812, Taiwan
1
Abstract
Environmental and technical problems have led to increasing attention being paid to
the subjects of secondary solid materials in the construction and maintenance of road
infrastructure. The industrial by-products from the metallurgical industry have been
frequently applied to base and surface courses in the road structure. These byproducts are primarily constituted of electric arc furnace (EAF) slag and basic oxygen
furnace (BOF) slag. This study was motivated by concerns in Taiwan that stone
mastic asphalt (SMA) mixed with BOF steel slag might not perform well under the
environmental and traffic conditions. SMA is designed to be a tough, stable, rutresistance mixture that relies on stone-to-stone contact to provide strength; however,
it was feared that the rut resistance, friction, durability and environmental benefits of
SMA with BOF would be lost due to heavy trafficking. The Tainan City Government
joined the China Steel Corporation in 2011 to begin a multi-year research program to
evaluate the performance of BOF steel slag as aggregate substitute in road rehabilitations. This paper is to report the test results of the first phase.
Test sections were constructed by using three different types of asphalt mixtures as
follows: stone mastic asphalt with BOF (BOF-SMA), dense-graded asphalt concrete
with BOF (BOF-DGAC), and dense-graded asphalt concrete with natural aggregate
(NA-DGAC). These sections completed in March, 2012, were extensively evaluated
133
for their performance in terms of durability and safety. Traffic loading and densification were shown to be the main reason to cause an increase in permanent deformation of asphalt pavements. For the BOF-DGAC and NA-DGAC sections, the rut
depth resulted from traffic compaction was much higher than that for the BOF-SMA
section. The increase in the international roughness index (IRI) corresponded well
with the increase in rut depth. Because of the significant amount of macrotexture
produced within BOF-SMA pavement surfaces, BOF-SMA layers maintained adequate frictional characteristics even after become condensed. No raveling, cracking
or other failures have been observed on the BOF-SMA section to any significant extent since open to traffic, which suggests that BOF-SMA be a viable pavement surface type for use on roads to provide good performance, including good friction, reduced rutting, and improved durability. In addition, BOF could be well encapsulated
by a rich mortar binder in SMA. Asphalt cement is a highly hydrophobic substance
and is capable of forming an immobilizing barrier that can prevent BOF from any expansion. The use of BOF steel slag as aggregate is shown to present appropriate
technical solutions for road applications, which could ensure both an excellent level
of performance together with a harmless environmental impact.
2
Introduction
Basic oxygen furnace (BOF) steel slag, a by-product of the steelmaking process, is
readily available in the southern urban area in Taiwan. If unused, the slag material
could end up in landfills, increasing the expenditure of public and private agencies as
disposal facilities reach capacity and new landfills are required. One way to utilize
the steel slag is to incorporate it into flexible pavements. This process has been
used successfully by other researchers with reported improved pavement performance [1-6]. Prior to this study, Taiwan had limited experience in handling, testing,
and constructing steel slag pavements.
The primary objective of this study is to evaluate the use of BOF in hot mix asphalt
(HMA) concrete. All aspects of using BOF in asphalt pavements will be evaluated,
including:
134
 Asphalt concrete mix design and testing,
 Constructability, and
 Performance characteristics.
3
Materials and Mix Design
3.1
Aggregate
Natural aggregate used in this study was a limestone obtained from the Kao-Ping
River, and BOF was supplied by the CHC Resources Corporation. The basic properties of BOF and natural aggregates were listed in Table 1. Both materials meet the
specification requirements stipulated by the roadway agency.
Test
BOF
Natural
Aggre- Spec.
Methods
gate
LA Abrasion (%)
10.26
17.75
< 30
Flat and Elongated (%)
AASHTO T96
ASTM D4791
1：3 3.12
3.84
< 15
1：5 0.52
1.56
<5
Bulk Gravity
3.41
2.62
-
AASHTO T85
Absorption (%)
2.00
1.28
-
AASHTO T85
Soundness (%)
0.65
0.73
< 12
AASHTO
T104
Crushed Content (%)
ASTM D5821
One face 100
100
100
Two face 100
95
>90
Table 1: Basic properties of BOF and natural aggregate
The LA abrasion test provides an indication of the relative quality of competence of
various sources of aggregate. Roadway agency uses the LA abrasion test as an indication of aggregate wear resistance. BOF seems to possess better LA abrasion
value than natural aggregate. Both aggregates have few flat and elongated particles
135
although the shape of BOF aggregate is more cubic. The specific gravity of BOF is
about 30% more than that of natural aggregate.
Because of the porous surface of BOF, the absorption value of BOF is slightly higher
than that of natural aggregate. The sodium sulfate test measures the soundness of
aggregates subject to weathering action, by immersing samples in sodium sulfate.
The test results indicate that both samples are resistant to weathering; however, the
steel slag has a better soundness value as compared to natural aggregate. BOF and
natural aggregate have a high content of crushed particles that could provide the interlocking mechanism.
3.2
Bitumen
A total of three mix designs were performed in this study as follows: the stone mastic
asphalt with BOF (BOF-SMA), dense-graded asphalt concrete with BOF (BOFDGAC), and dense-graded asphalt concrete with natural aggregate (NA-DGAC).
Three types of bitumen were used as follows: Pen 85/10, Pen 60/70 and polymermodified binder (PMB) for NA-DGAC, BOF-DGAC and BOF-SMA, respectively. The
properties of asphalt binders listed in Table 2 meet the requirements set by the roadway agency.
Test
Fresh
Penetration (25C, 0.1mm)
Viscosity
(60C, poise)
(135C, cSt)
Flash point(C)
Solubility (%)
TFOT residue
Weight loss(%)
Retained penetration (%)
Ductility (cm)
Pen 85/100
Pen 60/70
PMB
95
1,137
321
>232
99.9
63
2,693
512
>232
99.5
25
10,753
1658
>232
99.7
0.1
58
100+
0.01
55
100+
-
Table 2: Binder properties
136
3.3
Mix design
Both the control mix (i.e., NA-DGAC) and the slag-modified mix (i.e., BOF-DGAC and
BOF-SMA) were designed by the Marshall procedure according to ASTM D1559.
The Marshall compactor was used to compact the DGAC samples 75 times each
side, but 50 times each side for the SMA samples. The Marshall stabilometer was
used to determine the stability after compaction. The mix design criteria and mix
characteristics at design binder contents were listed in Table 3.
The steel slag mix was prepared substituting steel slag for coarse (19 to 4.75 mm)
aggregate. The test mix included about 53% and 75% steel slag aggregate for BOFDGAC and BOF-SMA by total weight of the mix, respectively. The greatest difference between the control and test mix was with the maximum specific gravity. The
usage of a relatively high content of BOF results in an increase in the bulk gravity of
an HMA mix. The bulk gravity of the control mix was 2.36, whereas that of the steel
slag mix was 2.81 to 2.88 depending on the BOF content. Care should be paid to the
difference in unit weights when calculating bid quantities for overlay construction
when using BOF as aggregate.
137
Characteristics
NA-DGAC
BOF-DGAC
BOF-SMA
100
100
100.0
3/4"(19)
95
97.4
94.9
1/2"(12.7)
82
78.3
57.9
3/8"(9.5)
74
70.6
45.4
#4(4.75)
51
47.0
24.9
#8(2.36)
37
31.5
18.6
#16(1.18)
31
25.4
16.6
#30(0.6)
22
18.7
14.7
#50(0.3)
12
11.7
12.5
#100(0.15)
7
7.3
11.0
#200(0.075)
5
4.9
9.3
Binder content (%)
5.1
4.3
5.2
Bulk gravity
2.36
2.81
2.88
Stability (kgf)
1111
1822
931
Flow value (mm)
2.2
3.2
3.6
Voids (%)
4.3
4
4
VMA (%)
17.6
11.3
15.9
VFA (%)
70
69.7
75.6
Retained strength (%)
81.2
84.3
84
Sieve (mm)
1"(25)
Table 3: Mix design criteria and design mix characteristics at design binder contents
An increase in stability is observed for the BOF-DGAC mix as compared to the NADGAC mix. Coarse BOF mixes may provide more stability because the larger components would include angular and rough textured particles that would increase the
interlocking friction. However, the stability value cannot be used to evaluate the potential resistance to rutting for the SMA mix, since field performance is more representative of strength of SMA mixtures.
138
4
Construction of Test Sections
4.1
Pavement structure
All three test sections constructed are located in the southern region of Tainan City.
The stone mastic asphalt with BOF (BOF-SMA), dense-graded asphalt concrete with
BOF (BOF-DGAC), and dense-graded asphalt concrete with natural aggregate (NADGAC) test sections are located adjacent to each other on an urban roadway, with
no exits among them. These three mixtures were used to build test pavements about
750 m long with each section of about 250 m. All other variables (i.e., geometric and
structural design) were held constant. Each section was consisted of a 10-cm asphalt surface course over a 30-cm coarse aggregate base course as shown Figure 1.
The surface course was constructed with two 5-cm lifts thickness including tack coat
and prime coat sprayed in between to achieve smooth requirements.
Figure 1: Pavement structure of test sections
4.2
Aggregate storage, mixing, placement and compaction
Quality control of hot-mix asphalt mixture begins with the stockpiles of aggregate that
are to be processed through the asphalt plant and incorporated into the mix. To reduce the amount of moisture that accumulates in BOF, especially from rain, a roof or
a shed should be provided to cover the BOF stockpiles. The moisture content inside
139
BOF will directly affect the quality of asphalt concrete as well as field performance.
The mixing plant was a 2,000-kg batch plant that produced mix at 120 ton/hr. BOF
steel slag was added to the mix like the natural aggregate, substituted for the 4.75- to
19-mm aggregate. The mix was dropped into end-dump semi-trucks and hauled to
the paving site, about 30 minutes away. Because of different types of bitumen, the
mixing temperatures were 155, 160, and 175C for the NA-DGAC, BOF-DGAC and
BOF-SMA test section, respectively.
End-dump semi-trucks laid the mix in a hopper in front of a paving machine. For the
NA-DGAC and BOF-DGAC test sections, compaction was provided with a 12-ton
pneumatic-tired roller making six passes, and a 10-ton vibratory roller making two
passes in vibratory mode and two passes in static mode. The finish roller was an 8ton steel wheel roller which made at least two passes.
Only steel rollers were used when compacting the BOF-SMA section. Breakdown
rolling began immediately behind the paver, and the roller stayed close behind the
paver at all times. Six passes in static mode were needed to seat the BOF-SMA test
section because over-rolling could lead to aggregate breakdown. Due to the gapgraded nature of the SMA mix, there is extensive stone-to-stone contact between the
coarse aggregate particles, with very few fine materials to cushion the coarse aggregates. Pneumatic-tired rollers are not allowed for use on SMA. The rubber tires tend
to pick up the mortar causing surface deficiencies.
5
Field Performance
All test sections were completed in March 2012 with typical construction equipment
and operations. This urban road has four lanes and an average traffic volume of
36,000 vehicles per day with about 15% truck traffic. These test sections are heavily
trafficked, and the terrains are essentially straight and level. The relative humidity
varies from 50 to 90% with annual average rainfall of about 2000 mm. Following
construction, distress surveys have been carried out on a regular basis on each section during trafficking. The field performance includes friction, rutting and ride quality.
140
5.1
Friction
Pavement skid resistance was measured by the British Pendulum Tester according
to ASTM E303 and expressed by a British Pendulum number (BPN). The tests were
all adjusted by the exact pavement temperature at measurement to an equivalent
BPN value at 20C, as shown in Figure 2. The measurement of the friction showed
an initial BPN of 63 to 64. Skid resistance was relatively low just after construction
because of asphalt binder film coating the aggregate at the pavement surface.
As a consequence of the disappearance of the binder film covering the surface of the
aggregate, skid resistance was improved after test sections open to traffic. A BPN
value higher than 45 is considered sufficient and safe for roadway pavements. According to test results in Figure 2, all three test sections provide good wet weather
friction. The BPN value of test sections built with BOF appear to provide good skid
resistance after one year in service. Because of the significant amount of macrotexture produced within BOF-SMA pavement surfaces, BOF-SMA layers maintained
adequate frictional characteristics even after become condensed.
141
Figure 2: Pavement friction
5.2
Rutting
A rut is a surface depression in the wheel path. The mean rut depth is calculated by
laying a straightedge across the rut, measuring its depth, then using measurements
taken along the length of the rut to compute its mean depth in millimeters. As shown
in Figure 3, rutting increases with increasing service time. Rutting of these test sections stems from the permanent deformation of the pavement surface, primarily
caused by consolidated movement of asphalt mixtures due to traffic load. The severity level is considered to be low when the mean rut depth is less than 12.5 mm,
moderate when rutting is between 12.5 and 25 mm, and high when rutting is higher
than 25 mm.
With the rut depth lower than 12.5 mm, test results indicate that all test sections possesses good resistance to plastic deformation after one-year in service. In particular,
the BOF-SMA section has a coarse gradation that results in stone-on-stone contact.
Rutting ranged from 4 mm for the BOF-SMA section to 8 mm for the NA-DGAC section after one year in service.
The rut depth of the NA-DGAC section was highest
among the three sections, that of BOF-SMA the lowest, and that of BOF-DGAC in
between. In the BOF-SMA section, the angularity and toughness of BOF was shown
to improve the rutting resistance of SMA mixtures.
Asphalt cement is a highly hydrophobic substance and is capable of forming an immobilizing barrier that can prevent BOF from any expansion. The use of BOF steel
slag as aggregate is shown to present appropriate technical solutions for road applications, which could ensure both an excellent level of performance together with a
harmless environmental impact.
142
Figure 3: Permanent deformation on pavement surface
5.3
Ride Quality
Roughness is an important index of pavement performance evaluation, which affects
the comfortableness of drivers and passengers.
It is an index involving human-
vehicle-road interaction, often evaluated by the International Roughness Index (IRI).
The ICC Surface Profiler used in this study is a multi-wheeled inclinometer-based
system that is pushed by an operator at a walking speed of 1.2 km/h. For these sections, the IRI value expressed by m/km increased with time as shown in Figure 4. It
is required that the IRI value of a roadway after construction be lower than 3.5 m/km.
All test sections meet the specification requirement.
The IRI value of BOF-SMA was the lowest, while the roughness of NA-DGAC was
larger than that of BOF-DGAC. The increase in the international roughness index
(IRI) corresponds well with the increase in rut depth. No raveling, cracking or other
failures have been observed on the BOF-SMA section to any significant extent since
open to traffic. Test results suggest that BOF-SMA be a viable pavement surface
type for use on roads to provide good performance, including good friction, reduced
rutting, and improved durability.
143
Figure 4: Ride quality of pavement surface
5.4
Performance summary
A comparison of in-place properties, skid testing and ride testing does identify some
differences in performance between the control and steel slag pavements. No raveling, cracking or other failures have been observed on the BOF-SMA section to any
significant extent since open to traffic. Test results indicate that BOF-SMA is a viable
pavement surface type for use on roads to provide good performance, including good
friction, reduced rutting, and improved durability. According to field observations,
BOF is well encapsulated by a rich mortar binder on the SMA surface.
144
6
Conclusions and Recommendations
Asphalt concrete was produced and test sections were constructed when basic oxygen furnace slag was used as a portion of the aggregate. On the basis of the test
results and the field performance conducted in this work, the following conclusions
and recommendations appeared warranted:





7.
BOF steel slag could meet the specification requirements set by the highway
agency as road construction aggregate. Using conventional paving techniques,
test sections were successfully constructed with steel slag replacing coarse aggregate in hot-mix asphalt mixtures.
The BOF-SMA section exhibited superior performance compared with the control
section built with natural aggregate. Field data indicated that the use of BOF as
aggregate is a viable option for hot-mix asphalt mixtures.
The gradation of the steel slag should be monitored to assure that a uniform mixture of hot-mix asphalt concrete is produced. To reduce the amount of moisture
that may accumulate in BOF, especially from rain, a roof or a shed needs to be
provided to cover the BOF stockpiles at an asphalt plant.
The specific gravity of the produced slag aggregate should be monitored. Care
should be taken to unit weights of asphalt concrete mixed with BOF when calculating bid quantities for overlay construction.
Since the period of time for data collection only lasts one year, this study is a
preliminary report. More field data should be collected to address the long-term
performance of steel slag as aggregate for asphalt pavements.
References
[1] Lewis, D.W.: Resources conservation by use of iron and steel slag, ASTM STP
774 (1982) pp.31-42.
[2] Noureldin, A.S., R.S. MacDaniel: Evaluation of surface mixtures of steel slag and
asphalt, Transportation research record 1269 (1900) pp.133-149.
[3] Bagampadde, U., H.I. Al-Abdul Wahhab, S.A. Aiban: Optimization of steel slag
aggregates for bituminous mixes in Saudi Arabia, Journal of materials in civil engineering 11 (1999) 1 pp.30-35.
[4] Motz, H., J. Geiseler: Products of steel slags an opportunity to save natural resources, Waste management 21 (2001) 3 pp.285-293.
[5] Maslehuddin, M, A.M. Sharif, M. Shameem, M. Ibrahim, M. Barry: Comparison of
properties of steel slag and crushed limestone aggregate concretes. Construction and building materials 17 (2003) 2 pp.105–12.
[6] Wu, S., Y. Xue, Q. Ye, Y. Chen: Utilization of steel slag as aggregates for stone
mastic asphalt mixtures, Building and environment 42 (2007) 7 pp.2580-2585.
145
E. Nagels, S. Arnout and B. Soete
Slag properties – Easy access using new dedicated software
InsPyro N.V., Kapeldreef 60, B-3001 Leuven, Belgium
Abstract
InsPyro has developed a toolbox for slag properties based on state-of-the-art thermodynamic calculations. Based on slag composition and temperature, it is now possible to calculate the following important slag properties:
 Phase formation as a function of temperature
 Expected minerals in the slag after cooling
 Phase fractions at high temperature
 Liquidus and solidus temperature, which determine the solidification interval
 Heat content of the slag, that could be recovered
 Corrective additions needed for basicity control and avoiding powder formation
 Viscosity of the slag as a function of temperature, which determines how
easily the slag can be handled
 Metallurgical properties, such as the sulphur capacity
The Spark-software gives the slag engineer or slag end-user a must-have toolbox
with the following unique functionalities:
 Fast calculation with a single click
 Fully integrated in Excel
 No expert interpretation needed
 Best quality data and most reliable results
 User-friendly interface and customized results overview
InsPyro believes that this toolbox will help slag handlers to develop smarter slag recuperation and engineering schemes, and help slag producers improve their metallurgical process. In this presentation, we discuss the models behind Spark, its functionality, and how it can help to increase profitability in the slag business.
146
Introduction
The idea that slags are only an annoying side effect of the process luckily belongs to
the past. Nowadays slags are no longer waste but a by-product which requires attention to achieve an optimal value. The day at which this idea is picked up by metallurgical plants, “slag engineering” is born. The possible added value is manifold: slag
engineering to improve the metallurgical processes, slag engineering to reduce energy requirements, slag processing to recover energy, giving slags a new life as construction material, slags as a fluxing agent… It is clear that understanding the behavior and properties of slag is crucial to have an optimal benefit.
What does a “slag engineer” need to know? Let’s start with looking at the metallurgical side of slag. Undoubtedly slag is needed to perform proper metallurgy, and above
all a slag engineer should not forget that the main aim of a metallurgical plant is producing metal. So if you are the metallurgist responsible for the slag, you’ll have to
work within the boundaries allowed for metallurgy.
As a consequence slag engineering is bound to the limits of the metallurgical process. Before the slag becomes a byproduct, it is mainly the phase to collect all unwanted elements…the bin of the metallurgy. At this point the slag needs to be designed for the metallurgical process in terms of basicity, desulfurization capacity, captation of inclusions… Then the slag is separated from the metal. Usually this is done
by tapping or deslagging the slag into a slag pot. Here, viscosity becomes an issue.
Viscosity at high temperature is difficult to measure and it is impossible to follow up
the viscosity in line. Once noticing a higher viscosity, it is likely that it is too late to
change something. Literature offers several models for viscosity as a function of
composition and temperature, also taking into account to effect of partial solidification. As a result, calculations before charging can help to avoid tapping problems.
The tapped slag has a specific heat content which could be used for heat recovery.
Or in case the final slag application requires a different composition, some fluxing
may be required. In this situation it is necessary to know if the energy content of the
slag is sufficient to melt the flux.
InsPyro developed a program which allows obtaining slag properties as a function of
composition and temperature. These are proven to be instrumental for both metallur-
147
gical industries as for slag handling. This paper aims to inform on the background of
the slag property calculation and to illustrate the results that can be expected.
Figure 20: Example of a calculation result (online demo version).
Figure 21: Example of a possible input and output lay-out of the commercial Spark
software.
Composition and phase formation in slags
The composition and thus the properties of a slag depend on the metallurgical origin
(Figure 22). It is well known that slags from iron blast furnaces are well suited to be
used in cement replacing the Portland clinker as CEM III type cement [1]. This application was already published over a century ago [2] and still captures the attention of
researchers around the world today. In the iron-steel making chain multiple slag
types are formed [3]. Each slag type (from steel making and from non-ferrous metallurgy) has its typical compositional range and specifications.
148
Figure 22: Overview of compositional differences of different slag types in steel making [11].
The visualization of slag compositions in a phase diagram as in Figure 22 has proven
to be instrumental in documenting the variation of the slag composition in terms of
basicity or degree of oxidation. When also the isotherms and primary phase regions
are indicated as in Figure 23, follow-up of the slags melting behavior becomes easier.
If the visualization indicates trends or problematic evolutions, these can be linked to
changes in the charge composition (raw materials) and fluxing strategies, and corrective measures can be taken. To obtain such phase diagrams, with isothermal projection, equilibrium calculations are performed based on thermodynamic data.
149
Figure 23: Visual reporting on slag variation
Phase equilibrium calculations are based on the laws of thermodynamics, which state
that a system will evolve to the situation with the lowest energy content. The result,
at constant pressure, is the single specific combination of phase fractions with their
individual composition, which gives the system the lowest Gibbs free energy of all
possible combinations at a certain temperature. This information is visualized in conventional literature phase diagrams. They are, however, only available for limited
combinations of components, and are not always mutually consistent. For the Spark
tool reliable, accurate and state-of-the-art multicomponent thermodynamic databases
are used, which do not have these limitations [4].
Liquid
Most metallurgical processes require a sufficiently liquid slag phase. Hence, the metallurgical process defines the boundary conditions of the liquid slag properties and
thus a compositional range. Visualization of the compositional range for which the
slag is fully liquid is typically done by polythermal projections of the liquidus temperature as in Figure 23. These type of ternary and even some quaternary slag diagrams
(mainly relevant to iron and steel processes) can easily be found in the Slag Atlas [5].
150
However, seen the multicomponent constitution of the slag, it can be quite complex to
find the diagram which suits your system best. The Spark software produces the
most suitable diagram and accounts for all relevant impurities. This allows to systematically keep track of the slag liquidus temperature.
The isothermal liquidus lines also indicate the solubility limits of a compound in the
slag at a certain temperature. This solubility is an indication for the possible chemical
wear of the refractory (= dissolution of the refractory in the slag). From the metallurgical point of view, it is beneficial to saturate the slag with the lining material (MgO,
Al2O3…) in order to reduce the aggressive behavior of the slag towards the refractory. The chemical corrosion of refractory is limited near the saturation point and the
lifetime of the furnace lining is extended [6]. In steelmaking, dolomitic lime (CaOMgO mix) is used as a flux to increase the MgO amount of the slag so that the MgO
concentration is close to the solubility limit and chemical degradation of the refractory
is minimized.
When looking at slags as a byproduct, the properties of the liquid are still highly relevant. For maximal value of the slag, the liquid slag may offer opportunities over the
cooled solid slag. Indeed, besides the material itself, the liquid slag still has its heat
content. This heat content could be validated in heat recovery systems or allows for
further optimization of the slag for its final destination. At this point the slag composition can be changed without interfering with the metallurgical process where it originates, without high demands for energy. This intervention at high temperature can
still change the composition, e.g. by the addition of a stabilizing flux to avoid dusting
or leaching problems [7,8,9,10]. Once cooled, it is too late to intervene in such a way.
Solidification: from liquid to solid
Models are available that can predict phases forming during the cooling of slag.
When crystalline solidification is concerned, two extreme situations can be defined.
The first is called equilibrium solidification, in which it is assumed that at every stage
the material is in equilibrium conditions. This situation can occur when cooling is very
slow and no kinetic constraints are present. Equilibrium solidification assumes that all
atoms can move freely over all the phases. The other extreme situation is called
Scheil-Gulliver solidification [11], in which it is assumed that once a solid is formed it
will not change composition. In terms of diffusion, this means that solid atoms are not
151
able to move. In reality, the phase composition will be between these situations
(Figure 24). These models do not include glass formation as they are both based on
thermodynamic equilibrium and diffusion constraints. The kinetic conditions in which
a glassy phase is formed largely depend on cooling conditions. It is possible to take
this aspect into account by virtually dividing the slag in a fraction which crystallizes
and a fraction which remains glassy. Sufficient industrial data has to be available before a reliable model can be made. In the Spark software, standard assumptions or
more advanced industrially corrected models can be included.
Figure 24: Equilibrium and scheil solidification model compared to an experimental
sample [11].
Solid phases and the second life of slag
It is very well known that the properties of a slag largely depend on the mineralogy
[7]. The short term degradation (dusting) of steel slags is assigned to the C2S phase
which upon cooling undergoes a phase transformation and expands about 12%. The
mechanical stresses caused by this expansion reduce the slag to dust. A similar
problem, although with a different and slower mechanism, is seen in hydrating slags.
Steel slags containing free CaO or MgO will hydrate over time. The expansion of these phases makes the slag unstable. This effect is not seen immediately due to the
slow kinetics of the hydration reaction (possibly months). Also non-ferrous slags, as
the sodium slags in lead battery recycling, exhibit this hydration behavior. As the
origin of this stability problem is identified, a next step would be to avoid the problematic phases. At this point thermodynamic insight in the slag system becomes valuable.
152
From the resulting phase equilibria, it can be deduced which fluxing strategy or other
strategy is suitable to convert the slag into a stable and consequently more valuable
product, for e.g. application in building materials [13,14].
Viscosity
Viscosity is a measure for the resistance of a fluid against flowing. Measurements at
high temperature are possible but very difficult to achieve. Multiple estimation models
exists which relate the viscosity to composition and temperature. Spark makes use of
the Urbain model for liquid viscosity [5].
This model is based on the knowledge that the internal structure of a liquid slag has a
large influence on the viscosity. A liquid slag is an ionic structure, a chaotic mixture of
ions, but with short range order. As different atoms have different affinities for oxygen, the liquid structure can minimize its internal energy by making an ordered structure with this charge. Acid oxides as silica can take up additional oxygen ions, resulting in a negatively charged ion (SiO44-), with a rigid tetrahedral structure. As a consequence, the silica will share oxygen ions to form a three dimensional structure. These
oxides are known as network formers. Basic oxides, such as lime, are able to release
their oxide ion resulting in the creation of uncharged ends of the structure. Therefore,
they are called network breakers. The presence of network formers will tend to increase the viscosity while network breakers decrease the viscosity, making it flow
more easily. This leads to a formula for the viscosity as a function of a complexly corrected basicity ratio.
A further adjustment is made when solid particles are appearing in the slag. The
Spark software makes use of the Roscoe equation correction which takes into account a fraction of solid and a correction for particle aggregation (equation 1)
(1)
0 is the liquid viscosity, f is the fraction of solid which is a result of the equilibrium
calculation [14]. The fraction of solids can be accurately calculated, thanks to the link
with thermodynamic phase equilibria calculations. In Figure 25, the viscosity is calculated as a function of slag basicity. It is clear that once the slag is saturated in lime
and the first solids appear, the viscosity increases rapidly even when the viscosity of
the liquid decreases further.
153
Figure 25: Spark calculation result for viscosity of a CaO - SiO2 - 8% MgO - 10%
Al2O3 slag as a function of basicity
Heat content
In the framework of heat recovery of slag, an important quantity is the available heat
content of slag. A distinction has to be made between slow cooling, which allows
crystallization of the slag, and fast cooling, resulting in a glassy slag. The difference
between the two situations is the release of the solidification enthalpy, which may
amount to more than 20% of the available heat.
The maximum heat release (i.e., for crystallized slag) corresponds to the enthalpy
difference between the original state of the slag and the crystallized state. In Figure
26, the heat release is shown for a slag with basicity (CaO/SiO2) of 1.5, as a function
of the Al2O3 and MgO concentration. The Al2O3 concentration of the slag has a large
influence on the heat content, while the influence of MgO concentration is only limited.
154
Figure 26: Heat release (= enthalpy change) of a slag as function of composition as
calculated with Spark
The actual heat recovery potential largely depends on the technique used. About
50% of the heat is reported to be recovered in trials [15]. Based on industrial/experimental data an efficiency factor can be applied to the thermodynamic heat
content (split in solidification heat and sensible heat). The heat content also defines
to which extent fluxes can be added without the need for additional heating. It is, for
example, possible to calculate the theoretical temperature drop of a slag pot when
sand or alumina rich material is added for stabilization [9].
Sulfur capacity
The sulfur capacity of a slag is mainly a metallurgical concern. To desulfurize steel,
an artificial slag is added to the ladle, which is designed to pick up sulfur from the
steel. Sulfur capacity is defined as the amount of sulfur the slag can pick up from the
steel:
CS= (PO2/PS2)1/2. (wt% S)slag
(2)
where P stands for the partial pressure in the gas phase. The equation can be rewritten in function of sulfur and oxygen activities in the steel:
155
C’S = (a[O]/a[S]). (wt%S)slag = exp (-2154/T+3.166).CS [5]
(3)
Figure 27: Sulfur capacity as a function of basicity for a CaO - SiO2 - 8% MgO - 10%
Al2O3 slag with an oxygen activity in steel of 0.0002%.
Conclusions
Modern software tools can give the slag engineer and his operators easy access to
slag properties. Liquidus temperatures, solid fraction and solidification structure can
be predicted based on reliable thermodynamic data of slag systems. Also models for
slag viscosity and sulfur capacity are available in literature. The combination in a dedicated package allows to have all required information with a single click. Also the
visualization in fully adjusted multicomponent phase diagrams is a large step forward
compared to the more conventional drawing on hard copy phase diagrams. The main
application areas of the tools are improving both the high temperature slag practice
as well as maximizing the reuse. The tools can be used for production-related research purposes, as well as for daily follow-up of the slag. When powerful tools are
used in production, this results in faster feedback, better decisions, and very importantly, an easy way to involve a larger number of people in the understanding and
follow-up of slag practice, amplifying the impact.
156
References
[1]
European standard EN 197-1:2000.
[2]
Blast furnace slag, Journal of the Franklin Institue, vol. 108, issue 6
(1879) 410-415.
[3]
Euroslag statistic publication 2010: available online
(www.euroslag.com).
[4]
C.W. Bale et al., FactSage thermochemical software and databases,
Calphad vol 26 No2 (2002) 189-228.
[5]
Slag atlas 2nd Edition, Verlag Stahleisen GmbH, 1995.
[6]
P.T. Jones, Degradation mechanisms of basic refractory materials during
the
secondary refining of stainless steel in VOD ladles, PhD dissertation, 2001.
[7]
D. Durinck et al., Hot stage processing of metallurgical slags, Resources,
conservation and recycling 52 (2008) 1121-1131.
[8]
D. Mudersbach et al, Improved slag qualities by liquid slag treatment,
proceedings 2nd international slag valorization symposium 2011, 299-
[9]
311.
R.I. Iacobescu et al, A secondary alumina source for the stabilization fo
CaO- SiO2-MgO slags, proceedings 3rd international slag valorization
symposium
[10]
2013, 311-314.
F. Engström et al., Review: Hot stage engineering to improve slag valorization
options, proceedings 2nd international slag valorization
symposium 2011,
231-251.
[11]
D. Durinck, High temperature processing of metallurgical slags; a method
to
promote recycling, PhD-dissertation, 2008.
[12]
D. Van Mechelen, Valorisation of stainless steel slags: Zero waste concept,
proceedings 2nd slag valorization symposium 2011, 145.
[13]
L. Boehme et al., Ferromolybdenum slag as valuable resource material
for
the production of concrete blocks, proceedings 2nd slag valorization
symposium 2011, 129-144.
[14]
L. Wu, Study on some phenomena of slag in steelmaking process, PhD
dissertation, 2011.
[15]
H. Motz et al., Dry solidification with heat recovery of ferrous slag,
proceedings 3rd international slag valorization symposium, 37-55.
157
Dr N Ghazireh(1), B Kent(1) and J Smith(2)
Behaviour of Slag Bound Mixtures in Road Construction
(1) Lafarge Tarmac, Millfields Rd, Ettingshall, Wolverhampton, WV4 6JP, UK.
(2)
Lafarge Tarmac, School Road, Miskin, Pontyclun, Mid Glamorgan, CF72 8PG,
UK
Abstract
Slag Bound Mixtures (SBM) are widely used throughout Europe for sub-bases and
base layers in the construction of all categories of roads and other civil engineering
applications. In the UK, the use of bound foundation layers using bitumen or cement
as binders are becoming increasingly costly. Alternative hydraulic bound mixtures
using granulated blast furnace slag as binders are becoming attractive to highway
authorities, being more cost efficient. These mixtures are produced and laid cold,
and continuously develop strength over a period of one year. Whilst numerous studies have demonstrated the potential use of SBM in road construction, there appears
to be little previous research, if any, looking at developing early life high strength gain
mixtures achieving a comparable strength with cement bound mixtures.
This paper presents the results of a series of laboratory tests obtained on selected
mixtures of SBM. Effects of mix design on the early life strength in particular have
been investigated and the results are compared to cement bound mixture using the
same aggregates. The paper also presents the performance of these mixtures following an extensive site trial which investigated the un-surfaced characteristics of
these high strength gain SBM and their performance with a thin asphalt overlay.
158
1. Introduction
The objective of this paper is to enable more efficient and innovative use of slag
bound mixture (SBM) incorporating blast furnace slag (BFS) and Basic Oxygen slag
(BOS) in road construction. The development of high performance slag bound mixtures is of particular interest in road construction as it can provide direct performance
comparison to cement bound mixtures. The use of these materials in construction
applications supports Government policies of sustainable construction. SBM, which
are defined as mixtures that use slag from iron and steel production as the main constituent or constituents for the hydraulic combination or binder, are excellent examples of these sustainable mixtures as they are only reliant on the minimal use of
manufactured resources.
In this age of global warming and limited resources, there is increasing pressure on
businesses and material suppliers in particular to reduce their impact on the environment. The use of recycled materials and industrial by-products, particularly for
construction, is viewed as an integral part of this drive to greater sustainability. In
road construction, one of the ways to help attain this goal is the greater employment
of hydraulically bound mixtures (HBM) like SBM in particular.
SBMs are generally characterized by the use of granulated blast furnace slag (GBS)
as the main binder. A wide range of aggregates and activators can be used in these
mixtures
providing
a
range
of
properties,
different
rates
of
setting
and
strength/stiffness development. The flexibility in material selection and properties
provide many opportunities for developing SBM mixtures at competitive cost to
CBGM. Unlike cement bound mixtures, slag bound materials have the added advantage of being immediately trafficked after placement avoiding the need for curing
period to achieve a set strength. In addition they are more tolerant to increased water content and usually have a wider workability window than cement bound mixtures.
These advantages will accelerate the construction period leading to cost savings in
labour, plants and traffic management.
One particular aspect of the SBMs materials is that the mixture can be designed to
achieve a desired strength development. Unlike cement bound materials, increasing
159
the amount of binder in SBM mixtures to achieve higher strength will not increase the
risk of thermal cracking leading to reduced durability. This is generally attributed to
the slower hydration rate and reduced generated heat of the GBS comparing to cement hydration.
Prior to 2006, UK practice for the construction of pavement foundations employed a
method specification, which prescribed the materials to be used for the constituent
layers of the foundation, their thicknesses and how they should be compacted. This
approach generally restricted the contractor’s choice of materials to conventional materials with known behavior. For the majority of pavement types, there was no reduction in the thickness of the foundation layers or the pavement when superior materials were used in the foundation. At that time European countries specified minimum
elastic stiffness values at various levels in the foundation as a performance measure.
Within the UK, various research activities have been carried out to develop a performance based specification for road foundations.
It was asserted that lower class
foundations with the lower pavement support or stiffness could be built with unbound
granular materials, whereas higher class foundations of higher pavement support or
stiffness would require construction with hydraulically bound materials in their upper
layers. Thinner pavements were proposed for the higher quality foundations.
Conventionally slag bound mixtures have been demonstrated over the last 4 decades
to perform as foundation layer and their behaviour is well understood. However,
when compared to cement bound mixtures, slag bound mixtures failed to demonstrate high strength classes and early life strength development. The purpose of the
information presented in this paper is to demonstrate that Slag Bound Mixtures can
be designed to deliver a high strength class and also achieve a rapid strength gain
and in particular demonstrate equivalence with Cement Bound Granular Mixtures .
This paper briefly presents some of the findings from an internal study carried out by
Lafarget Tarmac on Slag Bound Mixture behaviour and performance.
160
2. Slag Bound Mixtures
Slag Bound Mixtures are versatile products which offer benefits in the efficient use of
materials; including the potential to increase the value of the material being recycled
(up-cycling), and the ability to consume a range of long stocks materials including
aggregates and quarry dusts.
The selection of the SBM type is based on whether a particular market or specification is to be satisfied or whether the production process is being established to exploit
a certain source of possible constituents. The latter will dictate the type of SBM that
can be produced and thus the market and/or specification that can be targeted.
SBMs are mainly characterised by the type of binder in being a reactive slag. The
use of SBM in the UK has largely centred round 2 materials: (1) the hydraulic constituent, ground granulated blast-furnace slag (ggbs) which is often used as cement replacement, and (2) the hydraulic binder is the granulated blast-furnace slag (GBS)
with no cement additions. The former mixture where ggbs is used as partial cement
replacement is actually classified as CBGM with CEM II cement type, however, these
mixtures can be developed with 100% ggbs as binder providing an alkaline activator
is added to the mixture which render them as SBM mixtures.
Ground granulated blast-furnace slag (ggbs), is the product that results from the
grinding of granulated blast furnace slag (GBS). GBS is the by-product of the production of iron in a blast-furnace. Initially it is a molten material that it ‘tapped’ or drawn
off from the furnace. If left to cool in air, the slag is known simply as air-cooled blastfurnace slag (ABS), a largely inert product, which is crushed and screened to produce excellent aggregate.
Alternatively, the molten blast-furnace slag can be quenched with water, which produces a glassy sand-like material. This quenching actually locks in the cementitious
potential of the slag producing a material with slow hydraulic (or cementing) properties. The constituent is a complex combination of calcium, silicon and aluminium oxides. This material is known as granulated blast-furnace slag (GBS). Effectively, GBS
161
is ‘cement’, which is illustrated perfectly by its ground form, ggbs, which is a constituent of approximately 50% of the ready mixed concrete produced in the UK.
As far as road construction is concerned, ggbs has been used widely in the last 20
years in soil stabilisation. It has proved particularly valuable in the UK when used in
combination with quick lime for the stabilisation of sulfate-bearing soils, being more
effective than either lime or cement alone or in combination or also providing resistance to soil volumetric expansion.
The Slag Bound Mixtures described in this paper are focused on the use of GBS as
binder.
3. Advantages of using SBMs
SBMs are composed of aggregates, granulated slag as binder, an activator and water. They are used to replace cement bound materials (CBGM) and Fly Ash Bound
Mixtures (FABM) in pavement construction. Whilst it is known that SBM possesses
lower strength characteristics than CBGM at early age, they can actually be designed
to provide accelerated early life strength development. SBMs are generally cheaper
than cement bound mixtures and have considerable advantages over CBGM. These
are summarised below:

SBM contains higher binder content than CBGM which facilitate the more homogeneous distribution of the binder within the mixture.

The binder in SBM is a fine aggregates and not a powder. This provides a
better skeleton distribution, aggregates interlock and more importantly reduced
interlinked voids.

The setting time of SBM takes a relatively longer time than cement bound mixtures, leading to increased workability duration, storage time and longer working period between production and final placement

Being a granular mixture, the aggregate interlock once compacted will enable
the SBM to be immediately trafficked with site traffic and plants. This will not
influence the integrity and performance of the monolithic mixture once set
162

SBM can be placed in wet conditions. Excess water is simply allowed to drain
off before placement and compaction without washing away the binder. Increased water content in the mixture is usually beneficial for the full hydration
of the GBS

The slower rate of setting allow the stiffness of SBM to increase progressively
with the increased traffic avoiding thermal cracking usually caused by rapid
hydration of the binder. The speed of the GBS hydration is mainly controlled
by the type of activator and nature of aggregates used in the mixture

SBMs are usually characterised by their one year strength and unlike cement
bound mixtures there is no need to wait for 7 days after placement to achieve
the set strength.

When compared to CBGM, the use of SBM reduces the emission of carbon
dioxide usually associated with cement manufacture.

Under frost, the hydration more so for cement than GBS binder is halted and
setting will continue once normal temperatures are reached. However, depending on the duration of the frost period it is likely that, once the hydration
re-commences, cement will not be fully hydrated comparing to GBS which will
impact on performance.
4. Designing SBMs for performance and durability
For HBMs, durablity is defined as “capablity of achieving the required performance
over the design life”. For example, a pavement foundation layer is required to support
the overlaying layers both during construction and during subsequent in-service life,
and to adequately distribute stress over the lower layers thereby preventing permanent deformation of the pavement foundation. The durability of an HBM layer can be
viewed as a function of the aggregate and of the mixture itself. When considering the
aggregate, durability can be quantified and evaluated as mechanical deterioration
(wear) and physio-chemical activated deterioration (soundness). When considering
the mixture as a whole, the volumetric stability allows structural stability, and hence
durability, to be evaluated.
163
If the HBM is not durable, the subsequent loss of performance can result in a lack of
support to adjacent layers in the pavement. Features such as surface deformations
(including rutting) and/or reflective cracking can occur. The actual durability requirements are dictated by the design performance. These might include whether a pavement layer is expected to break-up/crack, or whether the layer is expected to behave
more like a slab (in which case pre-cracks may be required to accommodate anticipated volumetric change), or how exposed the HBM will be to the influence of weather (temperature and water). Obviously, this will vary between applications; for example, a buried foundation layer below the depth of frost penetration maybe in a less
aggressive environment than an exposed working platform or erosion protection layer. The following (environmental and non-environmental) factors are likely to influence the durability of HBMs:

Volumetric changes;

Deterioration (including fragmentation) of unsound aggregate;

Sensitivity of the HBM to water;

Chemical attack/aggressive ground;

Adverse curing conditions (low temperatures, lack of water and so on); and

Deleterious substances, which may inhibit or limit strength gains (for example,
certain types of organic material and clay) or result in chemical attack on the
mixture (for example, sulfates sourced from within an HBM).
The factors that affect the durability of HBMs can be divided between those attributed
to environmental conditions (such as temperature and water), and those associated
with the material components of the HBM itself. Other factors include chemical reactions, which are often influenced by the presence of water but do not directly result
from the weather. If the durability of the HBM is inadequate, both environmental conditions and insufficient mechanical performance attributed to the HBM can result in
shortened in-service life spans, increased maintenance requirements or, ultimately,
costly remedial works. It is therefore vital that durability is evaluated at design and
construction stages.
164
5. Laboratory mix design
The laboratory mix designs for the CBGM and SBM were carried out in accordance
with BS EN 14227 ‘Hydraulically bound mixtures – Specifications’, Part 1 (CBGM)
and Part 2 (SBM). The CBGM is designed to be suitable for foundation and flexible
composite base layers and therefore Type B grading was adopted. For the SBMs
type B1 grading was adopted as being the only grading permitted for use in foundation and flexible composite base by HD26/06. For the SBM varying levels of GBS
content were adopted at design stage with hydrated lime being the activator in those
mixtures at different addition levels. A further SBM blend was considered at design
stage to achieve class B2-G2 with a targeted GBS and activator contents. For the
CBGM two levels of cement content were adopted. The maximum size aggregates
used in all mixtures was 20mm and the aggregate type used in the CBGM and SBM
was air cooled blast-furnace slag aggregates.
These mixtures were assessed
based on their compressive strength which has been harmonised to ‘cylinder’
strength rather than ‘cube’ strength. The graph below provides a comparable summary of the findings:
165
Corrected results using EN206 conversion factor
40
35
Compressive Strength (N/mm2)
30
25
CBGM B C8/10
CBGM B C16/20
SBM B1-1 + GBS1
SBM B1-1 + GBS2
SBM B2-G2 + GBS1
20
15
10
5
0
0
5
10
15
20
25
30
Age (days)
Figure 1: Compressive Strength development at design stage based harmonised to cylinder strength
The design requirement was to achieve a strength class equivalent to C12/15. Based
on the findings from the laboratory design CBGM B, mix C8/10 and SBM B1-1 with
GBS2 achieved the target strength. Those mixtures were proposed for the site trials.
An additional mixture was also proposed for construction which has an equivalent
mix design to SBM B!-1 with GBS1 and BOS slag as coarse aggregates
6. Site trial Layout
The trial layout was specifically designed to address the following criteria:







Performance of CBGM as sub-base layer overlaid by CBGM as road-base
layer
Performance of SBM as sub-base layer overlaid by SBM as road-base layer
Performance of CBGM as sub-base layer overlaid by SBM as road-base layer
Performance of SBM as sub-base layer overlaid by CBGM as road-base layer
Effect of pre-cracking on all layers
Effect of no pre-cracking on the SBM layers
Assess the potential of using thinner asphalt overlay of 100mm thick
166
The layout below presented in Figure 2 illustrates the constructed pavement:
Daily traffic volume:


45 loads of 44tons HGV
45 loads of 32 tons HGV
Asphalt overlay
100
Roadbase: SBM B1-1 or CBGM B C8/10
mm
200
Sub-base: SBM B1-1 or CBGM B C8/10
mm
150
mm
Existing Unbound Sub-Grade layer – BFS slag
Figure 2: Cross Section of the pavement layout
All CBGM materials were pre-cracked at 3 m centres. The SBM layers were also
pre-cracked at 3 m centres except one area 12m by 12m of the SBM laid at sub-base
and roadbase levels was not pre-cracked. Figures 3a and 3b show the plan layout of
these materials and the un-cracked area relative to the pre-cracked layers
150mm
CBGM
C8/10
150mm
B
SBM
150mm
B1-1
(BFS)
(BOS)
No
70 m
SBM B1-1
12m
pre-
cracking
12m
167
Figure 3a: Plan showing the Sub-base layers
200mm
CBGM
200mm
B
SBM B1-1
(BOS)
C8/10
70 m
No
12m
cracking
12m
200mm
SBM
pre-
B1-1
(BFS)
20 m
Figure 3b: Plan showing the Roadbase layers
7. Trial construction
The construction of the base layers commenced on Tuesday 2nd April 2013 and
completed on Friday 5th April 2013. On completion the CBGM and the SBM layers
were sealed to prevent water ingress during the hydration.
No solid edging was used at the perimeter of the trial sections. Instead the material
was laid un-compacted at a gradient of approximately 450 in order to provide lateral
168
restrain to the constructed layers. The compaction of all the materials commenced
immediately after laying using a 14 tonnes ballasted PTR for a minimum of 10 passes. A steel drum roller followed on and provided further surface compaction. The
PTR compaction is needed to provide high compaction level at depth leading to an
optimised aggregates interlock throughout the layer and the steel roller is usually required to provide compaction and sealing finish of the surface. This method of compaction enables the granular SBM to withstand site traffic immediately after compaction avoiding the need for curing periods of usually 7 days for a CBGM.
Pre-cracks were created in all the CBGM materials using a rotating disc cutter
mounted to a small excavator (known as ‘Pizza’ cutter). All the cracks were created
at 3m centres and to a depth of half the layer thickness. All cracks were filled with
K140 bitumen emulsion. Dedicated areas of the SBM in the sub-base and roadbase
layers were specifically not pre-cracked to enable the evaluation of these mixtures in
uncracked status.
On completion all HBM materials were sealed using a tack coat to ensure no loss in
moisture is achieved through evaporation. The top of the roadbase layers were left
exposed for a period of 6 months prior to the installation of the asphalt overlay in order to evaluate the surface characteristics of both CBGM and SBM materials. Only
limited traffic was allowed on these sections during the surface monitoring period.
During the construction of the site trial the ambient temperature ranged between 60C
and 80 C. At night the temperature dropped to just above freezing and a temperature
of 10C was recorded overnight. The weather was generally dry and windy.
8. Material testing and performance
A testing strategy was developed ahead of the site works to ensure quality control
over the laid materials. The testing protocol included the following procedures:
1- During construction
169
(a) LWD testing
The existing conditions of the unbound sub-grade layer are evaluated using LWD
testing. It is observed that the sub-grade layer is made of mixture of BFS slag and
some steel slag. Prior to testing the layer was compacted using a steel drum roller.
LWD drop points were carried out at a grid of 10 m by 10 m across the whole site
area. Through this test testing a soft low spot was detected. This area was built to
the same level as the rest of the site using a mixed CBGM material. The LWD testing provided the horizontal mapping of the surface stiffness of the sub-grade. The
graph below presented in Figure 4 shows the spread of the surface stiffness of the
sub-grade across the site. The averages of this data are included in the Table 1
During the installation of the sub-base and base layers, LWD testing was also carried
out soon after the completion of compaction and before the laying of the next layer.
None of the drop points aligned above the pre-cracks. Sufficient LWD data was generated for each layer to identify any variations within the same layer which may be
due to irregularity of laying, uneven compaction, etc. A summary of the results are
presented in the Table 1 below.
LWD Results - Unbound Sub-grade
350
300
Su rfa c e Stiffn e s s (M P a )
250
200
Eo MPa
Eo MPa
150
100
50
0
0
2
4
6
8
10
12
Drop Points
Figure 4: Surface stiffness modulus for the sub-grade across the site area
170
Mean E0
No of
Max
Min
(MPa)
tests
Sub-Grade – Day 1
153
10
235
81
Sub-Grade – Day 2
161
10
284
93
CBGM Layer 1 – Age=2 hours
107
16
156
81
CBGM Layer 2 – Age=2 hours
106
7
115
100
CBGM Layer 1 – Age=24hours
224
4
255
187
CBGM Layer 2 – Age=7 days
1931
16
2476
1357
SBM (BFS) Layer 1 – Age=2 hours
163
13
234
86
SBM (BFS) Layer 2 – Age=2 hours
158
30
296
106
SBM (BFS) Layer 2 – Age=7 days
1247
24
1921
703
SBM (BOS) Layer 1 – Age=2 hours
122
3
142
105
SBM (BOS) Layer 2 – Age=2 hours
110
10
134
85
12
839
412
Layer
SBM (BOS) Layer 2 – Age=7 days
638
0
0
(*) Note: Surface temperature varied between 8 C and 10 C during the LWD testing
Table 1: Development of Surface stiffness modulus with time for the sub-base
and Roadbase layers based on LWD
(b) In situ moisture and density measurements
During the construction of the sub-base and base layers the in situ moisture and
density for each layer of the laid mat were measured. These measurements were
taken on each layer after compaction is completed
In addition and during the mixing of the CBGM and SBM materials the moisture content was also determined throughtout the prodcution process to ensure that the set
moisture target for each mixture is achieved.
(c) Site Specimens
171
CLASSIFICATION TESTING
During the laying of each material, bulk samples were safely extracted from each material and duplicate laboratory cylindrical specimens (100mm diameterx 200mm
height) and cubes (100mm x 100mm) were manufactured, each were wrapped in
cling film and placed in sealed plastic bag. All the CBGM specimens were prepared
on site and within 3 hours from mixing. All the SBM bulk samples were placed in
sealed plastic bags as loose materials and brought back to the laboratory where all
the specimens and cubes were made. This was carried out within 1 day after production and sampling.
Sufficient bulk samples were extracted from each material and throughout the loads
to manufacture duplicates cylindrical specimens and cubes. The cylindrical specimens were dedicated for the tensile and elastic stiffness testing and the cubes were
tested for compressive strength. For the CBGM, all specimens were cured at 200C.
For the SBM materials one set of specimens was cured at 200C and the other set
was cured at 400C. Three specimens were prepared for each age, and from each
material and for each type of test (i.e., compressive, TS and EM):
The specimens were tested at the following ages:
- 7 days after manufacturing (cured at 20C for CBGM)
- 7 days after manufacturing (cured at 40C for SBM)
- 14 days after manufacturing (cured at 40C for SBM)
- 28 days after manufacturing (cured at 20C for CBGM)
- 28 days after manufacturing (cured at 40C for CBGM
- 90 days after manufacturing (cured at 20C for SBM)
- 180 days after manufacturing (cured at 20C for SBM)
- 360 days after manufacturing (cured at 20C for SBM)
Not all genertaed data is presented in this paper and only selected findings are reported. Figure 5 below shows a comparison in compressive strength results between
‘at design’ stage and ‘as built’ stage
172
Comparison between laboratory and site produced materials - Compressive Strength
35
Compressive Strength on cubes (MPa)
30
25
SBM (BOS) (Site)
CBGM (Design)
CBGM (Site)
SBM (BFS (Design)
SBM (BFS) (Site)
20
15
10
5
0
0
5
10
15
20
25
30
Age (Days)
Figure 5: Comparison in compressive strength development between design
and site stages
DURABILITY TESTING FOR SBM MIXTURES
Additional cubes were manufactured from the SBM mixtures for durability testing
(EFFECT OF WATER IMMERSION ON ELASTIC MODULUS BY COMPRESSION). Ec is the
mean average elastic modulus of 3 specimens after 14 days sealed curing at 400C
followed by 14 days full immersion curing in still water at 400C for SBM mixtures. For
CBGM mixture Ec is the mean average strength of 3 specimens after 14 days sealed
curing at 200C followed by 14 days full immersion curing in still water at 200C. All the
specimens were manufactured from the laid materials, using the same method of
manufacture. Table 2 presents a summary of the results
Mix Type
Method
curing
of Age at test Elastic
(days)
by
modulus Retained
compression rate (%)
(GPa)
173
CBGM C8/10
14
days
at
0
20 C at 95%
28
40
76.6%
28
13.5
65.2%
28
9.3
Data not
room humidity followed by
14
days
at
0
20 C in water
SBM
(GBS2)
B1-1 14
days
at
400C at 95%
room humidity followed by
14
days
at
400C in water
SBM
(BOS)
B1-1 14
days
at
0
40 C at 95%
room humidi-
available at
ty followed by
design stage
14
days
at
0
40 C in water
Table 2: Retained rate of Elastic Modulus by compression after soaking
9. Conclusions
One of the advantages of HBMs is that they can be mixed using mobile plants and
paved using conventional equipments. Irrespective of the binder type used in the
HBM materials, the nature of the aggregates whether it is limestone, Blast Furnace
slag, Basic Oxygen slag, or other, this will impact on the mechanical performance
and durability of these mixtures. The selection of the mixture components is usually
based on the proximity of the aggregates source, commercial viability, supply cost
and overall competitive bidding cost, it is vital that the impact of material selection on
the mixture performance is also evaluated. Most contracts are generally priced on
cement bound mixtures and based on compressive strength requirements. However,
considering the restrictions that CBGM material impose on contracts in terms of
workability window, curing period of 7 days, immediate pre-cracking requirements,
174
high cement cost, avoiding immediate trafficking by site plant, high susceptibility to
wet conditions, these issues have caused conflicts on numerous contracts which lead
in certain cases to major claims. Most of these issues which are related to construction and installation can be largely avoided by using SBM.
It is evident that the site trial has demonstrated that slag bound mixtures can be designed to achieve high strength and rapid strength gain which are comparable to
CBGM.
The surface of the HBM materials have been exposed for nearly 6 months with no
asphalt overlay and limited traffic loading applied, these materials have not shown
any sign of deterioration in terms of deformation, cracking, softening and failures at
surface level and at the un-supported edges. This demonstrate that these mixtures
are robust and resistant to weather conditions when exposed. The asphlat overlay is
planned for installation by 2nd October 2013 and the full traffic loading will be applied
thereafter.
175
Theme 4
1. From Research to Applications
176
Feldrappe, V. and Ehrenberg, A.
Development of new CEM X cements based on Ground granulated
blast furnace slag, fly ash and clinker
FEhS – Institut für Baustoff-Forschung e.V., Bliersheimer Straße 62, 47229 Duisburg,
Germany
Abstract
Today the manufacturing of cements with less Portland cement (OPC) clinker is one
of the most important levers in order to reduce the specific CO2-emissions within the
cement industry. Currently such cements are already standardised in EN 197-1 as
CEM II- to CEM V-cements. But these cements consider only a certain selection of
potential combinations of standardised cement constituents.
The present article illustrates impressively the potentials but also the limitations to
manufacture high-performance cements with less OPC clinker, ground granulated
blast furnace slag (GGBS) and fly ash. The consideration of EN 197-1 technical requirements was only one important aspect. But the performance compared to the
established CEM III/A-cements was rather the focus of the investigations. This includes workability properties but also the results of preliminary concrete tests in respect to fresh and hardened concrete properties as well as durability aspects. Furthermore the specific CO2-saving potential was calculated in comparison to
CEM III/A-cements.
Introduction
Sustainability gets an increasing importance in the global society. In this context cement industry is endeavours to limit the need of natural resources and to continuously reduce the specific CO2-emissions of the cement production. One of the most important actions on this is the production of cements with a significant reduced OPC
clinker content. Although CEM III-, CEM IV- und CEM V-cements are already stand177
ardised in EN 197-1, only less experiences are available in Germany for production
and application of CEM IV- and CEM V-cements. However those cements are already successfully applied in different applications in ready-mixed concrete and precast industry in other European countries. Furthermore the currently standardised
cements cover only a fraction of all potential compositions. From this point of view it
is reasonable to check also the performance of those combinations, e.g. out of
GGBS, fly ash and OPC clinker, which were not standardised yet.
The FEhS-Institute together with the Association of the German Cement Industry
(VDZ) carried out an extensive research program, which received public funding by
the AiF [9]. The following paper deals with potentials but also with constraints of optimised cements produced with a reduced OPC clinker content as well as GGBS and
fly ash as further main constituents.
Scope of work
The market share of Portland cement amounted still to approximately 35 % in Germany 2011 [10]. It is comparatively high compared to other European countries. As of
the year 1997 production of cements with GGBS was pushed. In Germany GGBS as
by-product of hot metal production is almost exclusively used in cement production
[11]. In contrast, fly ash, originated as exhaust gas filter residue of hard coal combustion during power generation, is mainly used as concrete addition. It is less applied in
CEM II-cements [12]. The production of CEM IV- and CEM V-cements, however, is
negligible in Germany.
In general cement production is not only an energy and raw material intensive process. Moreover it is also linked with high specific CO2-emissions. Therefore the industry endeavours to reduce the specific demand for raw material and primary energy as well as the emissions of greenhouse gases like CO2. Process optimisations of
[9]
[10]
[11]
[12]
178
kiln and grinding facilities were a successful lever in this context [13]. But their potential is nearly exhausted meanwhile [14]. Positioning of new cements with less clinker
and further main constituents in the market was another action during the last years
[15,
16
]. That way will become increasingly more important in future. However it has to
be kept in mind that the performance of new cements has to be comparable to those
of conventional cements in order to ensure durable and sustainable concrete constructions also in future.
Current European standardisation activities by CEN/TC 51 take this development into
account. So new cement types which consist out of GGBS, clinker and limestone or
clinker, limestone and fly ash respectively will be included in the revised European
cement standard EN 197. Thereby the clinker content can be further reduced relative
to the currently defined cements [17]. In contrast to that the scope of the presented
research work was to assess the binder performance of combinations consisting of
GGBS, OPC clinker and fly ash systematically, because the current cement standard
covers only a marginal part of these combinations. Furthermore they are also not in
the scope of current standardisation activities.
Testing program
Main parameters affecting the performance of cements are the reactivity and the
fineness as well as the quantitative composition of the constituents. Two different
reactive GBS, OPC and fly ashes were chosen. The GBS covered a typical spread of
available qualities in central Europe. Both GBS were ground to Blaine fineness of
4200 cm²/g and 5500 cm²/g respectively. The different C3S-content of the OPC clinker was used as a unit, to distinguish the reactivity of both CEM I 42.5 R – well aware
that it is only one of many factors and the cements are controlled to a similar performance by adjusting e.g. fineness or sulphates. Both fly ashes showed a considerably
different Blaine fineness besides the different reactivity measured as reactive SiO2.
13]
[14]
[15]
[16]
[17]
179
All raw materials were chemically, mineralogically, physically and performance orientated tested. The relevant properties important for characterizing the material parameters are compiled in Table 11.
In order to handle the wide range of potential compositions and material properties
systematically and efficiently, a design was planned with statistical methods. The experimental area, which reflects the defined limits of cement compositions, is shown in
Figure 28. Based on the 15 compositions, shown in Figure 28, 108 material combinations were considered. All cements were produced by mixing the separately ground
raw materials. In order to adjust the sulfate content of the CEM X-cements anhydrite
was added up to a SO3-content of 3.5 M.-%. As a first step mortar strength according
to EN 197-1 after 28 days and heat of hydration development during 7 days with an
isothermal heat flow calorimeter were determined. All measurement results were statistically assessed and relevant influencing factors were identified. With these factors
mathematical models were established for each result parameter (mortar strength
and heat of hydration) [9].
Unit
(C+M)/S
GGBS (S)
S1
S2
1,53
1,15
-
Al2O3
CEM I 42.5 R
(C)
14,0
11,8
C1
C2
V1
V3
-
-
-
-
-
-
23,2
TiO2
0,81
0,47
-
-
C3S
-
-
55
-
-
C3A
-
C4AF
Glass content
26,2
0,94
0,31
64
-
-
20
9
-
-
-
7
8
-
-
-
-
6
5
-
-
100
99
-
-
74
64
M.-%
C2S
Fly ash (V)
Reac. SiO2
Vol.-%
-
-
-
-
36,4
41,3
True density
g/cm³
2,917
2,923
3,142
3,116
2,675
2,543
-
-
6654
19141
3540
4620
3010
4970
BET
Spec.
cm²/g
fineness Blaine
4230 5210 4260 5430
PSD
d'
µm
15
12
16
13
21
12
33
23
n
-
0,79
0,99
0,73
0,88
0,85
0,87
1,00
0,84
180
Activity
index
7d
28 d
%
81
-
60
-
-
-
-
-
104
-
88
-
-
-
-
-
Table 11: Properties of raw materials
Figure 28: Experimental area for the
GGBS (S), OPC (C), fly ash (V)
three
component
system
Newly developed cements should meet quality requirements and customer expectations of current CEM II/B- and CEM III/A-cements which are well accepted on the
market. Based on the proved mathematical models 23 additional cements were defined with a compressive mortar strength of ≥ 32.5 MPa. Detailed cement and preliminary concrete tests were carried out in order to assess their performance in comparison to currently established cements. The examinations were completed by an assessment of possible CO2-emission savings depending on the compositions and performance of the cements.
Results
Mortar strength after 28 days
As expected, the statistical assessment of the 108 compositions showed a significant
impact of the cement composition on the mortar strength after 28 days. Also the reactivities of OPC clinker and GGBS were important influencing factors. In respect to the
181
chosen boundary conditions, GGBS fineness and kind of fly ash had, however, only a
secondary influence on 28 d mortar strength. A model of high quality was established
with a regression coefficient of 96 % and a standard deviation of 3 MPa. Figure 29
contains the graphical evaluation of the model for the example of cement C2. The
area of standardised cements according to EN 197-1 was marked in white for an easier interpretation of the cement compositions. The model was assessed with 7 additional tests which were not part of the statistical design. According to that the mortar
strength after 28 days is predictable to a good approximation depending on cement
composition and raw material reactivity.
In order to ensure a mortar strength after 28 days of > 42.5 MPa, especially the fly
ash content has to be limited. Independently of quality and reactivity of the other
main constituents, a fly ash content of 30 M.-% is maximum allowable for all cements
containing less than 65 M.-% OPC clinker and respecting the boundary conditions.
The maximum content can be exceeded if raw materials of higher quality and reactivity are used. As the example of Figure 29 shows, this value can be increased up to
40 M.-% when a more reactive GGBS is used. The influence of quality and reactivity
increases continuously with lowering OPC clinker content of the cements. As Figure
29 also illustrates, a mortar strength of 42.5 MPa after 28 days was not achievable in
any case with inappropriate combinations. All in all, one can say that higher GGBS
reactivity at simultaneously constant OPC clinker quality allows, therefore, the use of
higher quantities of lower performing cement raw materials like fly ash or limestone.
182
Figure 29: Predicted 28 d mortar strength for cements with
clinker C2 and GGBS S1 (left) and S2 (right)
* C1; S1 or V1
** C2; S2 or V3
Figure 30: Strength development of some of the additional tested cements
An excerpt of the strength development of the 23 additional cements is shown in Figure 30. According to the selection criteria all cements had compressive mortar
strength of more than 35 MPa, so that the standard requirement concerning strength
class 32.5 is achieved. However 28 day mortar strength of commercial German cements is normally close to the upper strength level of the relevant strength class. Figure 30 clearly shows that such a strength level can be achieved if the fly ash content
is limited as already mentioned before. Mortar strength of 50 MPa after 28 days is
achievable even with CEM X/B- or CEM III/C-cements containing only 31 or 20 M.-%
OPC clinker respectively.
Heat of hydration and early strength
The statistical findings concerning the tests of heat of hydration development after 2
days were that also the GGBS and fly ash fineness are highly significant beside of
the cement composition. Also some interaction between portions of main constituents
and material parameters were significant. The regression coefficient of the model
established out of the significant parameters was 81 % with a standard deviation of
183
23 J/g. Similar findings were gained concerning the heat of hydration development
after 7 days. The fly ash fineness, however, was not relevant any more. The quality
of the model was also high even if the regression coefficient was slightly lower. All
cements within the boundary condition released specific heat of hydration of at least
50 J/g after 2 days and 150 J/g after 7 days respectively. As expected the heat of
hydration development increased with increasing fineness of the main constituents
[9].
Figure 31: Relationship between compressive mortar
and specific heat of hydration after 2 and 7 days
strength
The early compressive mortar strength after 2 and 7 days was determined on 23 additional cements, which were chosen on the basis of the mathematical models. A linear correlation between specific heat of hydration and early strength exists as presented in Figure 31. This correlation ensures a transformation of the modelled specific heat of hydration into compressive strength as Figure 32 illustrates by way of example.
184
Figure 32: Transformation of specific heat of hydration into strength by way of
example for cements with GGBS S1, cements C2 and fly ash V3
The early strength of all 23 additional cements were in a wide range between 5 and
21 MPa (after 2 days) and between 18 and 39 MPa (after 7 days) respectively. The
majority of the 2 day strength results, however, were higher than 10 MPa. Consequently the early strength requirement of strength class 32.5 N is fulfilled by all cements. Most CEM X-cements, however, reached even the requirements for strength
class 32.5 R and 42.5 N easily.
Based on the mathematical models as well as on the results obtained by testing 23
additional cements it can be assumed that competitive cements are producible also
in respect to their early strength. But the restrictions e.g. concerning fly ash content
already made for the 28 d compressive strength are also valid for the early strength.
Properties according to EN 197-1
Beside of the strength development also the physical (setting, soundness) and chemical (loi, insoluble residue, SO3- and chloride content) properties of EN 197-1 were
tested. Furthermore the water demand and the consistency measured as mortar flow
spread at constant w/c ratio of 0.50 were also determined. All requirements of EN
197-1 were fulfilled by all 23 additional chosen cements. The initial setting time increased while a smoother consistency, that means a increasing flow spread arose
185
with decreasing OPC clinker content of the cements. The water demand of all cements was in the range between 26.0 and 29.5 M.-%. All in all the physical cement
properties which are market relevant concerning handling in concrete are similar to
those of commercial cements.
Preliminary concrete tests
Even if the development of new cements is initially done by paste and mortar tests
the performance in concrete is decisive. Consequently preliminary concrete tests
concerning fresh and hardened properties as well as durability aspects were carried
out. 5 cements (2 CEM X/A, 2 CEM X/B and 1 CEM III/C) having a mortar strength
after 28 days of approximately 50 MPa were used. The concrete composition was in
line with minimum requirements of certification tests defined by Deutsches Institut für
Bautechnik (DIBt) for exposition class XF3 according to EN 206-1. The cement content was 300 kg/m³ concrete and the w/c ratio 0.60. The aggregates fulfilled the MS18
criterion of EN 12620. All concretes were stored after producing 1 day in mould, 6
day under water at 20 °C and 21 days in climate at 20 °C and 65 % rel. humidity according to the national annex of EN 12390-2.
The fresh concrete properties – flow spread, air content and bulk density – were in
the range of normal concrete. The compressive strength follows the relationship between concrete strength, w/c ratio and cement strength which is well known as
"Walz-curve". The strength development of the cements in concrete was comparable
to those of commercial cements.
Beside of the fresh and hardened properties also the carbonation behavior was tested. For concretes stored in lab climate at 20 °C and 65 % rel. humidity the carbonated surface layer was between 4.0 and 5.5 mm after 180 days. A comparison with
literature values indicates that the carbonation behavior is comparable and a sufficient carbonation resistance should be given.
186
Ecological examination
To put it simply the CO2-emissions of the cements were calculated from cement
composition and specific CO2-contributions of cement main constituents. For the 3
main constituents the specific CO2-contributions were assumed for GGBS by 0.10 t/t,
for OPC by 0.89 t/t and for fly ash by 0.05 t/t. The result is shown in Figure 33 for the
compositions out of GGBS S1 and cement C2. But this figure is valid for all compositions because no differentiation between the different GGBS, OPC or fly ash was
made. The accounting of specific CO2-emissions is, of course, strongly depending on
the chosen input parameter. For this purpose a clear decision support was given in
EN 15804. Referring to GGBS it was declared in a position paper by VDEh "Blast
Furnace Committee", to perform allocations for CO2-emissions on the basis of economical values according to this standard [18].
Figure 33: Calculated CO2 saving potential for cement compositions
with GGBS S2 and cement C1
Due to the significant lower specific CO2-contributions of GGBS and fly ash the CO2emissions of the cements decrease, of course, with declining OPC content considerably as it is shown in Figure 33. This behavior is independently of the relative contents of the other main constituents. In order to calculate CO2-emission savings the
de facto cement strength is deciding. Therefore the areas for a predicted cement
[18]
187
strength of ≥ 42.5 MPa (to the left of solid line) and ≥ 52.5 MPa (to the left of dashed
line) were marked in the example of Figure 33.
According to this model calculation for the example of Figure 33, a CO2-saving potential of up to 55 % compared to CEM III/A with 50 % GGBS is possible for CEM Xcements which have a compressive strength of more than 42.5 MPa after 28 days.
Even for cements with more than 52.5 MPa after 28 days a saving potential of up to
20 % is still realistic. Especially the substitution of OPC by GGBS contributed to constant cement strength. This example indicates that the non-standardised area between CEM III/A- and CEM V-cements is in particular very promising concerning realisable cement strength and potential CO2 savings.
Conclusions
By using design of experiments and statistically assessment tools a general relation
between the different impact parameter and the performance of cements consisting
of GGBS, OPC and fly ash was developed statistical safeguarded. Within the chosen
boundary conditions the potential performance of any cement composition can be
estimated with the established mathematical models. As a result potentials and constraints are easily assessable depending on the available raw materials.
Manufacturing of cements according to strength class 42.5 is possible within a wide
range. This includes also compositions deviating from the current cement standard
EN 197-1. If the fly ash content is limited accordingly properties can be achieved
comparable to those of commercial cements. It could be shown by the preliminary
concrete tests that the properties of fresh concrete as well as the strength development of hardened concrete are similar to concrete made with cements according to
EN 197-1. The carbonation behaviour of the concrete was also proved similar to
those of concrete with commercial cements. Further research demand is, however,
still given in particular concerning durability aspects to lay the conditions for a later
industrial use as well as the basis for standardisation. A new research project is currently planned for this purpose.
188
Beside of their technical potential cements with a high content of GGBS and fly ash
may also make an important ecological contribution. Depending on performance of
the raw materials the specific CO2-emissions can be reduced up to 55 % compared
to a CEM III/A-cement with 50 % GGBS which already has a significant advantage
concerning consumption on resources and specific CO2-emissions compared to
OPC. An application of those cements will preserve natural resources and accelerate
the utilisation of industrial by-products in high quality applications.
Acknowledgement
The IGF project 16148 N of the "VDEh-Gesellschaft zur Förderung
der Eisenforschung" was founded via the AiF within the scope of
the programme for the support of the cooperative industrial research (IGF) of the Federal Ministry for Economy and Technology
according to a decision of the German Bundestag.
Literatur
[1]
[2]
[3
[4]
[5]
[6]
[7]
[9]
[10]
Gemeinsame Nutzung von Hüttensand, Steinkohlenflugasche und Portlandzementklinker zur Herstellung optimierter Zemente und Betone, Abschlussbericht des AiF-Forschungsvorhabens 16148 N, 2012
Verein Deutscher Zementwerke e.V. (ed.): Zahlen und Daten 2010-2011,
Düsseldorf, 2011
Ehrenberg, A.: Hüttensand – Ein leistungsfähiger Baustoff mit Tradition und
Zukunft, Beton-Informationen 46 (2006) No. 5, p. 67-95, No. 6, p. 35-63
Hugot, A.: Flugaschemarkt der Zukunft, BVK/VGB Fachtagung: Flugasche im
Beton – Neue Anwendungen, Proceedings 708-08 (2008) V 5, p. 1-8
Hoenig, V., Schneider, M.: CO2 Reduction in the Cement Industry, Process
Technology of Cement Manufacturing, VDZ Congress 2002, Verein Deutscher
Zementwerke VDZ (ed.), Düsseldorf, 2003, p. 499-505
Verein Deutscher Zementwerke (ed.): Tätigkeitsbericht 2005-2007, Düsseldorf, 2008
Ludwig, H.-M.: Entwicklung und Einführung von CEM II-M-Zementen, Proceedings of 15th Internationale Baustofftagung (ibausil), Weimar 2003, Vol. 2,
p. 1415-1430
Ehrenberg, A., Geiseler, J.: Ökologische Eigenschaften von Hochofenzement,
Teil 1, Beton-Informationen 37 (1997) No. 4, p. 51-63
Wolter, A.: Trends in the field of low CO2 cements, Proceedings of 6th International VDZ Congress 2009, p. 78-81
189
[11]
Steel Institute VDEh: Position of the Steel Institute VDEh Blast Furnace Committee on the allocation for the production of hot metal and blast furnace slag /
granulated blast furnace slag, Düsseldorf, March 2012
190
Jin-man Kim1,Sun-mi Choi1,Ha-seog Kim2,Sea-hyun Lee2,Sang-yoon Oh3
Hydration Properties of Rapidly Air-Cooled Ladle Furnace Slag with
Gypsum
1 Kongju National University, 330-717, 275 Cheonan-daero, Cheonan-city, Republic
of Korea
2 Korea Institute of Construction Technology, 411-712, 283 Goyang-daero, Ilsnaseogu, Goyang-city, Kyonggi-do, Republic of Korea
3 Ecomaister Co., Ltd., 404-250, 104, 250 street, Geonjiro, Seo-gu, Incheon-city, Republic of Korea
Abstract
Steel making processes using an electric arc furnace include oxidation and reduction
processes. Although slag containing a large amount of iron oxide is discharged in the
oxidation process, slag with various compositions is discharged in the reduction process according to the type of reducing agents. This study was conducted to evaluate
hydration properties when mixing gypsum and LFS(Ladle furnace slag) powder produced by powdering Calcium Aluminate Based Slag discharged in the process using
AI as a reducing agent.
For the measurement of hydration properties, the setting, hydration heat and hydrates were analyzed in paste phase and compressive strength and length change
were measured in mortar phase, and the properties were compared with ordinary
potland cement(OPC) and regulated set cement(RSC).
The results showed that the RC-LFS with high content of quick setting mineral have
high reactivity while have various problems. In the case of adding gypsum, however,
hydration properties including hydration heat, early age strength, long-age strength
and length change were improved.
191
Introduction
LFS is generated in ladle furnace, a secondary refining furnace, when the degasifying, deoxidation and desulphurization of molten steel produced in steel making furnace is performed. For the desulphurization process of molten steel, the sulfur may
be removed by adding elements having strong affinities with sulfur, such as Ca and
REM, however, for the economic reason, a method called reduction fining process
using high-basicity slag having CaO as a main component is used mostly2). The LFS
may become useless in this process, however, due to self-pulverization caused by
increment of free-CaO, within few days when it was exposed to atmosphere at a molten state. The Al and Si species, a material added for deoxidation, other than CaO, is
effective resource as a construction material, however its value has been widely ignored and sent to landfill until recently.
The results of analysis of oxide composition contained in LFS generated domestically
showed that the component ratio of major oxide effective as a inorganic binder, such
as CaO, Al2O3 and SiO2, is similar with that of quick setting binder. The previous
studies, indeed, reported that LFS may be used as a inorganic binder with both
quick-setting and strength development properties when it is amporphized at high
cooling speed together with mineral compositions such as C12A7 and C2S3,4).
In previous study of our group, accordingly, a Rapidly cooled Ladle furnace slag
powder(RC-LFS) beads were produced by cooling LFS rapidly using high pressure
air and were powdered by pulverization, in order to make LFS into amorphus state
with reactivity. The powder, in the analysis of hydration properties, showed quick setting within few minutes and the developed strength was high enough to be expressed
during initial three hours. In the case of using only RC-LFS powder without adding
water, however, some problems including low development of long-age strength, high
shrinkage due to CAH hydrate which is major hydration product and high hydration
heat over 100 ℃ occurred5).
192
The C12A7, a major mineral of RC-LFS generates CAH in isolated hydration reaction,
however, various effects may be expected by generation of expansive ettringite when
the gypsum is added. The purpose of this study was, therefore, to seek a way to improve the hydration properties of RC-LFS powder and for sustainable reaction of hydration in order to propose a fundamental data used in using RC-LFS as a raw material of Regulated set cement(RSC). The gypsum, for the purpose, was added into
RC-LFS powder and the change in hydration properties as a binder with quick setting
property was analyzed.
Experimental plan
- Experimental plan
The plan of experiment is shown in Table 1. The Ordinary portland cement(OPC),
RSC and RC-LFS were used as binders. The optimal molar ratio of gypsum was calculated as 30%, therefore, the gypsum displaced RC-LFS powder at four ratios: 0,
20, 30 and 40%. The OPC and RSC were used as comparison group, therefore processed without adding gypsum. For the test items, the setting test and semi-adiabatic
temperature test were performed and hydrates was analyzed in paste state, and
compressive strength and length change by age were measured in mortar state.
Factors
Levels
Test items
⦁Setting time
Substitution
ratio of
gypsum(%)
OPC1), RSC2)
LFS3)
(0, 20, 30, 40)
Paste
⦁Hydration Heat
(Modified adiabatic temperature)
⦁Hydrates analysis
Mortar
⦁Compressive strength
⦁Length variation
1) OPC : Ordinary Portland Cement as the control without gypsum.
2) RSC : Regulated Setting Cement as the control without gypsum.
3) LFS : LFS with gypsum of pre-decided rate
Table 12 : Experimental plan
193
- Materials
The RC-LFS as experimental binder, and OPC and RCS as comparison binders were
used in experiments. Table 2 shows the results of analysis of physical properties and
oxide of each binder, and Table 3 shows chemical composition of gypsum used as a
admixture. RSC and CA-species binders were diluted, before using, by mixing 0.5
wt% of retarder and water, to obtain workability by quick setting property.
Physical properties
Binders
Density
Fineness
(g/㎤)
(㎠/g)
OPC
3.15
RSC
LFS
Oxide content(Wt.%)
SiO2
CaO
Al2O3
3,200
17~25
60~67
3~8
2.87
4,800
10~16
47~53 14~20
2.97
5,500
10.9
44.5
Fe2O3
MgO
MnO
0.5~6 0.1~4
26.6
-
3.0
2.5
-
4.3
6.6
0.6
Table 13 : Physical & chemical properties of binders
Type
HG
(Hemihydrate Gypsum)
MgO
Al2O3
SiO2
0.32
0.88
2.57
SO3
CaO
55.79 39.99
Fe2O3
SrO
0.41
0.04
Table 14 Chemical composition of gypsum
- Experimental Methods
The setting test of paste, mortar test and length change test were performed according to KS L ISO 9597 Determination of setting time and soundness of cements, KS L
ISO 679 Methods of testing cements-Determination of strength and KS F 2424 Testing method for length change of mortar and concrete, respectively. The analysis of
hydrates was performed using Scanning Electronic Microscope(SEM) and X-ray Diffraction(XRD).
194
Results and Discussion
- Setting time
Fig.1 shows setting properties of LFS and a binder made by adding 30% of gypsum
into LFS. The mixing of gypsum with LFS powder was impossible because it reacted
and was set immediately after hydration due to CA-species minerals with quick setting property. The gypsum-containing LFS binder showed approximate two minutes
of delay effect compared to LFS powder without containing gypsum. Fig. 2 shows the
setting property when 0.5 wt% of retarder was added into binder, indicating that the
organic acid component increased the ratio of fluorine-containing aluminum hydroxide gel in hydration products and that those gel inhibited formation and growth of
ettringite by covering surface of unhydrated particles6). The LFS powder without gypsum shows final setting after approximately 25 minutes, meaning faster initial and
final setting than RSC. The increase of displacement ratio of gypsum improved setting time and the addition of retarder accelerated the trend; displacement ratio of
40% and addition of retarder achieved over two hours of workability.
Figure 34 Setting time of LFS with or without gypsum
195
Figure 35 Setting time of LFS binders according th the use of retarder
196
- Heat of hydration(Simple adiabatic temperature)
The results of hydration heat test measuring semi-adiabatic temperature of binders
are summarized in Fig. 3. The maximum peaks of hydration heat were achieved within an hour in RSC and LFS due to initial rapid reaction. The LFS, especially, exhibited
hydration heat as high as 110 ℃, which leading to fluctuation of temperature causing
crack and affecting durability. The addition of gypsum, however, decreased the initial
heat of hydration as high as 110 ℃, to only 70 ℃ level, similar 60 ℃ in case of RSC,
meaning the alleviation of initial rapid calescence by hydration.
Figure 36 Hydration heat of LFS with gypsum compared with OPC and RSC
- Compressive Strength
Fig. 4 shows compressive strength by age. The three days strength of binder displaced with gypsum by 20 and 40% was similar to that of 28 days strength of OPC,
and rather lower in three hours and one day strength while higher in three days
strength than that of RSC. In the case of binder with 30% of gypsum displacement
ratio, especially, the hydration activity index(HAI) calculated as a ratio to 28 days
strength of OPC was 129%, much higher than that in case of LFS without containing
gypsum. It is considered, however, that, over 30% of gypsum displacement ratio is
undesirable because some cracks were observed in specimen with 40% of gypsum
197
displacement ratio, raising concern for risk of durability deterioration due to strength
deterioration and expansion caused by excessive formation of ettringite.
Figure 37 Compressive strength of binders
- Length Variation
Fig 5 shows length change by binder types. The specimens for length change were
measured after demolding for initial three hours in RSC and LFS, and after one day
in OPC. The length change was highest in specimen using only LFS, meaning higher
shrinkage compared to OPC and RSC, while the addition of gypsum decreased the
change into less than that of OPC, and in case of 40% of gypsum displacement, the
volume even increased. It is presumed that, in the LFS without gypsum, the CAH gel
and CSA hydrates formed near by unhydrated particle stop hydration reaction by inhibiting penetration of water, causing decrease of volume by CAH hydrates binding.
In the case of adding gypsum, on the contrary, the stability of volume is obtained by
compensation effect of shrinkage by expansion due to formation of ettringite from
reaction of C12A7 and gypsum.
198
- Hydrate Analysis(XRD, SEM)
Fig. 6 shows the component analysis of LFS and LFS with 30% of gypsum, by using
XRD. In the case of LFS without gypsum, unhydrated C12A7, β-C2S and CAH hydrates were observed even after one day, while in case of gypsum displacement ratio
of 30%, the ettringite was formed initial age(one hour) of hydration, and it increased
by one day.
Figure 38 Length variation of binders
199
Figure 39 X-ray diffraction patterns of LFS specimens according to the time
Fig. 7 shows images of state of hydration of LFS and LFS with 30% of gypsum at ten
and 30 minutes and one day after. by using SEM. In LFS, the progressive binding of
CAH hydrated generated within initial ten minutes and the formation of hydrates film
with plate form extended from unhydrated particle surface appear to contribute to
early-age strength, while the film inhibit hydration reaction by surrounding unhydrated
particle surface, and it seems to be a main factor affecting long-age strength and durability.
In the case of adding gypsum, the ettringite on short fiber at early age of hydration
was identified, and it continuded as the hydration time increased.
200
Figure 40 SEM Image of LFS according to the time
Conclusions
The RC-LFS powder was mixed with gypsum and the possibility of using as an inorganic raw material, and the results are as follows;
1) The use of both gypsum and retarder resulted in longer initlal and final setting time
compared to use of only retarder, some hydration delay effect was also observed.
2) The addition of gypsum decreased the maximum heat peak of initial age from over
100 ℃ to 65 ℃, meaning the alleviation of initial rapid calescence by hydration.
3) The compressive strength in case of adding gypsum was 20 MPa for three hours
strength and the three days strength was similar to that of 28 days strength of OPC,
and rather lower in three hours and one day strength while higher in three days
strength than that of RSC.
4) The length change was highest in using only LFS, while the increase of gypsum
displacement ratio enabled the stability of volume by compensation of shrinkage using expansion.
201
5) The results of analysis of hydrates showed the formation of ettringite by addition of
gypsum, accordingly the long-term durability was obtained by sustainable hydration
reaction.
Acknowledgments
This study was supported by the R&D Center for Valuable Recycling(Global-Top Environmental Technology Development Program) funded by the Ministry of Environment. (Project No. : GT-11-C-01-210-0) & a grant(Code 11-Technology InnovationF04) from Construction Technology Innovation Program(CTIP) funded by Ministry of
Land, Transportation and Maritime Affairs(MLTM) of Korean government
Reference
[1] Korea Iron & Steel Association, 2012
[2] J.K Yoon, J.D Shim, Ferrous Metallurgy for Specialists, 2004, pp.240-262
[3] F.M. Lea, Formerly Director of Building Research, The Chemistry of Cement and
Concrete, 1997
[4] H.F.W. Taylor, Thomas Telford, Cement Chemistry, 2nd edition, 1997
[5] S.M. Choi, J.M. Kim, ZEMCH 2012, Hydration Properties of Inorganic Binder Produced from EAF Reducing Slag, 2012
[6] P.Kumar Mehta, Paulo J.M.Monteiro, Concrete, 4th edition, 2013
[7] J.H. Kim, J.M. Kim, ZEMCH 2012, A Study on the Hydration Characteristics of
EAF Reducing Slag by The Use of Gypsum
[8] J.S. Choi, Inha University, A Study on the Manufacture of CSA Cement Clinker
Using EAF Reducing Slag, 2010.
[9] Y.J. Ahn, I.K. Han, J.S. Choi, J. of Korean Inst. of Resources Recycling, Hydration
Property of Electric Arc Furnace Reduction Slag, Vol.19, No.6, 2010
202
J. Roininena,* and V. Kuokkanenb
USING GRANULATION (PELLETIZING) TO INCREASE THE USAGE
OF SLAGS
a
Research Engineer, Laboratory of Process Metallurgy, University of Oulu, P.O. Box
4300, 90014, Oulu, Finland
b
Researcher, Department of Chemistry, University of Oulu, P.O. Box 3000, 90014
Oulu, Finland
* Corresponding author. E-mail address: juha.roininen@oulu.fi
Abstract
Large quantities of industrial fine sized particle by-products remain without effective
utilisation within the European Union. Several research projects aiming to increase
the utilisation of fine particle by-products have been completed, but successful commercial and industrialized activities are still lacking due to the prevailing economic
conditions and the absence of proper business incentives. However, the current development in waste taxation is providing these incentives with possibilities for considerable innovative solutions to utilise these material flows. As a result, two separate
ash granulation plant projects have been launched simultaneously in Northern Finland. This will increase production capacity as earlier (for more than a decade) there
have been only two granulation plants (in Eastern Finland) serving the whole country.
This kind of breakthrough for fine sized industrial material handling is starting to take
place in Finland, and hopefully also in other European countries as there are, especially with the usage of bio-based ash as a forest fertilizer, proper business incentives. This breakthrough is reality at the moment, even though it can provide a solution for only one third of the cultivated forest area. In order to increase the suitability
of bio-ash -based products for the rest of the forest area, its composition and some of
its properties could be enhanced with fine sized slag products. Using recyclable materials from the industry can also mean affordable solutions for the forest owners,
especially if the forest soil also has a need for liming. Also, from an occupational
health and safety point of view it is very important to develop granule/pellet -based
203
products for the safe spreading and usage of fine sized dusting waste and industrial
by-products in many practical applications such as in forest fertilization, soil conditioning as well as road construction material.
The granulated steel slag products studied in this paper were manufactured by mixing and granulating steel industry by-products with bio-ash. In this article, which is is
part of a Finnish research project RAE (Rakeistaminen Avartaa Ekologisuutta / Granulation Expands Ecology), the authors present an idea and details of currently ongoing experiments on how fine sized particle steel slags can be used as environmentally sound construction materials or as forest fertilizers. To increase the usage of
waste material flows as a valuable resource, there is an immediate need for mixing
different materials in certain proportions and afterward use granulation (pelletizing) to
form ash- and slag-containing symbiosis pellets. Also, there is a clear idea how combining bio-ashes and metal industry by-products results economically and technologically viable solutions. These are the true innovations of this study, when test results
(currently only preliminary results have been obtained), not yet finalised, are ready.
One idea presented here is to use granulated symbiotic by-products as forest and
agricultural fertilizer after enrichment through sorption of run-off nutrients (nitrogen
and phosphorus). Another idea is to use pellets as road and building construction
material. An important topic concerning this application is to study some geopolymer
reactions and thus achieve high enough mechanical strength for the product at a
reasonable price. Porosity is another desirable property of the material, so that it
could be used as an insulating material.
Oulu region in Northern Finland has a long history of investigating how ash pellets
together with different additives (but not with slag earlier) can be utilized preceding
their final positioning as a forest fertilizer. In common, the basic idea for different approaches is to be eco- and material efficient, since certain chemical elements (mainly
phosphorus and nitrogen) are transferred from being detrimental (water bodies) to
being beneficial (forests). The latest aim of this research area in our research group
is to optimize and find commercially viable material combinations for the maximally
efficient removal of nutrients from water or wastewater onto the symbiosis granule
materials.
204
Introduction and background
Slag materials
Fine sized slag products are industrially manufactured from slag material flows that
are obtained from metallurgical processes mainly with high calcium level. These material flows are normally obtained from ladle, AOD or other converter slags. For many
manufacturers these are the most difficult material flows to manage, as long as they
should also be usable as by-products after primary use in the melt shop. The formation of fine sized materials is tried to be avoided for example by using stabilization
materials or with additions of different material flows together or even with changes in
handling efforts, for example in cooling methods (or mixing different slags together).
(Kühn & Mudersbach 2004, Durinck et al. 2008)
Figure 1. Ladle slag pouring after cast.
One of the solutions for slag utilization, which have been presented many times, is to
use these materials in other industrial processes. This is particularly feasible for pro205
cesses, which have a need for increased total calcium concentration. For example in
ore based steel plants, most of the blast furnace slag is used externally nowadays
and at least almost half of the BOF slag is recycled to the blast furnace process as a
slag former. The reason for the usage of slag is the unreacted lime it contains, which
makes its recycling economically feasible for manufacturers. In some cases, this kind
of recycling has been found as the most successful slag management method, both
economically and technically. Despite of this and its eco-friendliness, most of the processes in which slag recycling could be used (in theory) often have very strict environmental limits. Thus, this is not a realistic solution for achieving a total 100 percent
usage of slag material flows with high calcium oxide levels. (Kühn & Mudersbach
2004, Durinck et al. 2008)
There is a need, set by the authorities and the manufacturers themselves, to achieve
a 100 percent usage for slag material. Therefore, new innovations are needed in this
field. Even though 100 % utilization has been in aim, after decades of intensive R&D
efforts; comprehensive solutions for several manufacturers are unfortunately not
available.
206
Figure 2. Comparison of contents of slags produced in different industrial units using
a CaO-SiO2-MgO phase diagram.
Granulation/pelletizing in the by-product handling of fine-grain materials
In this paper, the terms granulation and pelletizing are used to point out, that in the
by-product handling sector, there are these two different concepts, which stand for
totally different processes, both of which are widely used to convert fine materials into
materials of larger size with a rounded shape, even though granulation has a totally
different meaning in the metallurgical sector than pelletizing. Granulation/pelletizing
are often mixed with each other, and therefore literary searches concerning slags
should be made using both concepts.
207
Although pelletizing process is a worldwide known technique in the production of artificial aggregates, it has not been used very widely in construction and other sectors.
Using materials directly in the form of fine material and utilizing the liming effect of
materials has been presented in earlier articles. Furthermore, some ideas of increasing the material size for certain usages by pelletizing have also been presented earlier, but higher temperature processes have mostly also been used in those cases.
(Cheeseman 2011)
However, fine sized materials that could be turned to pellet form remain without effective use due to previous lack of novel utilization ideas. If we estimate that earlier on,
easier solutions have already been tested, we should be looking for more difficult
value chains in order to find more value for the materials with multiple properties to
choose from in the future. In this kind of value chain, the material could be used as a
filter material, or specific infra building properties should be preferred to achieve. Also, from an occupational health and safety point of view it is very important to develop granule/pellet -based products for the safe spreading and usage of fine sized
dusting waste and industrial by-products in many practical applications such as in
forest fertilization, soil conditioning as well as road construction material.
If suitable material flows are combined in correct ratio, the agglomerated materials
will achieve preferred properties and a sustainable process efficiency, which could
mean a feasible price level for the new material. A reasonable aim for the granulation
process should be that artificial aggregates should be produced (with adequate engineering performance) by moisture treatment under atmospheric conditions and there
should not be a need for external heating of the material. In this study, the performance of the bio-ash- and slag-containing symbiosis pellets, including the effects of
lime and cement additions, both for geotechnical applications and concrete production purposes will be investigated. These results are needed to be obtained in order
to satisfy the related design requirements. (Baykal & Döven 2000)
The usage of bio-ashes in the form of pellets in forest fertilization
The use of wood-ash as a forest fertilizer has been researched comprehensively in
studies in Northern Finland, and also in other parts of the Scandinavian area. In
208
Northern Finland, pure wood ash from small-scale combustion units is observed to be
available. Bio-ashes are often a combination of wood and peat residuals, because
peat is also an important domestic fuel in Finland. Large-scale industrial peat harvesting for energy production started in the 1970’s in Finland due to the proliferation
of larger district heating systems and was also related to the worldwide oil crisis.
Bio-ash has been reported to enhance forest growth and the earliest studies are up
to 40 years old. Some of the studies have lasted for more than 20 years and the
longest-lasting of them have been started even as early as in the 1940’s. All the studies have similar conclusions, which is that bio-ash contains all the essential nutrients
for plant growth except for nitrogen (N), which is vaporized during the combustion. In
addition, the studies show unambiguously that pure wood-ash can be used as a fertilizer on peat land forests without any significant negative impacts on the environment, but the recycling and utilization of wood-ash in these areas is still insignificant.
Many combustion processes produce mixed ashes, because large investments have
been brought to enable the integrated use of wood and peat in existing peat-fired
power plants. Despite all of the above, all kinds of ash have traditionally been considered as waste (see below). For decades, this has been reality for a material,
which, when used in correct amounts can decrease the acidity of soil and produce a
long-lasting increase in the total nutrient stores of the surface soil. (Huotari 2012)
Figure 3. Consumption distribution between different fuel raw materials for the district
heating producers in Finland in 2012. (Finnish Energy Industries 2013)
209
According to the present European Union and Finnish strategies on waste materials,
all kinds of waste must be utilised primarily as material (reuse, recycling) and secondarily as energy, and if neither of these methods is possible, they can be disposed
of using ecologically beneficial methods. The present goals of the Finnish waste
strategy given in the national waste plan approved by the Finnish Council of State on
the 10th of April 2008 and in the new waste law took effect on the 1st of May 2012
are to: 1) reduce waste, 2) expand waste material recycling and biological reuse, 3)
increase combustion of non-recyclable waste and 4) ensure harmless treatment and
final placement of waste. In addition, material efficiency is nowadays an essential
topic in promoting sustainable use of natural resources, waste materials, and industrial by-products in agreement with the principle of sustainable development and LCA
(life cycle assessment). (Kuokkanen 2013)
A major change in the status of bio-ash has taken place during the last years, as earlier the main treatment method for all kinds of ash was to place them in landfill and
dumping places—all kinds of ash were considered to be unusable waste or even
hazardous waste. Wood and wood-based ashes are bio-ashes which are generated
as by-products of biomass combustion or bio- gasification processes in heat and
power generation. With the change in the EU/Finnish waste strategy, there is currently a tremendous interest in Finland to substantially increase the utilization of bio-ash
and to develop new applications and bio-ash-based products, particularly granulated
bio-ash-based products or symbiosis ash pellets. Today, bio-ash has potential e.g. as
a soil fertilizer, as a liming agent, in soil remediation, in the manufacture of concrete,
and in new applications such as in road construction and in waste water purification
when used in granulated form. Utilizing pellet ash instead of disposing of it in landfills—with increasing landfill costs and a new waste tax of 50 €/ton for all industrial
landfills—will increase substantially the profitability of the whole chain of bioenergy
production.
According to the REACH regulation (1907/2006/EC), all chemical substances of
which at least one tonne per year is produced or imported must be registered. Thus,
if ash materials are placed in landfills and dumps, they need not be registered but
belong under the waste taxation instead. The registration of European wood-based
ashes has been done recently, including altogether 71 large-scale ash producers.
210
This broad consortium indicates the importance and timeliness of the potential utilisation of bio-ashes. (Kuokkanen 2013)
This article will present some of the latest ideas to increase the usage of this kind of
unutilized material flows in Finland. It is proposed, that industrial by-product materials
from traditional metallurgical industrial sector could be utilized as innovative symbiosis pellet products with bio-ash from forest and energy industry, which also remain
without effective utilization within the European Union. (Kuokkanen 2013)
Numerous research efforts have been made in the fields of the economical utilisation
of biomass ash (or bio-ash), slag and sludge, but successful commercial application
has not yet been achieved mainly due to the lack of proper business incentives or
insufficient application potential. However, as the current trends in waste taxation are
increasing the costs of by-products disposal as waste (e.g. landfill), further efforts
should demonstrate utilization possibilities of these materials within the European
Union. For example (biomass) fly ash production is growing rapidly (e.g. currently,
600 000 tons of bio ash is produced in Finland annually) since every country has to
promote the use of energy from renewable sources (directive 2009/28/EC).
Material characteristics require mapping by means of intensive laboratory testing.
Test constructions and fertilizations are also needed. In addition, there is a specific
need for the pelletized materials to have certain characteristics so that they would be
able to compete with natural materials in road construction applications. The two
most important features of bio-ashes in road construction applications are their low
thermal conductivity and enhanced bearing capacity compared to natural soils.
Because of the suggested very strict leaching limits in Finland, the leaching properties of bio-ash and slag were a particular subject of research in this study. In slag
valorisation, when these kinds of symbiotic pelletized materials are used for the applications mentioned above, leaching of potentially harmful compounds is a key issue
to be taken into account. This leaching process is generally characterised to begin as
a surface reaction, which is then followed by a solid-solid diffusion process. As the
basicity difference between pelletized materials and soil becomes remarkable
enough, the material will then dissolve into the soil, with greater reaction speed and
211
rate in the beginning of the dissolution. The rate of the leaching decreases with time
as the diffusion from the bulk of the solid material to the surface becomes slower.
However, the exact mechanisms of leaching remain still unclear. The solubility of individual minerals contained in the bio-ash and slag as well as the distribution of metal
elements in the microstructure of the symbiosis pellets are of the greatest importance
when it comes to being able to fully explain the leaching reaction. (Kuokkanen et al.
2006, Kuokkanen 2013)
Aimofthisstudy
This study is part of a Finnish research project RAE (Rakeistaminen Avartaa
Ekologisuutta / Granulation Expands Ecology). There is a need to research the effects of symbiosis pellet fertilization on plant and forest growth and to monitor and
document its long-term environmental impact, as well as to establish techniques and
methods for cost- and eco-effective spreading of ashes and other fine sized byproduct material flows. In addition to this, there is also a need to communicate with all
the stakeholders on the spreading of ash to forest lands.
The main aims of fertilizing forests with symbiosis pellets consisting of bio-ash and
slag are:
 Returning the nutrients P, K, Ca and Mg to the forests and even increasing the
amount of certain components in forest soil
 The liming properties of the symbiosis pellets are even better than those of
bio-ash (because slag has higher total calcium content than bio-ashes)
 Decreasing the acidification of soils and water streams and simultaneously recovering valuable nutrients from the water
 Increasing the growth rate on organic matter and some minerogenic soils, thus
achieving enhanced carbon capture
 The high pH of the pellets affects the nitrogen balance in the soil and enables
its usage by plants for growth
To perform correct actions, all projects should be analyzed with the NABC method.
Based on this analyze (Table 1), experiments carried out to combine and pelletize
bio-ash and slags as well as application tests were seen as vital in order to achieve
higher usage level for these material flows. The Needs, Approach, Benefits and
Competition were analyzed as follows in this case:
212
Table 1. NABC analysis for the projects to enhance slag utilization with bio-ash in the
form of pellets.
Needs Approach  Larger and mobile units are needed in order to get by‐products into pelletized form in amounts required by the Directive 2006/12/EC  More flexibility to production units and security for the investments done by SMEs on service business  More flexibility to product variety  Increased material usage in the new utilization model  Fertilizers with better properties  Nutrients removal from water and wastewater is needed  Light stone material needed for several purposes
 Application of mobile, flexible (more silos and dosing equipment) and larger production units, with more variation possibilities in the products  Utilizing the results of long‐time studies concerning this research area, conducted by local and international research institutes and universities  Several laboratories are participating in this multidisciplinary study, field testing will also be conducted  Marketing benefits will be shown by LCA analysis  Second generation products with added value will be developed, harnessing the principles of eco‐, cost‐ and material efficiency Competition Benefits  The products are clearly more affordable than competing products  Production unit sizes will be sufficient after these projects, and there will be no further necessary need for enlargement in the future  Logistics costs will decrease  Larger production units mean more R&D professionals working at the SMEs (the experts used are usually from outside the SMEs), and therefore, in the future, SMEs must hire more own R&D people to handle several difficult details in production  A partial solution to the sustainable usage of phosphorus  Currently, there are no large and/or mobile granulation units available  Larger size of production units are necessary from the economical point of view, and intensive research is required to point out the demand for this change  There are similar needs across the whole Europe, and this Finnish project (RAE) will demonstrate the benefits that could also be achieved with even larger production units  Research will make the gap between the prices of by‐product based products and primary raw material based products significant enough – markets will bloom The by-product -based symbiosis products aimed for production, their intended use
as well as their advantages are presented in in a scheme in Figure 4. In addition to
the intended usage applications presented in Figure 4, symbiotic pellets could also
be used in water and wastewater treatment, and the pellets could be further used as
213
fertilizer material containing even higher amounts of nutrients. Furthermore, the term
fertilizer, as depicted in Figure 4, is thought to contain of concept of a soil remediation
agent (liming) as well.
Figure 4. The by-product -based symbiosis products aimed for production, their in-
tended use as well as their advantages.
Experimental
This paper contains some preliminary experiments with aims to combine different byproduct materials in an innovative way to produce granular symbiotic products suitable for forestry and construction. New pelletized products will be produced by combining bio-ash and steel industry slag as main components and certain other industrial waste/by-products as possible additives. The products can be first used as an adsorbent material to reduce nutrient runoff (nitrogen and phosphorus) from different
land uses and diffuse sewer sources, and then as a forest fertilizer and soil conditioning agent. The pelletized products can also be tailored to suit road and building construction needs. There is a need to demonstrate a high-environmental-and-marketpotential process that aims at solving the problem of managing by-product flows in a
sustainable way.
214
Large quantities of industrial by-products still remain without effective utilization. Several research projects have been performed, but successful commercial and industrial activities are still lacking due to the prevailing economic conditions and the absence of proper business incentives. However, the current developments in waste
taxation are providing these incentives with possibilities for a considerable breakthrough. One of the main aims of research should be scaling up the granulation process for by-product treatment. In the future, by-products should be produced by combining at least two by-product streams. With the large scaling up of the granulation
processes, almost 100 % of certain waste flows produced in certain industries could
be treated. If no change in the environmental policies concerning the utilization of
bio-ash and slag -based products can be reached in practice, the current landfilling
practices will continue. This will in turn not lead to – opposite to the target – achieving
emission reductions by the replacement of commercial virgin products manufactured
from scarce primary materials with more environmentally sound alternatives.
Details of the experimental part
In this chapter, the contents of currently on-going experiments in the project RAE are
presented. So far, only preliminary results have been obtained, and thus no final research data and results will be presented in this paper. However, these results have
been very promising and they will be presented later. In Figure 5, untreated bio-ash
and steel slag as well as symbiosis pellets (two different sizes intended for different
applications) containing bio-ash and steel slag used in this study are presented.
Figure 5. From left to right: bio-ash, steel slag and small- and large-sized symbiosis
pellets containing bio-ash and slag used in this study.
215
The following research parts are currently being or will be studied:
1) Small-scale manufacturing tests of the symbiosis pellets
- The effect of the proportional composition (bio-ash/slag) of the symbiosis pellets
- The effect of granulation time
- The effect of certain additives (e.g. waste lime)
- The effect of water addition (volume added, method of addition) and its mixing efficiency
- Required drying time of the manufactured symbiosis pellets before use
- Results of these studies will be utilized and tested in pilot scale granulation
2) Properties of the symbiosis pellets
- Physicochemical properties such as pH, electrical conductivity, dry matter content,
moisture content, total organic carbon (TOC), buffer capacity, liming capacity,
specific surface area, particle size distribution, etc. These analytical methods are
in common use in our research group and are described in detail in earlier publications (Kuokkanen et al. 2006, Kuokkanen 2013).
- Concentrations of easily soluble nutrients (Ca, Mg, Na, K, S, P, Mn, Cu, Zn, etc.)
- Total concentrations of heavy metals (Cd, Cu, Pb, Cr, Zn, As, Ni, Hg, etc.)
- Sequential 5-stage leaching tests (as depicted by Kuokkanen et al. 2006)
3) Application tests of the symbiosis pellets
- As a fertilizer material
- As an adsorbent material for water and wastewater treatment (and further usage
as a fertilizer material)
- As a liming agent in soil remediation
- In road and building construction
- In concrete manufacture
216
Conclusions
Technical and environmental properties of bio-ash as a product could be influenced
by additions of slag. From an occupational health and safety point of view it is crucial
to develop granule/pellet -based products for the safe spreading and usage of fine
sized dusting waste and industrial by-products. Carbon storage potential of forests
can be increased, if by-product materials with low carbon footmark are used to enhance fertilization and thus also increase the growth of the forests. Sustainable production of renewable materials and energy can be increased with the use of byproducts. The symbiotic pelletized materials are seen as an affordable fertilization
material to increase forest growth and an eco- and cost-efficient soil conditioning
agent. Some of the elements (e.g. Mo and B) contained in bio-ash are currently seen
as being detrimental, but could even be seen as vital in the longer run.
Acknowledgements
The Centre of Environment and Energy (CEE) and the Department of Chemistry,
both at the University of Oulu, Finland, are kindly acknowledged for taking part in the
research work. The authors want to thank the RAE-project (EAKR/Tekes) for financial
support and adj. prof. Toivo Kuokkanen for scientific assistance.
References
Baykal G & Döven A G (2000) Utilization of fly ash by pelletization process; theory,
application areas and research results. Resources, Conservation and Recycling,
30, (1), 59–77.
Cheeseman C (2011) Production of sintered lightweight aggregate using waste ash
and other industrial residues – 2nd International Slag Valorisation Symposium, 18–20
April, Leuven, Belgium, 89–101.
217
Durinck D, Engström F, Arnout S, Heulens J, Jones P T & Björkman B (2008) Hot
stage processing of metallurgical slags. Resources, Conservation and Recycling,
52, (10), 1121–1131.
Finnish Energy Industries (2013) District heating year 2012. Finnish Energy Industries.
http://www.slideshare.net/energiateollisuus/energy-year-2012-district-heating,
accessed on 15.9.2013.
Huotari N (2012) Tuhkan käyttö metsälannoitteena (The use of ash as a forest fertiliser). The Finnish Forest Institute (Metla), Oulu. ISBN 978-951-40-2370-5. 48 pp.
Kuokkanen M (2013) Development of an eco- and material-efficient pellet production
chain—a chemical study, Doctoral Thesis (general part and papers IV & V),
Acta
Universitatis Ouluensis. Series A, Scientiae rerum naturalium 607, Oulu, 100 + 108
pp.
http://urn.fi/urn:isbn:9789526201047, accessed on 14.9.2013.
Kuokkanen T, Pöykiö R, Nurmesniemi H & Rämö J (2006) Sequential leaching of
heavy metals and sulphur in a bottom and fly ash from co-combustion of wood and
peat at a municipal district heating plant. Chemical Speciation and Bioavailability, 18,
(4), 131–141.
Kühn M & Mudersbach D (2004) Treatment of liquid EAF –slag from stainless
steelmaking to produce environmental friendly construction materials, SCANMET II –
2nd International Conference on Process Development in Iron and Steelmaking, Luleå 2004, 369–377.
218
Teresa Annunziata Branca1, Chiara Pistocchi1, Valentina Colla1, Giorgio Ragaglini1,
Cristiano Tozzini1 and Lea Romaniello2
Investigation of BOF slag use for potato and tomato cultivation with saline irrigation water in Italy
1
Scuola Superiore di Studi Universitari e di Perfezionamento Sant’Anna,
t.branca{c.pistocchi,v.colla,g.ragaglini,c.tozzini}@sssup.it
2
ILVA S.p.A, ecorifiuti.taranto@rivagroup.com
Abstract
The use of ironmaking and steelmaking slags in agriculture has a long tradition, particularly in Europe. Nevertheless further improvements are required in this field of
application. On this subject, a current project, with a financial grant from the Research Fund for Coal and Steel (RFCS) of the European Community, aims at assessing the effects of ironmaking and steelmaking slags as liming materials on the
quality and yields of some selected crops across Europe. In addition, selected soil
parameters, by taking into account the different soil and climate conditions, will be
evaluated. Along with field trials in Finland, Germany and Austria, a lysimeter trial has
been developed in Italy. Lysimeters are polyethylene boxes, equipped with an automatic drip irrigation system and a plastic tank for the drainage water collection.
The experiment aimed at assessing the effect of BOF (Basic Oxygen Furnace) converter slag, coming from the ILVA steelworks, on the reduction of ESP (Exchangeable Sodium Percentage) levels in saline-sodic soils, due to the high content in CaO
and MgO into this slag, by counteracting the absorption of Na from the soil. In fact in
Italy agricultural soils are rarely acidic, in coastal areas sodic or saline-sodic conditions are frequent. On the other hand, the risk of leaching into the groundwater of
some heavy metals, such as Cr and V, has been evaluated.
The experiment consisted in 18 lysimeters filled with saline sodic soil and irrigated
with saline water. The lysimeter were splitted in two subgroup for tomato and potato
cultivation and treated with different levels of slag: D0, in which no slag were applied,
D1 and D2 consisting in a dose of 3.5 and 7 g kg-1 of soil respectively. BOF slag were
sieved, characterized and then applied in the upper 10 cm of soil. Three additional
219
lysimeters per crop were used for the control test in non saline condition without slag
application.
Chemical analyses of soil and drainage water from the lysimeters have been carried
out on samples collected during the growing season, while the major cations and
trace metals content in the plant tissues was investigated on the biomass sampled at
the harvest.
Results of the first year of the experiment showed an indirect effect of slag application on K adsorption (higher contents in plant tissues) on both crops, probably due to
Ca effect on K mobilization by competition for soil sorption sites (higher K losses on
drainage waters has been observed too) and a lower adsorption of Mg. Moreover it
was pointed out that, despite slag addition provides more exchangeable Ca, it does
not produce any particular effect on sodicity in the short term, while the increase of K
leaching at the higher slag dose was observed, probably due to the competition of
additional Ca on soil K, instead of Na. Finally the higher concentration of V found into
the drainage water of the D2 level, should be considered as a factor of risk and
needs further investigation.
1. Introduction
According to the “zero-waste” concept, wastes and by-products are potential resources rather than issues which need to be addressed. The steel industry is committed to meet this ambitious but difficult goal, through the implementation of innovative
technologies in the industrial practices, in order to achieve the zero-waste vision. This
entails, on one hand, the solution of pollution issues as well as a more sustainable
exploitation of natural resources, and a shift to renewable sources; on the other
hand, from an economic perspective, zero-waste means more competitiveness and
higher efficiency.
In an integrated steelworks, the by-products production accounts for 450-500 kg per
tons of steel produced and slags represent about 80% of the total by-products produced (source: www.worldsteel.org).
Slags are by-products consisting of silicates, alumina silicates, calcium aluminium
silicates, iron oxides and crystalline compounds. Metallurgical slags can be generated either from integrated steel plants or scrap and DRI based steel production and
slags are classified as Blast Furnace (BF) and steelmaking slag.
220
Over the past decades the main destination of waste and by-products were landfills,
because this procedure was considered inexpensive and technically simple. Nevertheless, due to the significant increase of by-product volumes, more stringent legislation and landfill engineering requirements as well as landfill costs, such as landfill tax,
and social acceptance, this is no longer true. The slags dumping in landfills can
cause not only some environmental problems, through the release of harmful substances to the environment and consequently to the water, but also cost increasing,
due to the land occupation and disposal costs. For this reason, is very important to
recover metals from slags and subsequently to recycling them.
Over the last few years, the steel industry has significantly improved the use of byproducts, with the result that their disposal in landfills has been meaningfully reduced.
On the other hand, thanks to the increase of efficiency of production processes, the
slags amount produced during the iron and steelmaking processes has been reduced. For instance, the BF slag generation has decreased from 980 kg/tons of hot
metal to 270 kg/tons of hot metals [1].
Compared to the other by-products, such as dust and sludge, which are usually recycled internally to steelmaking processes, slags are especially used externally and
therefore they are considered marketable by-products. Due to the issues related to
their sustainability and to legislative constraints, this has led to develop and implement innovative solutions from the technical perspective, in order to make slags suitable for use in different field of applications.
The slag recycling is a common practice allowing to use a material with the same or
even better properties of its competitive natural materials. This leads to important
results, such as environmental impact, CO2 emission and cost reductions. Nevertheless it can produce some environmental issues due to the possible release of heavy
metals content.
The use of BFS (Blast Furnace Slag) and steel slags as a phosphate fertiliser dates
back to 1880.BF and steel slags can be recycled in the iron and steel making process
as well as, after recovering of metals, they are applied outside the iron and steel
making process in many areas, such as cement production, road construction, civil
engineering, fertiliser production, landfill daily cover, soil reclamation,etc.[2]. Steel
slags,coming from Basic Oxygen Furnace (BOF) and from Electric Arc Furnace
(EAF) steelmaking processes,represent about 10–15% by weight of the steel output
[3].The chemical composition of the steel slags includes CaO, Fe, SiO2, MgO and
221
MnO.Fe can be separated and recycled in sintering, BF and steel making; on the
other hand, the high content of CaO, MgO and MnO in steel slags can be used to
substitute for a part of limestone,dolomite and manganese ore,by allowing the reduction of iron and steel making costs[4]. Nevertheless the higher content of P2O5 and S
in some steel slags can prevent the direct recycling of the steel slags to the iron and
steel making process, because they negatively affect the steel quality. On the other
hand, the higher phosphorus content makes this slag suitable for use as fertiliser.
BF slag recycling is a consolidated and globally accepted procedure, whereas the
BOF slag use still features some issues, due to the content of free lime, which compromises its volume stability. Nevertheless today the recycling 100% of BOFS (Basic
Oxygen Furnace Slag) and EAFS (Electric Arc Furnace Slag) has been achieved in
some steelworks around the world (source: www.worldsteel.org). In particular, BOF
slag, from the LD (Linz-Donawitz) process, can be utilized in many fields, such as its
partial reuse as an aggregate for civil engineering, due to its good technical properties, the recovery of metal values, etc. [2]. Furthermore some experimental trials have
been carried out in order to assess the use of this slag in acidic soils as soil conditioner and fertiliser. The addition of this slag has led to increase the soil pH and some
changes to the exchange complex, with the result of quality and soil productivity improvement and consequently of possible economic advantages. However, particular
attention is paid to some heavy metal present in it, such as Chromium (Cr) and Vanadium (V) release, that are potentially mobile and toxic to the environment, depending of their speciation [5]. For this reason, one of the most important issues is to
study the environmental behavior of by-products, such as BOFS, in the long term and
the prediction of the contaminants release.
2. Use of metallurgical slag in agriculture
The use of metallurgical slags, such as BF and steel slags, in agriculture has a long
tradition, not only in the past when the Thomasphosphate was used as fertiliser, but
also today. Nowadays the Thomas process was replaced by the BOF process and
consequently steel slags contain 1 to 2 % P2O5. This has led to a significant decrease of the use of slag as fertiliser. Nevertheless the need for liming agents in European agriculture and forestry to be applied to acidic soils has increased the use of
metallurgical slags as liming materials. In fact, over the past few decades, a lot of
222
studies and researches have been carried out around the world, by investigating different applications of these by-products in agriculture [6].
Traditionally BFS is used as a raw material for silicate fertilisers. Furthermore it contains lime, magnesia, Fe and B as well as P and other microelements, which are potentially useful for plants. Both BFS and steel slags used to substitute commercial
chemical fertilisers can lead to reduce some environmentally detrimental processes,
such as mineral extraction and calcinations, that usually are manufactured to make
some natural resources suitable for their use in agriculture.
The use of metallurgical slags as liming materials for amending acid agricultural soils
is a frequent procedure, due to their alkaline properties [7]. This is very interesting as
soil acidity is an important factor limiting crop yield in about 30% of world’s ice-free
land area. Particularly steel slags contain 22 to 38% CaO and 3.5 to 6.5% MgO. These characteristics make such by-products as potential alternatives for lime, by providing not only the pH, Ca and Mg increase, but also exchangeable Al decrease in acidic soils [8]. Calcium, with soil organic matter, allows the formation of a stable structure and of favourable conditions for microbiological processes in soil and increase of
soil fertility. The use of steel slags, that decrease soil acidity, lead to the formation of
insoluble forms that reduce higher uptake of toxic elements in soils [9]. The main factors that control the trace elements in soils are organic C content, pH, CEC and Fe,
Al, Ca, Mg and P concentrations [10].
At European level a recent research work has shown that iron and steel slags can
improve the soil pH and structure as well as the plant yields. On the other hand, after
analyzing results achieved in 40 year experiments, harmful effects on soil and plants
have not been observed. In particular, the increase in Cr and V in the soil are bound
in immobile fractions of soils, therefore they did not affect groundwater [11]. These
achieved results have led to further investigations, particularly linked to the influence
of Cr and V on different parameters, such as soil fertility, respiration and the biological activity. On this subject, an ongoing RFCS (Research Fund for Coal and Steel)
project, RFSR-CT-2011-00037 “Impact of long-term application of BF and steel slags
as liming materials on soil fertility, crop yields and plant health” (“Slagfertiliser”) is
focused on the fertilizing and liming properties of iron and steel slags, such as BFS,
BOFS and LFS (Ladle Furnace Slag) through the improvement of the environmental
and economic aspects. The assessment of these effects on crop yields and quality,
biological and chemical parameters in different soils and climate conditions is the
223
main purpose, as well as the effects of Cr and V on soils and plants and their immobilisation in the soil. In fact, particularly in BOFS, although there are low concentrations
of heavy metals, higher levels of Cr and V can be detected, due to the quality of iron
ores and scraps used [12].
Particularly mineralogical and chemical soil investigations can clarify these bonds
tendency in comparison with Cr and V bonds in soils with high Cr and V contents
naturally, by providing important information on the effects of long-term use of slags
as fertiliser and liming materials in agriculture. Although leaching tests on slags are
usually carried out, in this project some research topics are focused on slags leaching behavior under conditions close to the field.
Concerning the project partnership, while in Germany, Austria and Finland fields trials
are carried out, in Italy lysimeter tests are performed, by fertilizing with the BOFS
coming from Taranto ILVA steelworks. Nevertheless in neutral or alkaline soils present in Mediterranean areas, such as in Italy, the liming effect of slags is less important. But, on the other hand, the evaluation of their potential ability in reducing the
sodicity, which could affect alkaline soil of coastal areas due to sea water infiltration,
is one of the main objectives of Italian trials. It represents an innovative research,
because the interaction with the irrigation water (in this case study it has high salt
content) has not been investigated so far. Only few studies have been carried out on
the effects of steel slags in neutral or alkaline, well drained soils. Nevertheless their
fertiliser effect has been reported [13], as well as a higher risk of trace elements
leaching.
Furthermore the heavy metals leaching in groundwater after BOFS treatment and the
interactions of crop-soil system, after BOFS treatment, and the irrigation water containing salt, are assessed. In fact, salinity and sodicity (measured as Exchangable
Sodium Percentage (ESP)) can negatively affect crop growth, directly because of Na
toxicity for plants and indirectly through the effect on the soil structure. In other
words, the aim of Italian lysimeters trials is to assess the application of different doses of BOFS on crops in saline-sodic soils irrigated with saline water as well as to assess the risk of heavy metals leaching in groundwater after BOFS application to the
soil.
The achieved results will allow not only to improve the ecological and economic aspects of metallurgical slags in agriculture, but also to prove their environmental compatibility, which can be advantageous also for other slag applications, such as build224
ing industry and road construction. Finally this research project will provide an European overview of the effects of metallurgical liming material, after their application in
different local conditions.
The need of homogenised legislation in the European Union is a very important topic,
because currently each European country has its own regulation concerning fertilisers. For this reason new and good results have to be achieved in order to apply slags
for agriculture purposes in the next future in different European contexts.
3. Lysimeter trials in Italy
Trials developed in Italy are focused on the evaluation of influence on soil sodicity of
the BOFS which has been supplied by ILVA Taranto steelworks.
Lysimeter trials have been carried out using 24 lysimeters, consisting in polyethylene
boxes of 1.00 m depth and 1.00 x 1.00 m surface, with an automatic drip irrigation
system and collector at the bottom in order to accumulate drainage water. The soil in
lysimeters is slightly alkaline, with an average pH value of 7.5, and it was treated with
solid NaCl (1.5 g NaCl kg-1 of soil) in order achieve an ESP of almost 35%.
Lysimeters, cultivated with tomato (Lycopersicon esculentum) and potato (Solanum
tuberosum), have been set up according to a completed randomised experimental
design with 3 replicates. In particular 3 doses of BOFS have been supplied: no slag
(D0), low dose (D1) and high dose (D2), corresponding to 0, 3.5 and 7 mg kg-1, respectively. A control with non-saline water and without slag has been also included.
The drainage sampling has been carried out every time the drainage volume exceeded 5% of the field capacity, which corresponds to 2 L. Each sample, after the
total volume measurement, has been subdivided in two subsamples of and divided in
acid-washed PTE bottles in order to analyse major cations (Ca, Mg, Na, K), nitrates
(NO3), chlorides (Cl) and trace metals (Cr, CrVI and V). One of them has been treated with 0.5 ml of HNO3 for V, Cr analyses. Furthermore Electric Conductivity (EC)
and pH were measured at the sampling time with portable devices, and, for each
sampling, a field blank has been prepared as well. All samples have been stored up
at 4°C before analyses.
In the early autumn, after the two crops harvest, 3 samples per lisymeter have been
collected at two different depths, such as 0-10 cm and 10-30 cm and, for each layer,
a composite sample has been obtained through mixing. Afterwards samples have
been dried at air temperature and then maintained at 4°C before analyses. Ex225
changeable cations (Xe), trace metals, anions and available P (Polsen) have been analyzed. A two-way ANOVA has been performed on the water results, using crop (2
levels) and dose (4 levels) as orthogonal factors. On soil, a three-way ANOVA has
been performed, using as orthogonal factors the crop (2 levels), soil layer (two levels)
and dose (4 levels). The Tuckey HSD test has been applied for the post-hoc means
comparisons. All tests have been performed by means of the R statistical software
(version 2.12.0, R Foundation for Statistical Computing, http://www.r-project.org).
3.1 Results of drainage analyses
Table 1 shows results of the drainage analyses. The value of hexavalent Cr (CrVI)
concentration is not shown, because it was below the detection limits (0.5 µg l-1).
Whereas the pH was not affected by the treatments, the Electric Conductivity (EC)
significantly differed between the controls (C) and the treatments with saline irrigation water.EC in drainage waters was also affected by the crop factor. Particularly in
the potato-planted lysimeters significantly lower values have been detected.
The major exchangeable cations (with the exception of K), Cl, and Cr, significantly
differed between potato and tomato crops. In particular lower concentrations of these
elements in the leachate of the lysimeters planted with potato have been found.
Those differences could be determined by a higher cation uptake by potato plants or
the weed community.
Concerning the effect of BOFS application, significant differences were detected for K
and V. Indeed, the K concentration was higher in D1 and D2 treatments, compared
with D0 and C, while V concentrations were higher in D2 than in D0, D1 and C.
As potassium has not been added with the irrigation water or BOFS, an indirect effect
was determined by the slag application, possibly due to the competition of bivalent
cations on the soil exchange sites, thus mobilizing the sorbed K. Conversely, the
higher V losses in the drainage are due to the V content in the highest dose of BOFS.
crop
dose
pH
TOM
C
9.37
EC
Ca
Mg
K
Na
Cr
Ni
V
Cl
µS cm-1 mg l-1 mg l-1 mg l-1 mg l-1 µg l-1 µg l-1 µg l-1 mg l-1
1318
161 15.5 6.2
64.0 1.1
51.0 16.0 104.00
NO3
mg l-1
1.10
D0
8.58
4540
464
62.3
23.7
485.0 1.2
110.7 23.3
1220.00 0.70
D1
8.61
5523
551
94.0
42.0
517.0 1.2
199.7 29.3
1544.00 < 0.1
226
POT
D2
9.28
5485
606
92.0
32.0
491.0 1.2
181.5 29.5
1553.00 0.30
C
9.06
1488
163
35.0
13.8
63.0
0.8
99.0
21.0
104.00
6.70
D0
8.72
3590
366
54.0
16.3
328.0 0.5
98.7
23.0
858.00
11.70
D1
8.67
2933
304
49.0
26.0
218.0 0.5
77.7
20.0
534.00
59.40
D2
7.19
2930
346
49.0
35.0
208.0 0.5
59.0
27.0
574.00
112.00
**
**
*
**
*
**
***
***
**
*
**
*
*
**
**
*
**
crop
dose
P value Interac.
***
Table 15. Mean values of the parameters measured in the drainage waters and
relatives P value from the ANOVA. C = control, D0 = saline irrigation water and slag
dose = 0 mg kg-1, D1 = saline irrigation water and slags dose = 3.5 mg kg-1, D1 =
saline irrigation water and slags dose = 7 mg kg-1. 0 = ***, 0.001 = **, 0.01 = *
3.2 Results of soil analyses
As observed in drainage water, also in the soil CrVI (Table 2) was below the detection limit (0.2 mg kg-1). Higher Nae content was found in the soil treated with the saline irrigation water respect to the control. Cae was higher in D2, due to the slag supply, while Ke was lower, probably due to the mentioned competition effect for the
sorption sites of bivalent cations, such as Ca. Slag did not affect Nae, thus ESP did
not decrease in D1 and D2 with respect to D0, as shown in Table 2.
Exchangeable cations did not show significant differences between the two crops,
with exception of Mge. The latter was lower in lysimeters with potato plants than in
the tomato ones. This could be due to the higher ability of potato in uptaking Mg,
consistently with the results from drainage water samples. The other cations did not
differ, possibly because excess supply by the slag (Ca), saline water (Na) and fertilisers (K). As shown also in the Table 2 some differences between the two soil layers
have been detected. In particular Na leaching was higher in the top 10 cm than below, with final ESP of 2.8% and 10%, respectively. However those values were much
lower to the initial ESP, after salt addiction (35%). Therefore, the added Na has been
gradually eliminated by rainfall in the early autumn. The concentration of Mge and Ke
were higher in the upper layer. K has been added in the surface of soil as fertilizer
(potassium-sulphate) and, for this reason, its residual effect was still present at the
end of the growing season. The enrichment of Mge was less clear, probably due to
227
the different root uptake. Furthermore for Mge the interaction between crop and soil
layer was significant, with a higher difference between upper and lower layer for the
tomato. In addition, the interaction crop x dose was also significant, with higher depletion of Mge at higher dose of slag in the potato.
As far as the major anions are concerned, Cl was lower in the upper layer of soil.
Probably, as for Na, this was due to the leaching from the top layer by the rainfall in
the early autumn. Furthermore, the effect of crop was important, resulting in lower
concentration in potato than in tomato. A different behavior has been detected for
NO3, with higher concentration in the upper soil layer, maybe due to higher mineralisation rate, otherwise to the residual effect of N fertilization with urea, during the
growing season. Results on major anions are consistent with EC results. In fact these
were higher for the saline irrigation treatment (D2, D1, D0) than the control (C). Consistently with the Na and Cl pattern, lower EC has been detected in the upper layer.
Trace metals showed different behaviors. V was significantly higher in the D2 lysimeters, while Cr did not increased with the high slag dose. Conversely the crop factor
was significant for Cr, while it was not for V. In fact the Cr content in the soil was lower in the potato-planted lysimeters. This result is consistent with the observation in
the drainage water (lower content of Cr in the drainage of potato). We can hypothesize that either this crop or the weed community had a higher Cr uptake.
profile crop dose
EC
µS/cm
L
TOM C
POT
U
Ca
Mg
K
Na
Cr
V
Cl
NO3
ESP
mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg
133.9
43.7
918.0
52.0 166.0 26.0
50.0
16.7
11.6
19.7
2.0
D0
155.6
42.7
828.0
46.0 182.7 184.7 54.0
19.0
13.1
29.5
13.4
D1
158.0
42.0
954.3
42.3 149.7 258.0 48.0
18.0
10.1
23.0
17.0
D2
169.5
42.0 1020.0 55.0 151.0 225.0 65.0
20.0
10.2
27.3
14.6
C
100.6
44.7
929.5
52.7 180.7 23.0
47.0
15.7
5.1
23.3
1.6
D0
136.7
42.0
816.0
45.3 145.0 193.3 50.3
17.0
6.5
34.7
14.5
D1
116.5
43.3
987.7
45.7 181.7 140.3 51.0
18.3
10.1
18.7
9.4
D2
119.8
43.7 1211.0 45.0 118.0 173.7 48.0
17.7
5.2
26.7
16.6
114.0
42.3
899.0
60.0 326.3 19.7
51.0
17.0
3.7
31.3
1.3
D0
112.1
46.0
876.3
63.7 292.3 48.7
54.0
16.5
4.8
36.3
3.0
D1
111.4
41.3 1046.7 55.7 280.3 59.7
60.7
18.3
2.7
23.3
3.0
D2
124.2
42.0 1051.0 62.7 280.0 60.0
67.7
21.0
8.9
33.3
3.6
C
86.1
43.7
965.7
56.3 293.3 19.7
56.7
16.7
3.7
29.3
1.3
D0
117.1
43.0
796.0
54.3 339.3 44.0
54.7
18.7
2.7
43.7
3.2
TOM C
POT
Polsen
228
D1
119.7
45.7 1037.7 48.0 288.3 49.0
54.7
18.7
7.4
18.7
3.7
D2
114.7
40.0 1071.3 48.7 243.7 55.3
44.3
18.0
6.1
27.0
4.1
**
***
dose
crop
profile
dose : crop
dose: profile
crop : profile
dose : crop : profile
*
**
***
**
*
**
***
*
***
**
*
***
***
**
***
*
**
*
***
*
Table 16. Mean values of the parameters measured in the soils and relatives P value
from the ANOVA. C = control, D0 = saline irrigation water and slag dose = 0 mg kg-1,
D1 = saline irrigation water and slags dose = 3.5 mg kg-1, D1 = saline irrigation water
and slags dose = 7 mg kg-1. U = upper soil layer, L = lower soil layer. 0 = ***, 0.001 =
**, 0.01 = *
3.3 Results of plants analyses
Figure 1. Ca content as the average of different plant organs in tomato (T) and potato
(P), letters indicate significant differences
Both crops showed significant differences in Ca content among dose treatments
(Figure 1). In particular at the higher slag dose (D2) for potato and at both doses (D1
and D2) for tomato corresponded the higher Ca concentrations. This effect was clearly visible in stems and roots of tomato and only in stems for potato, while berries and
tubers were not affected (data not shown).
229
6
b
a
ab
3
-1
b
mg g
4
a
4
5
a
3
mg g
-1
5
6
7
P
7
T
ab
2
1
0
0
1
2
b
C
D0
D1
D2
C
D0
D2
D1
Figure 2. Mg content as the average of different plant organs in tomato (T) and potato
(P), letters indicate significant differences
Conversely both crops showed lower concentration of Mg in the lysimeters treated
with slag (D2 and D1 for tomato and D1 for potato). (Figure 2). This is also consistent
with what observed on drainage waters of the same treatments (higher Mg losses)
and seems to indicate a competition effect of Ca, which is uptaken preferentially by
plants when in excess with respect to Mg. As for Ca, the effect of the slag treatments
on Mg contents was visible in stems and roots of tomato, and in stems of potato (data
not shown).
K concentrations varied among the different plant organs, and significant differences
where found among the treatments. In particular in the tomato berries, K contents
were significantly higher in D2 than in D1 and C, while the lowest value was found in
D0 (Figure 3). This could be explained with the already cited mobilization of K by
competition with Ca added with slags on soil sorption sites, thus increasing its availability for plants. This effect is clearly not present in D0, where no Ca was added. The
same effect of K concentration increase with slag dose is visible mainly in the potato
tubers, and partially (higher K content in D2) in the aboveground biomass.
230
Figure 3. K content in different plant organs in tomato (T) and potato (P), letters indicate significant differences
Figure 4. V content in different plant organs in tomato (T) and potato (P), letters indicate significant differences
Vanadium concentrations were higher in tomato roots and in potato aboveground
biomass (stem) for D2 lysimeters. Tomato did not transfer the absorbed V to the
stems and berries, while potato seemed to accumulate it in the above ground biomass. Nevertheless data on roots are not available. A higher V content was found
also in the control of potato, the reason is unclear.
No significant differences on Cr concentrations were found among dose treatments of
the two crops (data not shown). We hypothesize that the already mentioned differences in soil Cr content (lower for potato lysimeters) are due either to weeds uptake,
since the mean weeds dry matter production was significantly higher for potato crop
(160 g and 50 g of dry matter for tomato and potato lysimeters respectively) or to Cr
accumulation in potato roots (but not in tubers).
4 Conclusions
Results of lysimeter trials carried out in Italy, inside the “Slagfertiliser” project, have
shown that slag addition did not affect the exchangeable sodium percentage as hypothesized at the beginning, while improved the exchangeable Ca content in soil.
Instead potassium was affected indirectly by slag addition (competition of Ca for
sorption sites), resulting in its higher availability both for leaching and for plant uptake
(higher content in tomato berries and potato tubers). An indirect effect (competition of
231
Ca for root uptake) seemed to affect also Mg concentration, with lower concentration
of this element in plants treated with slags.
Strong differences were highlighted in the response of the two crops to trace metals
added with BOFS. In particular tomato seemed to accumulate vanadium in roots, with
significantly higher concentrations on plants grown with slags, while potato seemed
to transfer vanadium also in the aboveground biomass (higher concentration corresponding to the higher slag dose).
Finally vanadium was found in higher concentration in drainage water of the lysimeters with the higher BOFS dose (7 mg kg-1), highlighting a possible risk of leaching to
groundwater for this metal.
References
[1] Joulazadeh, M.H. and F. Joulazadeh, SLAG; VALUE ADDED STEEL INDUSTRY BYPRODUCTS. ARCHIVES OF METALLURGY AND MATERIALS.
55(4).
[2] Geiseler, J., Use of steelworks slag in Europe. Waste management, 1996.
16(1): p. 59-63
[3] Proctor, D.M., et al., Physical and chemical characteristics of blast furnace,
basic oxygen furnace, and electric arc furnace steel industry slags. Environmental science & technology, 2000. 34(8): p. 1576-1582.
[4] Motz, H. and J. Geiseler, Products of steel slags an opportunity to save natural
resources. Waste Management, 2001. 21(3): p. 285-293.
[5] Proctor, D.M., et al., Physical and chemical characteristics of blast furnace,
basic oxygen furnace, and electric arc furnace steel industry slags. Environmental science & technology, 2000. 34(8): p. 1576-1582.
[6] T.A. Branca and V. Colla: 'Possible Uses of Steelmaking Slag in Agriculture:
An Overview', in 'Material Recycling - Trends and Perspectives', (ed. InTech),
2012, Dimitris S. Achilias (Ed.), ISBN: 978-953-51-0327-1, InTech, DOI:
10.5772/31804.
[7] Lopez, F.A., Balcazar, N., Formoso, A., Pinto, M., and Rodriguez, M., The recycling of Linz-Donawitz (LD) converter slag by use as a liming agent on pasture land. Waste management & research, 1995. 13(6): p. 555-568.
[8] Das, B., Prakash, S., Reddy, P. S. R., and Misra, V. N., An overview of utilization of slag and sludge from steel industries. Resources, Conservation and
Recycling, 2007. 50(1): p. 40-57.
[9] Kabata-Pendias, A. and H. Pendias, Trace elements in soils and plants. 2001:
CRC PressI Llc.
[10]
Chen, M., L.Q. Ma, and W.G. Harris, Baseline concentrations of 15
trace elements in Florida surface soils. Journal of Environmental Quality,
1999. 28(4): p. 1173-1181.
232
[11]
Kühn, M., Spiegel, H., Lopez, F. A., Rex, M., Erdmann, R. (2006). "Sustainable agriculture using blast furnace and steel slags as liming agents"
EUR(22033): 1-152.
[12]
Rex, M. Environmental aspects of the use of iron and steel slags as agricultural lime, in Proceedings of the 3rd European slag conference€”manufacturing and processing of iron and steel slags, Keyworth, 2002.
Euroslag publication No. 2, pp. 137-150.
[13]
Wang X., Cai Q.S.: Steel slag as an iron fertilizer for corn growth and
soil improvement in a pot experiment, Pedosphere (2006) 16 pp. 519-524.
Acknowledgments
The work described in the present paper has been developed within the project entitled "Impact of long-term application of blast furnace and steel slags as liming materials on soil fertility, crop yields and plant health" (Contract No. RFSR-CT-201100037) that has received funding from the Research Fund for Coal and Steel of the
European Union. The sole responsibility of the issues treated in the present paper
lies with the authors; the Commission is not responsible for any use that may be
made of the information contained therein.
233
P. Drissen1), F. Chazarenc2), M. Fixaris3), M. Rex4), H. Rustige5), St. Troesch6),
Removal of Phosphorus from Wastewater by Steel Slag Filter Systems
1) FEhS ‐ Institut für Baustoff‐Forschung e.V., Duisburg, Germany, 2) Ecole des Mines, Nantes, France, 3) ArcelorMittal Shared Services BSME, Esch‐sur‐Alzette, Luxembourg, 4) Arbeitsge‐
meinschaft Hüttenkalk e.V., Duisburg, Germany, 5) AKUT Umweltschutz, Berlin, Germany, 6) Sarl Epur Nature, Caumont sur Durance, France Abstract The discharge from wastewater treatment plants serving communities with more than
10.000 inhabitants can be limited to 2 mg P / l or below according to the Water
Framework Directive 2000/60/EEC. In sensitive areas this limiting value might be applied even for smaller communities. Small scale wastewater treatment plants usually
run single step treatments like constructed wetlands or aeration tanks and often are
struggling with these limitations.
An option to further decrease phosphorus concentration in the outlet is a subsequent
wastewater treatment in fixed bed reactors. Common filter media in fixed bed reactors are sand, gravel or natural stones. Up to now industrial aggregates have been
tested exclusively on lab-scale. These tests revealed a high phosphorus removal capacity of metallurgical slag qualities. Despite positive results the technique has not
been tested in field- scale system under real conditions, at least in Europe.
Within the European project SLASORB, funded by the Research Fund for Coal and
Steel, BOF- and EAF-slag were tested as filter media for phosphorus removal on labscale and in field-scale systems. The technical feasibility of slag filter systems has
been established, showing good phosphorus removal performance even over long
term. Further investigations aimed at the valorisation of the slag with retained phosphorus in agriculture and economical aspects of slag filter systems.
234
Introduction Phosphorus is of essential importance for all biologic organisms. Its availability is a
limiting factor in plant growth, making phosphorus an important constituent of many
fertilisers. However, an excessive input of phosphorus in water bodies like rivers and
lakes causes eutrophication and an excessive growth of algae, especially in slow
flowing waters. As a consequence the treatment of phosphorus in wastewater is
mandatory in Europe [19, 20]. The discharge from wastewater treatment plants serving
communities with more than 10.000 inhabitants can be limited to 2 mg P / l or below [20]. In sensitive areas this limiting value might be applied even for smaller communities. Usually large scale wastewater treatment plants cope with statutory provisions by a highly sophisticated combination of chemical and biological treatments to
remove phosphorus. Small scale wastewater treatment plants usually run single step
treatments like constructed wetlands or aeration tanks and often are struggling with
these limitations due to the need of additional staff, technical equipment, energy,
chemicals or increasing volumetric flow rates.
A simple option to further decrease the level of phosphorus in wastewater is a subsequent treatment of the wastewater in solid bed reactors. It is common to use sand,
gravel or natural aggregates as filter media in such solid bed reactors. Aggregates
made of industrial by-products have been tested only on lab-scale and mostly with
synthetic wastewater [21, 22, 23]. Especially metallurgical slag qualities with high concentrations of CaO and FeOx showed a good removal performance for phosphorus in
19
20
21
22
23
235
these tests [24]. Despite this the use of metallurgical slag as filter media in solid bed
reactors has not been transferred into operational practice, at least in Europe.
Within the currently finished European research project "SLASORB" (Using Slag as
Sorbent to Remove Phosphorus from Wastewater) the use of BOF- and EAF-slag as
filter media to remove phosphorus from wastewater has been tested in long-term field
trials for the first time in Europe [25]. Other aspects of the project include the examination of phosphorus loaded slag as potential (P-) fertiliser and the feasibility of the
process in terms of efficiency, recommended amounts and lifespan as well as estimations on the economy.
Operational work
According to the state of the art slag qualities with high concentrations of CaO and
FeOx are most suitable for the elimination of phosphorus in wastewater. For this reason available data of BOF- and EAF-slag qualities from carbon steelmaking all over
Europe have been evaluated. Appropriate qualities were selected and their phosphorus removal capacity has been tested by simple batch trials along with the measurement of chemical behaviour of slag in wastewater conditions [26]. Effects of slag composition, contact time, pH and initial phosphorus concentration of wastewater were
analysed with respect to phosphorus sorption capacities. Based on these results representative BOF- and EAF-slag qualities from several production sites in Europe
were selected and used for lab-scale and field-scale investigations. The selected slag
qualities are customarily used as aggregates in road making and comply with relevant national environmental regulations.
Lab-scale filter systems
As a first step lab-scale investigations were performed to determine the long-term
performance of the filter media to remove phosphorus from wastewater and to gather
24
25
26
236
information on hydraulic properties of slag based filter systems. A schematic drawing
of the lab-scale filter systems is shown in Figure 1.
Figure 1:
Schematic lay-out of the lab-scale filter systems
Grain sizes used in the test were 20/50 and 6/16 mm for EAF-slag and of grain size
20/40 and 5/16 mm for BOF-slag. In addition combinations of EAF- and BOF-slag
and combinations with additional sand layers as separate filters were used. All lab
tests were performed with synthetic wastewater having around 11 mg P / l and an
average hydraulic retention time of one day at constant temperature of 20°C. Labscale filter systems have been in operation for 65 weeks with monitoring of phosphorous concentration in the water inlet and outlet.
Information on hydraulic properties like flow mechanisms, reaction volume, average
reaction time between wastewater and slag (hydraulic retention time = HRT), disposition to clogging by adsorption or precipitation and pH have been acquired prior to
design and lay-out of field- scale filter systems.
Field-scale filter systems
Field-scale filter systems with BOF- and EAF-slag were constructed and operated at
existing wastewater treatment plants of two small communities in France and Germany. The French slag filter system was connected to a two stage constructed wetland (CW) in La Motte d’Aigues, close to Avignon; the Germany system was connected to an aeration tank, operated as sequential batch reactor (SBR) at Kappe,
north of Berlin. Further details are listed in Table 1.
237
Parameter
Unit
French-system
German-system
number of reactors
[-]
2
2x3=6
reactor square
[m²]
12
0.44
reactor height
[m]
0.5
0.7
reactor volume
[m³]
6
0.31
BOF-slag size
[mm]
20-40
8-32
bulk density
[kg/m³]
1400
1883
pore volume
[%]
~ 50
48
EAF-slag size
[mm]
20-40
5-10
bulk density
[kg/m³]
1800
1652
pore volume
[%]
~ 50
52
inlet water
[-]
CW outlet
SBR outlet
mean HRT
[d]
2
1
loading operation
[-]
24 batches / day
1 batch/ day
loading rate
[l/h]
60 L / min
variable
loading operation
[-]
free flow
pumping
flow direction
[-]
horizontal
up- / downstream
Table 1:
Overview layout and operational parameters of the French and German
field-scale filter systems
The French slag filter system consists of containers, filled with BOF- and EAF-slag.
The filters were instrumented with multi parameters probes and online meters to determine values of pH, redox potential, O2, conductivity and flow as well as automatic
samplers. During operation a small percentage of the outlet water of the wastewater
treatment plant was used as input for the slag filter systems. A schematic drawing is
shown in Figure 2.
238
Figure 2:
Schematic lay-out of the French field-scale filter system
Outlet water of the slag filter systems was circulated back to the first stage of the
constructed wetland to avoid accidental increase of pH in the in outlet water.
The German slag filter system was an arrangement of 2 x 3 solid bed reactors, connected by a sophisticated system of pipes and valves that allowed different filling operations. A schematic drawing is shown in Figure 3.
Figure 3:
Schematic lay-out of the German field-scale filter system
In contrast to the French system flow direction was vertical and smaller grain sizes of
8 to 32 mm for BOF-slag and 5 to 15 mm for EAF-slag have been used. Further instrumentation was similar to the French system. The focus was set on efficient opera239
tion by minimizing dead volume to improve removal of phosphorus. For the outlet pH
adjustment below 8.5 two bubble-reactors had been installed, operating with CO2.
Valorisation of P-saturated slag as fertiliser
Slag samples from the lab-scale and field-scale slag filter systems were taken after
several months of operation and analysed with respect to their effect on soil properties and phytotoxic effects on plants. The impact of these slag samples on plant
yields and the plant availability of phosphate were investigated in vegetation trials.
Results
Absorption mechanism
Phosphorus removal capacity of slag qualities has been tested by batch trials. Maximum removal capacities ranged from 0.6 to 2.0 mg P/g slag corresponding to contact
times from 1 to 7 days. In these trials dissolved, precipitated or adsorbed phosphorus
can be analysed. In general decreasing concentrations of dissolved phosphorus in
the wastewater were correlated with increased Ca2+ ions concentration and pH. Besides precipitation of calcium-phosphate complexes absorption and/or crystallisation
on the surface of the slag aggregates was observed. Results showed different removal capacities for BOF- and EAF-slag due to divergent mechanisms. The main Premoval mechanism of EAF-slag appeared to be adsorption. The P-removal by BOFslag seemed to be a combination of adsorption and precipitation phenomena. According to investigations by X-ray diffraction and scanning electron microscopy observed crystals are assumed to be hydroxyl-apatite Ca5(PO4)3(OH).
This suggests that the main mechanism of phosphorus removal from wastewater by
steelmaking slag is related to the release of Ca2+ ions by slag and the related increase in pH. The high efficiency in phosphorus removal becomes obvious by the
simplified reaction (1)
5 Ca2+ + 3 PO43- + OH- → Ca5(PO4)3(OH)
(1)
240
taking into account the formation of Ca2+ and OH- ions by steelmaking slag in aqueous solution.
Lab-scale filter systems
All lab-scale filter systems have been operated and sampled for at least 65 weeks. A
failure of the filter systems in terms of reduced flow rates due to clogging was not
observed for all slag qualities and grain sizes in test. Tracer experiments proved a
stable water flow with respect to time and space, at least a few weeks after start of
operation.
All slag qualities and combinations with natural sand revealed a high phosphorus removal capacity. The average in-let concentration of 11 mg P / l of the synthetic
wastewater was decreased to less than 0.5 mg P / l in the out-let throughout most of
time within 65 weeks of observation. The removal efficiency was almost higher than
90 % (see Table 2). Small grain sizes showed higher efficiency compared to larger
grain sizes.
241
BOF
EAF
BOF
EAF
BOF +
EAF +
big
big
small
small
sand
sand
g/kg slag
0.84
0.66
0.79
0.62
0.99
0.77
mass P retained g/kg slag
0.83
0.59
0.78
0.55
0.98
0.76
99
89
99
89
100
98
mass P in-let
efficiency
Table 2:
%
Removal performance of phosphorus in lab-scale filter systems
Results of the lab-scale filter trials confirmed that the adjustment of wastewater qualities with less than 2 mg P / l by steelmaking slag is a realistic goal.
Field-scale filter systems
The potential efficiency of steelmaking slag qualities in phosphorus removal from
wastewater has been confirmed by the field-scale filter-systems operated in France
and Germany. The different lay-out of the French and German field-scale filter systems, different environmental surroundings and variation of operational modes gave
additional information for future design and operation of such systems.
The French system was operated over 15 months since the launch in September
2010. Starting with good removal performance the efficiency went down within the
first 10 weeks of operation. After 10 weeks the hydraulic retention time was set to 2
days, as general value for the remaining time of operation. With increasing hydraulic
retention time removal performance rose again, but decreased in the winter months.
In spring of the following year removal performance was better again. Changes in
removal performance clearly followed seasonal changes in temperature as could be
observed throughout the entire period of observation. Presumably this is related to
temperature dependency of pH and reduced reaction kinetics.
Within the last months of operation a better performance of BOF-slag compared to
EAF-slag was observed. This must be seen in view of the slightly higher pH of 8.5
induced by the BOF-slag, due to its higher concentration of CaO and free lime, compared to 8.0 of the EAF-slag. Another reason might be that the water in this area of
242
France has a high buffer capacity and EAF-slag might not be able to increase pH in a
sufficient way.
Removal performance and pH were influenced as well by the hydraulic retention
time. The hydraulic retention time was changed occasionally for some tracer experiments and accidentally due to dysfunction of the instruments. Hydraulic retention
times of less than one day usually were correlated with lower pH and reduced removal performance and reverse with hydraulic retention times of 3 days.
A save adjustment of 2 mg P / l in the outlet of the French slag filter systems has
been achieved only in the first weeks of operation when comparably high pH values
were achieved. Nevertheless, the average inlet concentration of 8.3 mg P / l was seriously decreased throughout the entire period of observation.
Figure 5 shows the amount of removed phosphorus as function of the total input of
phosphorus; both expressed in terms of kg slag.
Figure 5
Removed versus added phosphorus of French field-scale filter systems
over 15 months of operation
The diagram indicates that the removal efficiency of BOF-slag is 59 % on average
and that of EAF-slag is just 36 %. Especially the lower efficiency of EAF-slag after
approximately 22 weeks of operation must be seen in context with the lower pH as
discussed above. The slop of data points for BOF- and EAF-slag shows that both
243
filter systems have been still working and a saturation level has not been achieved at
the end of operation. Both systems could have been run for a longer period of time
than just 15 months.
Similar results have been achieved with the German field-scale filter systems. As the
sequential batch reactor of the community did not work properly the different test
campaigns had to be run with highly varying concentrations of inlet phosphorus. Outlet concentrations used to follow inlet concentrations and it turned out that phosphorus removal demands some time. Shortening of the hydraulic reduction time below 1
day shows lower removal performance. By using EAF-slag as filter media outlet concentrations of phosphorus below 2 mg / l have been achieved. In the case of BOFslag minimum concentrations of 4 mg per litre were analysed. This result seems to be
in contrast to results achieved with the French system. The simple explanation is the
smaller gain size of aggregates used in the German filter system.
Best removal performance was achieved with reactors operated in down-stream
mode. The slope of the P-retention curve shows, similar to Figure 5, revealed a constant high removal efficiency which reaches 91 % at a load of 800 g P/ m³ of EAFslag. The overall removal rate of BOF-slag was up to 66 % at a comparable load.
Valorisation of P-saturated slag as fertiliser
The use of slag coming from wastewater treatment after a distinct period of operation
is reasonable because of the enrichment in phosphorus. Various slag samples from
lab-scale and field-scale filter systems were chosen and analysed on their mineral
acid soluble and citric acid soluble phosphate content to select qualities for further
testing in vegetation tests on their phosphate plant availability. In parallel vegetation
tests without additions and with addition of reference phosphate fertilisers, like soft
ground rock phosphate, Thomas phosphate and Triple superphosphate (TSP) have
been done. Vegetation tests were repeated with two levels of phosphorus addition on
two different types of soils.
The phosphate fertilising effect of the finely ground slag samples was mostly similar
to the fertilising effect of Thomas phosphate and Triple superphosphate. The soluble
244
and mostly plant available phosphorus fraction in the soil was significantly increased
as shown by the results of different extraction methods (see Figure 7).
Figure 7:
Water soluble soil phosphorus in dependence of applied phosphorus
level and the product type, 18 days after start of trial
The yield of the crops rape and rye were increased significantly by all slag qualities.
An additional yield increase appeared on the higher phosphorus fertilisation level.
These increases were accompanied with increases of the plants phosphorus contents. Consequently the phosphorus uptake of both experimental cultures was increased significantly. It was partly even higher than the phosphorus uptake of plants
with Thomas phosphate and Triple superphosphate fertilisation and significantly
higher than phosphorus uptake of plants fertilised with soft ground rock phosphate.
Besides some yield decreasing influence of high pH values there was no negative
impact of slag application on plant growth, yields and nutrient uptake. A phytotoxic
effect of the slag samples from wastewater treated on plant development was not
detectable, even in high application rates.
The vegetation pot experiments showed that BOF- and EAF-slag used for phosphorus removal in wastewater treatment can be successfully used as liming materials
and phosphate fertilisers. Although the phosphorus contents are comparatively low
245
the phosphate fertilising effects are equivalent to phosphate fertilisers of high efficiency like Triple superphosphate or Thomas phosphate and much higher than the
effects of soft ground rock phosphate.
Prospects
BOF- and EAF-slag proved to be highly valuable filter media for phosphorus removal
in wastewater treatment. Operational results with lab-scale filter system confirmed
excellent removal performance on long terms with save adjustment of out-let concentrations below 2 mg P /l.
Some deviation in the overall removal performance of field-scale experiments compared to lab-scale experiments are due to the chosen layout and specific environmental conditions. An optimised hydraulic retention time of approximately 2 days with
slightly increased pH is beneficial with respect to low outlet concentrations of phosphorus. Further optimisation can be adjusted by the specific design of slag based
filter systems, like the adjustment of length to width aiming at a longer percolation
distance and the selection of suitable mono grain sizes, aiming at a higher specific
surface ratio without reducing the reaction volume.
The available results from field-scale experiments provided sufficient information to
further optimize these systems and to safely adjust outlet concentrations of less than
2 mg P / l in professional practice. With respect to average water consumption and
emission of phosphorus by inhabitants in Europe slag based filter systems should
have a reaction volume of 1.5 to 2.0 m³ slag aggregates, corresponding to roughly
2.5 to 3.5 tons of slag per inhabitant. The expected lifespan of such slag based filter
systems is in the range of 6 to 7 years until saturations with phosphorus results in
bad removal performance. The use of saturated slag from wastewater treatment as
liming agent and phosphate fertiliser bears promising aspect. Cost for transportation
and processing to suitable grain sizes will influence the feasibility.
Aiming at lowest phosphorus concentrations in the discharge of wastewater treatment plants a chemical treatment is customary. Comparing investment and maintenance costs slag based filter systems could be a competitive option at least for small
communities running wastewater treatment plants for less than 400 inhabitants.
246
References
1
Directive 2000/60/EG of the European Parliament and the Council of 23 Octo-
ber 200 establishing a framwork for Community in the field of water policy
2
Council Directive 91/271/EEC of 21 May 1991 concerning urban waste-water
treatment
3
Chazarenc, F., Brisson, J., Comeau, Y.: Slag columns for upgrading phospho-
rus removal from constructed wetland effluents
Water Science and Technology 56(3), 109-115, 2007
4
Drizo, A., Forget, C., Chapuis, R. P., Comeau, Y.: Phosphorus removal by
electric arc furnace steel slag and serpentinite
Water Research 40(8), 1547-1554, 2006
5
Bowden, L. I., Jarvis, A. P., Younger, P. L., Johnson, K. L.: Phosphorus re-
moval from wastewaters using basic oxygen steel slag
Environmental Science and Technology 43(7), 2476-2481, 2009
6
Chazarenc, F., Kacem, M., Gerente, C., Andres, Y.: Active filters - a mini re-
view on the use of industrial by-products for upgrading phosphorous removal from
treated wetland
Proceedings of the 11th international conference on wetland systems for water pollution control, Indore, India, 1. 7. Nov. 2008
7
Using
Slag
as
Sorbent
to
Remove
Phosphorus
from
Wastewater
Research Fund for Coal and Steel, contract no. RFCS-CT-2009-00028, draft final
report, submitted in March 2013
8
Barca, C.: Steel slag filters to upgrade phosphorus removal in small
wastewater treatment plants
PhD Thesis, Ecole des Mines, Nantes, submitted 17th October 2012
247
Theme 4
2. From Research to Applications
248
A. Ehrenberg
Does stored granulated blast furnace slag lose its reactivity?
FEhS - Institut für Baustoff-Forschung e.V., Bliersheimer Str. 62, D-47229 Duisburg
Abstract
The production of granulated blast furnace slag (GBS) and its use as a cement constituent is subject to cyclical fluctuations. This can result in the need to place the slag
in intermediate storage for fairly long periods. The question as to whether this is associated with a loss of reactivity has been discussed heatedly in the past. The FEhSInstitute therefore carried out investigations in which two granulated blast furnace
slags were stored for 4 and 5½ years respectively under both in dried and moist
state. The investigations showed that even after years the largest percentage of slag
grains by volume were present unchanged in a glassy state. This also means that
they are not easier to grind. Comparison of the cementitious properties at the start
and end of storage showed that there is no reduction in reactivity if the GBS is correctly processed, particularly with respect to a comparable particle size distribution.
However, the water demand of slag-rich cements may be slightly increased. Tests in
technical scale confirm the lab-scale results. A detailed publication in "Cement International" is given in [27].
Introduction
GBS,whichinGermanyisusedalmostexclusivelyasamaincementconstituent,is
heavilydependentontheseasonalandeconomictrendsintheconstructionindus‐
try.ThisrepeatedlygivesrisetothesituationwherefreshlyproducedGBScannot
beprocessedimmediatelywiththeresultthatlargequantitieshavetobeplaced
inintermediatestorage,sometimesformonthsorevenyears.GBSisaby‐product
ofhotmetalproductioninblastfurnaces,soproductionoftheslagisdependent
ontheutilizationoftheblastfurnaces.Thisisparticularlythecaseifthevastma‐
jorityoftheliquidblastfurnaceslagbecomesgranulated(Fig.1).Providedthata
blastfurnaceisnottemporarilyshutdownintimesoflowerhotmetalproduction
causedbyeconomictrendsitisinfactpossibletoincreasethespecificquantityof
27
[ ]
249
slagproducedpertonneof
hotmetalfromtheaverage
figureof270kg/tHMto,for
example,350kg/tHM.Howev‐
er,eventhisexpensivemeas‐
ureisnotsufficienttokeep
thetotalGBSproductionata
constantlevel.Duringthese
phasesitisparticularlyim‐
portantforthecementpro‐
ducerstobeabletomakeuse
ofstockpiledmaterial.Time
andagainthereistheques‐
Fig. 1: Hot metal and GBS production and cetionaboutwhetherstored
ment dispatch in Germany 2008-2012
GBShasthesameproperties
asfreshlyproducedslag,especiallywithrespecttoitsreactivity.Ithasinfactal‐
readybeenshowninsomeveryearlypublicationsonthissubjectthathigh‐per‐
formancecementcanalsobeproducedwithfairlyoldGBS[28,29,30].However,
therearealsoarticlesthatareapprehensiveaboutlimitations[31,32].
Investigative program
The samples of two different granulated blast furnace slags A and B, which were delivered to the FEhS-Institute with residual moisture contents of 7.9 M.-% and 6.9 M.%, were each split. One half of each sample was dried for 24 h at 105 °C and then
characterized in detail. The majority of these dried sample halves and of the moist
sample halves were then stored in closed airtight containers for 4 (GBS B) or 5½
(GBS A) years. The samples were not mechanically compacted. All the samples
were characterized again after 4 or 5½ years respectively. To verify the results of the
laboratory investigations 3000 tonnes GBS of Dillinger Hütte were stored in a separate outdoor stockpile (GBS D). The original slag and samples taken after
3 / 9 / 15 / 21 months storing time were characterized in detail, too.
[28]
[29]
[30]
31
[ ]
)
[32]
250
Results of the investigations
Fresh granulated blast furnace slag
The top row in Fig. 2 shows the freshly delivered granulated blast furnace slags that
had already been dried. The typical glassy lustre of the grains is clearly visible. The
chemical compositions and other characteristic values of the slags are listed in Table
17. After removal of the water the freshly produced granulated blast furnace slags
generally have contents of chemically bound H2O and CO2 in the range of 0.1 M.-%
to 0.3 M.-%. The higher content, compared with slag A, of chemically combined H2O
of 0.56 % in slag B was not confirmed during the analysis of the sub-sample after it
had been stored when dry for four years.
GBS A
as delivered
GBS B
as delivered
1 mm
dried, 5½ years storage
dried, 4 years storage
1 mm
residual moisture, 4 years storage
residual moisture, 5½ years storage
Fig. 2: Granulated blast furnace slags A and B under a reflected light microscope
251
The cementitious slag properties were determined in accordance with the guidelines of the granulated blast furnace slag database of the FEhS-Institute [33] using
blast furnace cements with a slag/clinker ratio of 75/25 and a sulfate content of
4.5 M.-%. The Portland cement clinker (CL) came from a single batch that was
stored unground under dry conditions. The characteristic granulometric values of
the ground granulated blast furnace slags and clinker as well as the cement properties are listed in Table 18. This could be an indication that the 24-hour drying at
105 °C carried out before the chemical analysis was insufficient for this comparatively porous GBS. It was also pointed out in [34] that under some circumstances
physically combined water remains in the sample during drying at 105 °C and
would then be recorded as "chemically combined" during subsequent analysis.
[33]
34
[ ]
252
Na2O-Equivalent 1
GBS A
GBS B
0.33
0.79
CO2
2
0.26
0.09
H2O
2
0.01
0.56
C+M+S3
85.5
C/S 3
(C+M)/S
3
F value acc. to Keil
4
Glass content 5
84.5
1.20
1.15
1.40
1.48
1.62
1.74
98.3
6
99.7
M.-%
1
2
Vol.-%
4
2.93
2.93
Apparent density 7
2.69
2.50
Bulk density 8
1.20
1.06
Total porosity 9
8.2
14.6
Vol.-%
Residual moisture
7.9
6.9
M.-%
True density
3
g/cm³
5
6
7
8
9
Na2O + 0.658 K2O
under N2
C = CaO, S = SiO2, M = MgO
(CaO+0.5 S2 +0.5 MgO+Al2O3) /
(SiO2+MnO)
light-microscopy
gas pycnometry
DIN EN 1097-6
DIN EN 459-2
calculated
Table 17: Specific values for granulated blast furnace slags A and B (as delivered)
GBS A
as
delivered
5½ years
GBS B
as
delivered
4 years
residual
residual
dried
moisture
moisture
4270
4290
5250
4160
4300
4910
cm²/g
Blaine (GBS)
15
15
17
17
16
17
µm
d' (GBS)
0.92
0.96
0.98
1.02
0.98
1.01
n (GBS)
0.95
0.87
2.12
0.99
0.96
1.60
m²/g
BET (GBS)
4220
4270
4250
4140
4270
4250
cm²/g
Blaine (CL)
14
15
14
16
15
14
µm
d' (CL)
0.91
1.00
0.89
1.06
1.00
0.89
n (CL)
24.5
26.0
27.5
26.5
26.5
28.5
M.-%
Water demand
5:104:254:304:104:205:30h:min
Setting time
6:15
5:35
6:25
5:20
5:25
7:00
234
226
202
223
223
212
mm
Mortar spread
14.5
11.5
13.4
16.2
17.2
15.9
RC (2 d)
34.9
30.4
30.2
34.5
34.7
33.1
RC (7 d)
MPa
49.2
45.0
42.7
48.9
48.7
45.3
RC (28 d)
57.0
54.1
52.3
55.2
56.6
53.5
RC (91 d)
Table 18: Specific values for blast furnace cements containing 75 M.-% GBS
GBS storage
-
dried
253
Stored granulated blast furnace slags
Consolidation behaviour
It is well known that granulated blast furnace slags stored in the open consolidate in
the course of time [28, 29]. The use of heavy clearing equipment and a crusher is
often necessary when these slags are picked up in order to transport them to the
grinding plant. It is stated in [28] that "the slag becomes lumpy, sticks together and
hardens like stone, often to such an extent that it can no longer be removed without
blasting". The extent to which consolidation occurs depends on various parameters
that are described in [35,
36
]. The granulated blast furnaces slags A and B that were
stored when moist consolidated significantly during the storing time, but the
agglomerates were still relatively easy
to break up. This is attributable to the
only slight compaction (unlike the conditions found in practice), the lack of
imposed load and the exclusion of any
further input of moisture (Fig. 3). The
residual moisture content dropped only
slightly during storage from 7.9 M.-%
to 6.4 M.-% (GBS A) and from 6.9 M.-
Fig. 3: Consolidated granulated blast
furnace slags A and B containing moisture, after storage in the
laboratory for 5½ and 4 years
% to 6.6 M.-% (GBS B).
Change in loss on ignition
For GBS that has been stored for a long time the content of chemically combined
H2O and CO2 is an indication of the duration of the storage [34,
37
]. According to [34]
calcium hydroxides and alkali hydroxides form during storage in the open and are
gradually transformed into carbonates. The way that the levels of chemically combined H2O and CO2 in slags A and B have changed is summarized in Table 19. As
expected, the granulated blast furnace slags stored with residual moisture contain
higher levels than those stored when dry. This is an indication of the reaction at the
[35]
36
[ ]
[37]
254
outer surface, at the surfaces of the pores and in the pore water. Earlier investigations with granulated blast furnace slags that had been stored for decades in the
open and then ground had shown that the loss on ignition had increased, especially
in the finer fractions [38]. From this it can be deduced that when the slag is ground the
readily grindable early hydration products accumulate in the fines. This result was
confirmed in the current investigations through the increase in BET surface area of
the slags that were stored when moist and then ground (Table 18).
Storage
As delivered
GBS A
H2O
CO2
H2O
CO2
0.01
0.26
0.56
0.09
after 5½ years
Dried
GBS B
0.12
0.16
after 4 years
0.20
0.11
M.-%
1.43
0.32
0.81
0.18
Residual moisture
Table 19: Levels of chemically combined H2O und CO2 in slags A and B
The BET surface area is determined to only a small extent by the granulometry of a
fine material and more by the structure of its surface and particles. In the ground
slags that had been produced from freshly delivered slag and from slag that had
been stored when dry it was 0.95 m²/g (GBS A) or 0.99 m²/g (GBS B) and 0.87 m²/g
(GBS A) or 0.96 m²/g (GBS B) respectively but it was significantly increased (GBS A:
2.12 m²/g, GBS B: 1.60 m²/g) in slags that had been ground to comparable particle
size distributions but had previously been stored when moist. This increase can only
be explained by the early hydration products when compared with the unreacted
GBS glass. The formation, and possibly also the structure, of these products produced by early hydration are not sufficient quantitatively (relative to the mass of the
overall sample) for appreciable crystalline phases to be detectable by X-ray diffraction analysis.
Physical changes
The surfaces of the granulated blast furnace slags that have been stored when moist
are conspicuously corroded. The difference in appearance under a reflected light microscope between the freshly delivered slag and the slags that were stored when dry
or with residual moisture can be seen in Fig. 2. In particular, the typical glassy lustre
is lacking in the samples stored with residual moisture. The differences between the
38
[ ]
255
fresh slag and the slags stored when moist or dry are even clearer under the scanning electron microscope (SEM). Only a few hydration products can be seen on the
largely smooth surfaces of the fresh delivered samples and the samples that had
been stored when dry (Fig. 4a).
GBS A
Surface as delivered, CO2+H2O = 0.27 M.-%
GBS B
Surface as delivered, CO2+H2O = 0.65 M.-%
Dried, 5½ years storage, CO2+H2O = 0.28 M.-% Dried, 4 years storage, CO2+H2O = 0.31 M.-%
Fig. 4a: Granulated blast furnace slags A and B under a SEM
256
The isolated reaction products that are visible are typical for wet-granulated granulated blast furnace slags [39]. In contrast, a thoroughly corroded surface and a large
number of reaction products, such as ettringite or calcite, can be seen on the samples that were stored when moist (Fig. 4b).
GBS A
GBS B
Res. moisture, 5½ years, CO2+H2O = 1.75 M.- Res. moisture, 4 years, CO2+H2O = 0.99 M.-%
%
Fig. 4b: Granulated blast furnace slags A and B under a SEM
The visual appearance of the corroded granulated blast furnace slag surfaces would
initially suggest that the glass content, which is important for the reactivity of the slag,
would be significantly reduced. However, the results of the optical microscope determination of the glass content show that no significant reduction of the glass content
can be detected (Table 18). GBS stored when moist has lower true density. In a similar way to the CO2 and H2O contents, the lower densities are found in the finer fractions of the ground slags. The change in density, even though still moderate after 5½
39
[ ]
257
or 4 years, of granulated blastfurnace slags A and B, is also an indication of the generation of early hydration products on the particle surfaces.
During grinding the finely divided, rough and inert weathering products accumulate in
the fine fraction of the ground slag, which can be observed in their lower true densities. According to [34] the reaction products should be mainly calcite (density: 2.6 to
2.8 g/cm³; GBS: 2.93 g/cm³). However, in addition to calcite there are also reaction
products present that contain water, as is shown by the content of chemically combined H2O and the scanning electron photomicrographs.
Grindability
It has sometimes been suggested that older GBS is easier to grind. During the investigations discussed here the grindability was determined by the Zeisel method on the
original material, not on a certain fraction. The specific grinding energy expended in
kWh/t is normally shown in relation to the Blaine specific surface in cm²/g that is generated after each grinding stage. The results show that in the fineness range from
about 3000 to 4500 cm²/g that is relevant for cements the grinding of the GBS that
was stored when moist apparently requires significantly less energy than is the case
with the fresh slags. However, it has already been shown in [40] that characterization
of the fineness by Blaine value leads to misinterpretations if the GBS has an increased loss on ignition. The effect is more strongly marked the higher the loss on
ignition (the age) of the stored slag. Apparently, the above-mentioned accumulation
of early hydration products in the very fine fraction combined with the changed structure compared with the unreacted slag, has the effect that higher Blaine values are
produced rapidly during the grinding, whether in an industrial mill or in the Zeisel
grindability tester. However, the associated particle size distributions (PSD) show that
the Blaine values simulate a falsely high fineness. For this reason not only the Blaine
value but also the PSD was determined after each grinding stage for the slags that
had been stored when moist and tested in the Zeisel tester.
Comparison of the PSD determined during the Zeisel test on GBS that had been
stored when moist with those obtained with the Zeisel test for the freshly delivered
slags confirmed that the Blaine values simulate a falsely high fineness and therefore
easier grindability (Fig. 5). The granulometric values from the grindability investigations are listed in Table 20. From the results it can be seen that it cannot be assumed
40
[ ]
258
that GBS that has been stored for a long time is easier to grind than the fresh slag.
This means that the output of an industrial grinding plant that is fed with slag that was
stored when moist should not be controlled by the Blaine value but by the PSD that is
needed for quality reasons. If this is not taken into account then the mill output will in
fact be increased but will contain a coarser, and therefore less reactive, ground slag.
Fig. 5: Particle size distributions of GBS A during the Zeisel test
GBS A
GBS B
Storage
Processing
Blaine
d'
n
Blaine
d'
cm²/g
µm
cm²/g
µm
5010
12
0.96
5040
13
As delivered 33 min. Zeisel test
after 5½ years
after 4 years
32 min. Zeisel test
4920
16
0.90
5010
14
27 min. Zeisel test
4620
19
0.89
4710
14
22 min. Zeisel test
4200
33
0.70
4370
16
Residual
moisture
17 min. Zeisel test
3760
27
0.84
3990
19
12 min. Zeisel test
3410
30
0.90
3400
24
7 min. Zeisel test
2730
29
0.94
2640
32
Table 20: Granulometric values for the grindability tests
n
0.94
0.90
0.92
0.91
0.93
0.95
0.93
Cement properties
In order to provide reliable information about whether or not the GBS stored when
moist lead to altered cement properties their particle size distributions were adapted
to the distribution that had been obtained at the start of the investigation with the
freshly supplied slags. However, the particle size distributions of the slags that had
259
been stored when moist were displaced slightly towards the coarser range. This
should be borne in mind when assessing the cement properties. The results of the
cement investigations, determined on CEM III/B blast furnace cements containing
75 M.-% GBS, are listed in Table 18. It can be seen that there is no serious drop in
strength when the granulated blast furnace slags stored when moist are used as cement constituents. The sometimes somewhat lower 28-day strengths of slag A can
be attributed partly to the fact that the PSD was not quite optimally adjusted. On the
other hand, although setting up comparable particle size distributions can be an important process step it is not an entirely adequate. If, as described above, there is an
accumulation of inert early hydration products in the very fine fraction then there is a
lack of reactive GBS in this section of the particle size range that is particularly important with respect to reactivity. This means that the PSD of the older GBS would,
depending on the degree of early hydration, have to be displaced further into the finer
range compared with the PSD of fresh slag. The water demand of cements made
with slag that was stored when moist was somewhat increased and the mortar flow
table spread was somewhat lower for the same w/c ratio. This can also be attributed
to an accumulation of inert early hydration products in the very fine fraction. As a
whole, however, all the values for the water demand lay at a normal to low level,
which is mainly attributable to the wide PSD that is advantageous for industrial cements.
Tests in technical scale at Dillinger Hütte
To verify the results of the lab-scale tests and to
evaluate the possible quality change during an interim storage under practical circumstances (open
storage, high load, compaction, weathering etc.)
3000 tonnes moist GBS of Dillinger Hüttenwerke
(GBS D) were stored. Different layers were placed
and compacted to simulate conditions given in the Fig. 6: Compacting the 3000 t
GBS storage
center of a storage of several hundred thousand
tonnes of GBS (Fig. 6). The results confirm completely the lab-scale tests described
above. With respect to the strength development of blast furnace cements
(GBS/CL = 75/25) it could be shown that the use of GBS which was stored for 21
months does not result in lower reactivity. Precondition is the adjustion of a compara260
ble PSD (Table 21). The tests in technical scale did not show an increase of the water demand.
fresh
3 months 9 months 15 months 21 months
8.20
6.30
5.30
19.90
22.40
Residual moisture
99.8
99.0
99.6
99.5
99.8
Glass content
0.40
0.64
0.59
2.15
2.46
CO2 + H2O
4120
4480
4600
5460
5190
Blaine
18
17
17
18
18
d'
8.8
9.8
8.4
11.3
13.1
RC (2 d)
28.9
29.9
31.1
31.9
31.7
RC (7 d)
42.3
42.7
41.3
45.8
46.0
RC (28 d)
50.6
50.6
50.8
54.1
52.5
RC (91 d)
Table 21: Specific values for GBS D and blast furnace cements
containing 75 M.-% GBS taken from the storage core
M.-%
Vol.-%
M.-%
cm²/g
µm
MPa
Conclusions
Seasonal and economic fluctuations means that it is often necessary to store GBS in
the open for fairly long periods because it cannot be processed promptly into cement
after it has been produced and de-watered. This inevitably gives rise to the question
about whether stored GBS has the same properties as freshly produced slag, especially with respect to reactivity. Comparative investigations carried out on two granulated blast furnace slags that were stored for 4 or 5½ years respectively after drying
or with residual moisture showed that the storage does not cause any essential loss
of slag reactivity and that high-performance cements can also be produced with "old"
GBS without having to increase its fineness – characterized by the particle size distribution – to a significant extent. However, it was not possible to confirm the statement that "the hardening capacity of stockpiled slag is substantially improved with
increasing age – especially with slag that originally had a low reactivity" [34]. On the
other hand, there was confirmation of the experience that older granulated blast furnace slags, which can be identified by their increased content of chemically combined H2O and CO2, exhibit Blaine values after grinding that, measured by the particle size distribution, are too high. The reason is the early corrosion of the grain surfaces that has occurred and therefore falsely simulate an apparently easier grindability in the grindability test or an adequately high fineness during the grinding. Only if
this is not taken into account in practice then granulated blast furnace slags that have
been stored for a long time could unjustifiably be designated as less effective. The
261
tests in lab-scale were completely confirmed by long-time tests in industrial scale with
GBS being stored for 21 months at Dillinger Hütte.
References
[1]
Ehrenberg, A.: Does stored granulated blastfurnace slag lose ist reactivity?
Cement International 10 (2012) No. 4, p. 64-79
[ 2]
Rün, R.: Haldenschlacken als Zumahlgut bei der Hüttenzementherstellung,
Zement 33 (1944) No. 4, p. 79-81
[3]
Frigione, G., Sersale, R.: Blastfurnace cement mortars manufactured with
fresh granulated and weathered slags, Cement and Concrete Research 24
(1994) No. 3, p. 483-487
[4]
Rostock, M.: Hüttensand vom Hochofen ins Zementsilo - ein Beispiel, ZKG
International 57 (2004) No. 6, p. 68-77
[5]
Schiller, B.: Mahlbarkeit der Hauptbestandteile des Zements und ihr Einfluß
auf den Energieaufwand beim Mahlen und die Zementeigenschaften, Schriftenreihe der Zementindustrie No. 54 (1992)
[6]
Battagin, A. F., Pecchio, M.: Blast furnace slag weathering study, Proceedings
of the 11th International Congress on the Chemistry of Cements, 11.16.05.2003, Durban, p. 905-913
[7]
Ehrenberg, A.: Überblick über die "Hüttensand-Kartei" der FEhS, Report des
Forschungsinstituts 4 (1997) No. 2, p. 6-7
[8]
Mußgnug, G.: Die Verwertung von gekörnten Haldenschlacken bei der Herstellung hydraulischer Bindemittel, Stahl und Eisen 69 (1949) No. 9, p. 301306
[9]
Numata, S. et al.: On the agglomeration of granulated slag sand and its storage stability tests, Transactions of the Japan Concrete Institute 3 (1981), p.
47-54
[10]
Lang, E.: Einfluß einer Verfestigung von Hüttensand auf seine Eigenschaften,
Cement International 5 (2007) No. 3, p. 84-94
[11]
Kollo, H.: Prüfverfahren zur Beurteilung des Frischezustands von Hüttensand,
Abschlussbericht zum AiF-Forschungsvorhaben 8224 (1991)
[12]
Ehrenberg, A., Israel, D., Kühn, A., Ludwig, H.-M., Tigges, V., Wassing, W.:
Granulated blast furnace slag: reaction potential and production of optimized
cements, Cement International 6 (2008) No. 2, p. 90-96, No. 3, p. 82-92
[13]
Schäfer; H.-U.: Does the method of grinding affect the hydration of ground
blastfurnace slag? Cement International 1 (2003) No. 5, p. 84-93
262
[14]
Ehrenberg, A.: Hüttensand - Ein aktueller Beitrag zur nachhaltigen Zementherstellung, Proceedings 17. Internationale Baustofftagung ibausil, Weimar,
23.-26.09.2009, S. 1-0097/1-0102
263
John J. Yzenas Jr.
Agricultural Utilization of Iron and Steel Slag in the USA
Director of Technical Services, Edw. C. Levy Company – Valparaiso, IN
Abstract
Iron and Steel slag has been utilized for agricultural applications in the United
States for many years. Much of its early utilization was as a liming agent, with some
emphasis on “Basic Slag” which also contained a level of available phosphorous. As
the steel industry changed and phosphorous contents in the slag were reduced the
use of slag declined. During recent years with the introduction of silicon, and its declaration as a beneficial substance for plants, the interest in iron and steel slag has
again increased.
Over the past years many environmental agencies have adopted Land Application Guidelines for the utilization of industrial by-products in agricultural applications. These guidelines typically include maximum metals concentrations for the byproducts. In many cases iron and steel slag applications fall under the solid waste
division of these agencies, and become governed by the guidelines. In 2009 a study
was initiated with the United States Department of Agriculture - Agricultural Research
Service office in Toledo, Ohio and the Edw. C. Levy Company. The study’s focus was
on the viability of various Iron and Steel slags as a source of silicon and to determine
if metals up-take was a potential issue. The study’s focus was the determination of
metals and silicon content of the slags and the level of up-take by various species of
plants.
This presentation will provide a history of slag’s utilization in agricultural applications in the United States and an overview of the metals research completed by the
USDA-ARS and the Edw. C. Levy Company.
264
1- History
Early steel slags, produced via the Bessemer process in open hearth furnaces, contained a level of available phosphorous that could be marketed as a fertilizer.
These materials came to be known as Basic Slag. The American Association of
Plant Food Control Officials defined this “Basic Slag” as containing at least 12% total
phosphoric acid (P2O5) or required it to be labeled "low phosphate". As the industry
converted to the Basic Oxygen Steelmaking process, the level of available phosphorus reduced to 1-3% making it unattractive as a phosphate source. The new opportunity became the BOF slags alkalinity which made it suitable as a liming agent.
Several universities, such as Michigan State University, studied these materials and
found them to be competitive, if not superior to locally available aglimes. In the October 1965 edition of Crops and Science, Boyd Ellis stated “Comparisons of both
BOF slag and open hearth slag with agricultural limestone were also made in greenhouse experiments with barley and alsike clover grown on acid soils. In all cases, the
BOF slag produced yield increases as good as agricultural limestone… “.
The State of Alabama had one of the earlier agricultural slag specifications
(1950’s). Figure 1 provides an overview of both Alabama’s specifications and the early history of the changes in slag chemistry. The concept of slag as an agricultural
liming agent was readily accepted and a large majority of the states began to include
slag into their liming specifications. Most slag in agricultural use today contains very
little phosphorus and is used primarily as a soil liming material. It typically has a CCE
(Calcium Carbonate Equivalent) between 50% and 70% and may contain some micronutrients. In the 80’s and 90’s, apparently due to logistical issues and the cost of
production, the agricultural slag market began to diminish.
265
Basic Slag
Analysis
Ground
Limestone
1957
1964
1975
1997
2003
(Minimum Quality)
Neut. Value (%CCE)
78
68
55
60
85
90+ (**)
Phosphorous (% P2O5)
10.9
7.4
2.1
0.3
0.7
--
Iron
--
17.9
--
24.4
26.2
--
23.6
--
Calcium
22.7
Magnesium
--
2.8
--
4.9
6.5
6+ in Dolomite
Manganese
--
1.8
--
2.6
1.2
--
Zinc
--
<0.1
--
0.1
0.1
--
Boron
--
<0.1
--
--
0.06
--
% Passing #60 Mesh
--
--
--
--
35
50+
% Passing #100 Mesh
80
70
80
50
--
--
Figure 1: Alabama Liming Specifications
2 – Land Application Permits
During the last portion of the twentieth century, as agricultural slag moved towards the background, the environmental movement began to take hold. This resulted in the establishment of criteria as to what could be placed in or on the air, land
and water. The agricultural arena was no exception. One of the early guidelines
published was 40 CFR Part 503 “The Standards for the Use or Disposal of Sewage
Sludge”. In this document the U.S. Environmental Protection Agency (EPA), utilized a
risk assessment to establish detailed requirements for the treatment and land application of biosolids, originally adopted in 1993. In 1995 the EPA removed Chromium
form Part 503 due to questions raised about supporting data. In 2005 Basta, Ryan
and Chaney published “Heavy Metal and Trace Element Chemistry in ResidualTreated Soil: a Review of Impacts on Metal Bioavailability and Sustainable Land Application”41 which further reviewed the affect of these requirements. Table 1 is a current summary of the Part 503 limits and includes the original 1993 limits for comparison.
266
Table 1: U.S.EPA Part 503 Heavy Metals Limits
While this work was targeted at biosolids many states also use the Part 503
biosolids rule as a guide for land application of other by-products, including slags. In
the case of slag, where a processor is required to obtain land application approval,
they must demonstrate their ability to consistently meet the Part 503 Maximum Concentration limits, prior to registering their product as either a liming agent or fertilizer.
This involves the submittal of historical chemical and leaching data. The required
data is made up of metals testing utilizing USEPA 846, Method 3051 and leaching
results from the EPA’s Toxicity Characteristic Leachate Procedure (TCLP). Many of
the slags produced at our integrated iron and steel mills in the USA do not have issues with the metals or leaching limits, although some EAF slags have had occasional exceedance. Table 2 provides typical values seen at our sites.
267
Typical Metals (mg/kg)
Tested via SW3051 Microwave Digestion
and SW846 6010B ICP Analysis
Arsenic (As)
Cadmium (Cd)
Chromium (Cr)
Copper (Cu)
Lead (Pb)
Molybdenum (Mo)
Nickel (Ni)
Selenium (Se)
Zinc (Zn)
Pool of Samples
BOF Slag
0.423
5.645
2372.43
100.92
46.287
13.01
32.92
<0.046
923.09
BF Slag
0.039
0.031
198.07
19.04
<0.034
1.73
7.49
<0.046
40.11
EAF Slag
5.614
5.798
3830.02
262.89
32.885
55.64
93.33
<0.046
576.00
LMF Slag
6.685
0.849
865.97
70.59
<0.034
11.29
30.13
3.099
48.43
150
140
150
87
Typical Leachable Metals (mg/kg)
Tested via SW846 3015 Microwave Digestion Arsenic (As)
Barium (Ba)
Cadmium (Cd)
Chromium (Cr)
Copper (Cu)
Lead (Pb)
Selenium (Se)
Silver (Ag)
Zinc (Zn)
BOF Slag
<0.095
0.510
<0.070
<0.060
<0.200
<0.110
<0.115
<0.070
0.240
BF Slag
<0.095
0.539
<0.070
<0.060
<0.200
<0.110
<0.115
<0.070
0.402
EAF Slag
<0.095
0.779
<0.070
<0.060
<0.200
<0.110
<0.115
<0.070
0.126
LMF Slag
<0.095
1.071
<0.070
<0.060
<0.200
<0.110
<0.115
<0.070
0.186
Table 2: Typical Metals and Leaching Results
3- Product Registration
When marketing a slag liming agent the rules for product registration vary greatly.
Some states require obtaining land application approval, while others just require registration. In several states both are required.
AAPFCO's SUIP #25
Heavy Metal Rule
Metal
ppm per 1% P2O5
Limit ppm/1%
Micronutrient
EPA Part 503
2011 YTD (ppm)
LSU (ppm)
Arsenic
13
112
75
1.1
<4
Cadmium
10
83
85
0.299
<0.2
Cobalt
136
2228
NA
4.1
4.4
Lead
61
463
840
11.8
<1.2
Mercury
1
6
NA
0.048
NR
Molybdenum
42
300
75
5.4
2.2
Nickel
250
1900
420
5.7
4.7
Selenium
26
180
100
3.3
<14
Zinc
420
2900
7500
87.5
99.7
268
Table 3: AAPFCO SUIP #25 "The Heavy Metal Rule
Fertilizers require registration in all states. The American Association of Plant
Food Control Officials (AAPFCO) is an organization of fertilizer control officials from
each state in the United States, Canada and Puerto Rico who are actively engaged
in the administration of fertilizer laws and regulations. This group developed AAPFCO's Statement of Uniform Interpretation and Policy (SUIP) #25 "The Heavy Metal
Rule". While the EPA guidelines focused on the land, SUIP #25 addressed metals
as one of the constituents of fertilizer. The metals limits in SUIP #25 are calculated
based upon the micronutrient contents of fertilizers. Table 3 illustrates the micronutrient limits at 1% versus EPA Part 503 limits and test data from one of our sites.
Many states have adopted these values when evaluating new fertilizer applications
that utilize by-products. In many states slag processors are required to meet the EPA
land application and then meet SUIP #25.
4- Field Trials
In 2010 we began a series of field trials in conjunction with the USDA to evaluate metals up-take and the efficacy of slag by a variety of plants in both a row crop
and horticultural setting. The original work at the USDA included varieties of zinnias,
while the field trials included corn, soy beans and sugar beets. A testing protocol
was established that utilized two controls (untreated and limed) as well as slag treated sections. The treatments were typically in the range of two tons/acre. Soil and
plant samples (root, stem and leaves) were taken at various points during the growth
stages to evaluate both nutrient up-take and metals content. The samples were prepared as per USEPA 846 and then analyzed by ICP-OES. The results obtained during the trials are summarized in Table 4 and Appendix 1, with all below regulatory
limits. The work was summarized by Dr. Jonathan Frantz in a presentation at the
2012 ASHA Conferece: “Based on these experiments, there is no/low risk of heavy
metal leaching from these slag types, even if applied at rates much higher than 2
tons/acre.”42
42
269
Date Received
10/05/10
10/05/10
10/05/10
10/05/10
10/13/10
10/13/10
11/04/10
11/04/10
Site
Bay Port
Bay Port
Bay Port
Bay Port
Bay Port
Bay Port
Bay Port
Bay Port
Soil
Soil
Soil
Soil
Harvest
Harvest
Harvest
Harvest
Treated WP3
Soil
Untreated
WP4 Soil
Treated WP2
Soil
Untreated
WP1 Soil
593-4
592-3
594-5
596-4
649-1
649-2
636-1
ND
5.7
5.4
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.0463
ND
ND
ND
16
19
14
9.3
0.216
0.109
0.157
0.117
14
16
9.6
5.8
11.6
11.3
1.59
1.06
10
11
9.8
9.2
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
13
16
11
7.2
0.9
0.780
ND
ND
ND
ND
ND
ND
0.104
ND
0.112
0.128
51
54
50
32
73.7
64.0
62.6
47.0
Material Type:
Sample ID:
Log #:
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Molybdenum
Nickel
Selenium
Zinc
Soy Treated Soy Untreated Corn Treated
Corn
Untreated
636-2
Table 4: Metals Analysis from Field Trials (mg/kg)
5- Liming and Calcium-Silicate Fertilizer
While the use of iron and steel slag as a liming agent has been common practice for a long time, the realization of its benefits as a silicon fertilizer is relatively new.
The materials have been known to be a source of micronutrients, but the added
component of silicon, as contributed by slag, has only recently been studied. Two of
the pioneers in 20th century research were Jian Feng Ma (Okayama University, Japan) and Emanuel Epstein (University of California at Davis). They helped arouse
the interest in silicon and set the basis for much of the current efforts. Their work,
along with that of Gaspar Korndorfer, “Extractors for Estimating Plant Available Silicon from Potential Silicon Fertilizer Sources43, and Lawrence Datnoff, “Silicon Products: At the Border Between Plant Nutrition and Plant Protection”44 has now brought
slag into the arena as a calcium-silicate fertilizer.
Much of the early fertilizer work was limited to a couple of slag types, but current work by Brenda Tubana at Louisiana State University and Jonathan Frantz at the
U.S. Department of Agriculture-ARS has expanded the potential for fertilizer utilization into a wide variety of slag types.
One of the key benefits of the calcium-silicate fertilizers is an ability to enhance plant health. A stronger cell structure can help the plants to overcome stresses from drought, insects and some diseases. Photo 1 shows the effect of a caster
43
44
270
slag fertilizer on corn root balls during a recent drought in Midwestern USA. These
benefits can also result in an increased yield even during normal growing years. After
a review of the literature and positive field responses AAPFCO has officially recognized silicon as a “beneficial substance”.
Photo 1: Corn Root Balls
6- Testing and Research
Research to further document the benefits of silicon and better define testing
protocols is continuing in the USA. While there is general agreement as to the benefits of calcium-silicate slag, the appropriate method of determining the level of available silicon is still under discussion for the slag products due to potential interferences
during colorimetric procedures. The following is a brief summary of some of the work
currently under way:
Louisiana State University (Tubana 201245):
Biomass and Silicon Uptake of Wheat in Response to Different Levels of
Plant-Available Silicon: The benefit of CaSiO3 slag application for crops requiring large Si supply can be offset by applying this type of Si source at rates large
enough to drastically change soil pH hence solubility of several plant essential nutrients.
Changes in pH and Mehlich-3 Extractable Nutrients of Selected Soils from
the Midwest and South USA; As Influenced by Different Rates of Iron Calcium
Silicate Slag: Results show that both liming potential and composition of CaSiO3
slag had significant effect on the amount of M3-extractable essential nutrients.
Estimation of Plant Available Silicon Using Different Extraction Procedures
for Selected Soils from the Midwest and South USA: The initial findings of this
study concur with previous studies which documented that among the solutions,
45
271
soil Si extracted using 0.5 M acetic acid can provide the best estimate of plantavailable Si.
Effect of Different Silicon Sources on Acetic Acid-Extractable Silicon Content of Two Alluvial Soils of Louisiana: Among these slag materials, only CS
influenced ryegrass biomass production on both soils linearly implying that CS
has the highest potential as Si source for crop production.
Agronomic and Environmental Impacts of Silicon Fertilizer Application On
Rice Grown in Louisiana Soils: Environmentally, CaSiO3 slag reduced methane
emission by 17-22% over that of the control in both soil types. The reduction in
methane emission can be attributed to the release of active iron oxide from the
CaSiO3 slag which is a potential source of electron acceptor eventually resulting in
decreased methane emission.
Utah State University (Bugbee, 201346):
Studies on the Beneficial Effects of Silicon on the Growth and Recovery of
Plants from Drought and Temperature stress: Studies investigating the release
rate of the BOF silicon (Plant Tuff) were conducted in a peat medium at a pH of 6,
indicate that there is a fairly rapid initial release of Si into the soil solution, followed by a long, lower- level release rate. It was also demonstrated that Silicon
has the potential to significantly reduce the detrimental effects of salinity (functionally analogous to drought stress) on corn growth.
The daily transpiration is
highly correlated with dry weight gain. Photo 2 (Bugbee 2012)
Photo2
46
272
USDA-ARS (Zellner, 201347):
Perform studies on the essentiality of silicon in plants and the use of PlantTuff as
a floriculture and/or nursery media amendment. Determine micronutrient availability from various slags in Greenhouse and Nursery crops.
Evaluate current available silicon test methods and potentially develop alternative
methods: It was observed that the use of various extraction methods utilizing the
Molybdenum Blue Colorimetric method for quantifying “available” silicon (Si) were
inadequate, not only for BOF but also for other slag-like materials. To continue
ongoing studies of improving detection methods for total plant Si and bioavailable
Si from PlantTuff, slags and other materials and verify by these methods by plant
uptake measurements.
7- Conclusions
Although work is still needed in the development of appropriate product test
methods, the ability of slag to meet the current environmental and agricultural regulatory requirements, combined with the research that confirms its efficacy as an agricultural product has opened a new range of opportunities for these products. This enables the industry to promote and document slag as a valuable product.
My thanks and appreciation to Professors Lawrence Datnoff and Brenda Tubana of LSU, Dr. Jonathan Frantz and Dr. Wendy Zellner of the USDA-ARS, and
Professor Bruce Bugbee of Utah State University for their support and continued assistance on this project.
References
1 Basta, N.T., Ryan, J.R., Chaney, R.L. 2005. Heavy metal and trace element chemistry in residualtreated soil: a review of impacts on metal bioavailability and sustainable land application. Journal of
Environmental Quality. 34(1):49-63.
2
J. Frantz, J. Yzenas and R. Friedrich (2012): “Evalusting the Potential for Slag as a Source of Sup-
plemental Silicon in Container Crop Production”. ASHA Conference
3
G.B. Buck, G.H. Korndorfer & L.E. Datnoff (2010): “Extractors for Estimating Plant Available Silicon
from Potential Silicon Fertilizer Sources”, Journal of Plant Nutrition, 34:2, 272-282
47
.”
273
4
L.E. Datnoff (2009):At the Border Between Plant Nutrition and Plant Protection:, New Ag Internation-
al 28-32
5
B. Tubana (2012): Abstract submissions for 2012 ASHA Conference
6
B. Bugbee (2012-2013): Research report summary
7
W. Zellner, J. Atland & J. Locke (2013): Research proposal “Studies on the essentiality
of silicon in plants and the use of PlantTuff as a floriculture and/ or nursery media amendment.”
274
SR van der Laan1), JBA Kobesen1), EJ Berryman2), AE Williams-Jones2)
Accelerated weathering of LD-slag using water and CO2.
1
TATA Steel Europe, IJmuiden, The Netherlands
2
Dept. of Earth and Planetary Sciences, McGill University, Montreal, Canada
1
Abstract
We have studied accelerated weathering of LD-slag under controlled conditions in
the laboratory to determine the changes in mineralogy, the extractable elements, and
the ultimate fate of slag after prolonged exposure to H2O-CO2 fluids.
In the experiments, batches of approximately 2 grams of converter slag (lime-free,
grain-size 2-3.3 mm) were subjected to leaching in a flow-through reactor using
5%CO2-95%H2O fluid at temperatures of 120, 150 and 180ºC. Experiments were
conducted at elevated pressures to maintain a single fluid phase, and employed a 3
ml/min flow rate for 7 days. The microstructure and mineralogy of the slag residue
has been analysed with SEM-EDS and the fluid composition with AAS. In this report,
we focus on the results of one experiment at 150ºC for which a mass balance has
been calculated between the solid residue and the extracted fluids.
In our flow-through experiments, the contact time of the fluid with the slag was insufficient for the fluid to reach saturation with C2S of the slag, as predicted from dissolution modelling (our Haenchen/Westrich model), nor to form calcite and amorphous
silica, as predicted from HCH equilibrium thermodynamic modelling. This is confirmed by the product mineralogy of the experimentally reacted slag which reflects
strong leaching. Only Al- and Fe-oxide/hydroxide remained at the rims of slag grains.
A Ca-phosphate phase developed in contact with the un-reacted cores of the slag
grains.
275
Mass loss by leaching has been assessed by comparing residual to original slag
composition as well as by determining the mass extracted by the leaching fluid. The
order of mobility of the components in the slag was established from the mass balance, as silica with ~approx. 63-70% extracted, followed by lime - 56%, sodium 50%, vanadium - 25%, and magnesia 5% extracted. Manganese was depleted at the
inlet and enriched at the outlet of the flow-through reactor. Several components,
namely, alumina, phosphate, titania and iron oxide accumulated in the slag residue
compared to the starting slag. The slag was slightly more enriched in all these elements at the outlet than at the inlet, consistent with an expected mobility in the flow
direction.
The source of the additional alumina and phosphate is unclear. If it was introduced
with the leaching fluid it would need to have been present at the 1-2 mg/l level. The
accumulated iron- and titanium-oxide may have been derived from the tubing or reactor (stainless steel and titanium alloy).
The above observations provide insight into the ultimate fate of slag during prolonged
weathering and the mobility of major and minor elements in slag
2
Introduction
Use of LD-slag for civil applications is tightly controlled by legislation on environmental impact. In addition to element-leaching pH and redox effects are also monitored.
Conditions imposed on LD-slag in civil applications are those of natural weathering.
The reactions occurring in slag at ambient atmospheric conditions are similar to
weathering reactions of natural rocks and likewise proceed too slowly to show major
bulk alteration on the time scale of years. Alteration progress is generally hindered by
reaction products accumulating at the rock surface. It is well established that at moderate temperatures and ambient pressure, carbonation produces calcite crusts on the
slag grains and a silicate residues on the reactive mineral, inhibiting further reaction.
Therefore grinding of slag to below 50 microns is required to achieve complete alteration [1],[2],[3],[4],[5].
276
At IJmuiden, a project was initiated to study the chemical breakdown of LD-slag using carbonic acid (CO2-H2O mixtures) in order to investigate the ultimate fate of slag
during weathering. The aim was to prevent reaction product accumulation by maintaining leachate compositions undersaturated with respect to common product phases, thus requiring the first step in the reaction sequence to be the congruent dissolution of the dominant Ca-bearing phase of slag, larnite/C2S (Ca2(Si,P,V)O4). The fate
of phosphorous and vanadium in the C2S phase was of particular interest. The question is whether phosphorous dissolves out of the slag, or remains in the insoluble residue. The vanadium mobility is of interest for long term predictions of the environmental impact of slag-leaching.
Using Thermochemical modelling, the conditions for congruent dissolution of C2S
were calculated to be ~180ºC for a 5%CO2-95%H2O fluid. At lower temperature, silica dissolution lags behind that of calcium, whereas the reverse is true at higher temperature. Elevated pressures (above 100 bar) are required to retain the CO2-H2O as
a single liquid phase. The upper limit for the solubility of C2S in this fluid, expressed
as the concentration of Ca and Si in equilibrium with C2S+calcite+amorphous silica
or in equilibrium with calcite and silica in CO2-saturated fluid, was calculated to be
about 1 10-2 mol/kg solute (Fig 1a, 1b) at 100-200ºC. The dissolution kinetics of C2S
(Fig. 2) were calculated from the experimental data of Westrich et al. [6], using the
Arrhenius-type model of Haenchen et al. [7]. The model predictions for equilibrium
solubility and dissolution kinetics were tested in laboratory experiments.
277
1.E+02
300 bar CO2 sat.
solute (mol/kg water)
solute (mol/kg water)
1.E+02
1.E+00
1.E-02
1.E-04
P=300 bar
CO2 undersat
1.E+00
CO2
Ca
Si
1.E-02
1.E-04
1.E-06
1.E-06
0
0
100 200 300 400
T (in C)
-log( r) mol/cm2s
Fig. 1a: HCH-model calculated fluid composition
saturated with reaction products silica+calcite
and CO2
100 200 300 400
T (in C)
Fig. 1b: HCH-model calculated fluid composition
saturated with C2S and the reaction products
silica+calcite
-4
25C
-5
90C
150C
-6
250C
-7
batch 25C
-8
pH stat 25C
-9
-10
-11
0
2
pH
4
6
8
Fig. 2: Solution rate model of Hänchen et al (2006) applied to
C2S, anchored on the data at 25ºC of Westrich et al (1993),
making use of the Arrhenius constants of Eact=52kJ/molK,
n=0.5. with a fitted pre-exponential constant A=45 mol/cm2s.
A series of slag dissolution experiments were carried out at McGill University at temperatures of 125, 150, 180 and 200ºC and elevated pressures using 5%CO295%H2O fluids. The experimental results are used here 1) to validate the predicted
solubility and kinetics, 2) to determine the mineralogy of the residue after C2S dissolution, and in particular, 3) to determine the mobility of the minor elements, phosphorous and vanadium, the behaviour of which is difficult to predict. Some results have
278
been presented elsewhere [8], [9]. A more in depth evaluation of the experiments is
the subject of the current report, in which we show the mineralogical change in the
steel slag and relate this to the change in the complementary fluid leachate.
2
Experimental methods
2.1
Experiments
Starting material for the experiments consists of LD-slag (no free lime) with a grain
size of 2-3.3 mm. The slag, as described in Table 1, is well characterized for mineralogy (Rietveld-XRD) microstructure and phase composition (SEM-EDS– PARC) and
bulk chemistry (XRF). The slag was packed in a reactor tube and a fluid was percolated through the sample (5%CO2-95%H2O, 3 ml/min over 2.15 g of sample). Fluids
were automatically sampled every 3 hrs for chemical analysis. Chemical analyses
were performed using Atomic Absorption Spectroscopy following standard protocol in
the geochemical laboratories at McGill University, after stabilizing dissolved components in solution [10]. After seven days of fluid percolation, the reactor tube was sectioned and analyzed for mineralogy of the residue. Optical microscopy showed strong
similarities in the residues of all the experiments. For this reason we have limited
ourselves to a full investigation only of the mineralogy of the experiment at 150ºC and
250 bar
Table 1: Characterization of the starting material using XRD-Rietveld, PARC and
XRF. Data subjected to consistency checks are shown in bold (XRD-based
bulk chemistry using indicated mineral formulae)
279
XRD Rietveld-analysis
RD9935
norm alized
SEM-EDS PARC-analysis
wt% 2σ
AvgArea
unclassified
amorphous*
3.0
6.0
Wuestite (FeO)
6.5
1.9
16.9
1.8
Mg-W uestite ([Mg,Fe]O)
normalized
AreaSdev Densities
0.22
0.16
0
empty spectra
0
0
0
embedding
0
0
0
19
2.4
5
24.1 Wuestite
wt%
1σ
26.0
3.4
Srebrodolskite (Ca2Fe2O5)
22.2
1.0
22.9 C2F
20
3.2
3.3
18.0
2.9
Larnite (Ca2SiO4)
40.4
1.9
51.7 C2S
59
4.4
3.3
54.0
4.3
a'-C2S (Ca2SiO4)
9.8
1.9
Lime (CaO)
1.1
0.4
1.6
0.69
3.4
1.5
0.6
Portlandite (Ca[OH]2)
0.1
0.1
Calcite (CaCO3)
0.1
0.2
Aragonite (CaCO3)
0.0
0.3
0.035
0.027
0
1.3 Lim e
CaS
total
100.0
chemical composition (wt%)
XRD based
GLT1000
XRF
PARC-based
1.9
Na2O
<0.2
0.1
10.1
7.78
8.7
2.72
2.4
18.1
14.8
16.0
P2O5
1.58
1.7
K2O
0.015
MgO
Al2O3
SiO2
CaO
43.4
TiO2
44.6
43.3
1.38
1.5
MnO
2.9
4.69
4.9
Fe2O3
25.6
23.1
22.3
100.1
100.7
100.9
total
2.2
99.9
Analysis of run products
Fluid compositions
Time series changes of fluid composition (Ca and Si) are presented in Figure 3. The
five experiments fall in three temperature ranges – 180ºC (Duplicate experiments),
150ºC and 120/125ºC (treated as a duplicate). Trend lines have been added to highlight the consistency in the five data sets. The Si-content of the leachate shows a
sharp drop in the first 30 hrs to an approximate plateau value. The plateau values
decrease with run temperature from about 1mmol/l at 180ºC to about 0.2 mmol/l at
120ºC. It should be noted that the fluid used for dissolving the slag has a blank Si
value of 0.12mmol/l, which should be subtracted from the plotted Si-values, if mass
balance calculations are to be performed. The trend lines for Ca display a different
behaviour, never reaching a plateau value. Instead the Ca-content of the leachate
drops continuously during the experiment, initially rapidly, starting at much higher
values than Si (approaching 10 mmol/l at the onset of the 180ºC experiment), but
280
after 20 hrs linearly, reaching much lower values than Si at the end of the experiments. Calcium concentration in the fluid increased with increasing temperature.
For one of the runs (2501010), a more elaborate set of analyses was performed,
which included determination of Mn, Fe, P and Mg concentrations in addition to those
of Si and Ca . The results of these analyses are shown in Figure 4 as a time series.
Of the additional elements, only Fe developed a plateau value. Manganese concentrations are similar to those of Ca, and Mg and P concentrations are consistently
lower than those of the other elements throughout the experiments. The P-values in
the leachate are about 50 times lower than those for Si. As the P-content of the unreacted slag is only about 10 times lower than its Si-content, this suggests that phosphorous was retained in the solid residue
100.00
Si-trendline
Ca-trendline
180C 310311 Ca
180C 090511 Ca
150C 250110 Ca
concentration mmol/l
10.00
125C 121109 Ca
120C 180610 Ca
310311 Si
090511 Si
180ºC
1.00
250110 Si
121109 Si
150ºC
180ºC
120ºC
0.10
180610 Si
090511 EC
250110 EC
310311 EC
121109 EC
180610 EC
150ºC
120ºC
0.01
0
20
40
60
80
100
120
Run time (hr)
140
160
180
Figure 3: Ca and Si contents of fluids versus time from slag dissolution experiments
at various temperatures. Trendlines have been added to improve readability of
the data. Note that the 120ºC lines also represent the 125ºC experimental points.
281
Experiment 250110
150C, 250 bar, 5% CO2, 7 days
10.000
mmol/kg
1.000
Si
Mn
0.100
Ca
Fe
0.010
P
Mg
0.001
0
20
40
60
80
100
120
140
160
180
Duration (hours)
Figure 4: Fluid composition sampled during a 7 day leaching experiment.
Slag residue compositions (solids )- Results of phase mapping – PARC analysis.
Slag residues were prepared as polished sections to examine the effects on the slag
grains of their progressive interaction with the leaching fluids over the length of the
reactor. To quantify the effects of leaching we used SEM-EDS analysis with PARC.
PARC (PhAse Recognition and Characterization) is an off-line software package developed at TATA Steel IJmuiden labs, which uses Spectral Imaging (SI) datasets as
input to automatically find the different phases followed by an accurate quantification
of elements detected in the phases yielding correct stoichiometry [11].
Average phase compositions obtained from PARC for two entire grains, one from the
inlet and the other from the outlet of the reactor, are presented in Tables 2 and 3.
The microstructure is illustrated in Figure 5, and shows the distribution of phases in
the grains as established with PARC. Both the inlet and outlet grains, the same rim to
core reaction zones, each distinguishable by a characteristic phase assemblage.
1. Outer zone containing primary wuestite and goethite + gibbsite-like amorphous
alteration products
282
2. Zone with primary wuestite, remnants of Ca-ferrite and alteration products (amorphous goethite, gibbsite + carbonate apatite-like products)
3. Zone with primary wuestite, Ca-ferrite and carbonate apatite with limited porosity.
4. Core with all primary minerals (including Ca-Silicate) and limited alteration.
The PARC model successfully assigns almost all pixels to the chosen categories of
phases. Only a few percent of the pixels, shown in white in Figure 5, were not classified. Pixels assigned to “embedding “ or “empty pixels” represent porosity in the
sample. The porosity development in the outer zones is very well reflected in the sum
of the pixels assigned to these two categories, ranging from 40 vol.% in the outermost zone-1 to none in zone-4. Furthermore, the original, unaltered compositions of
the primary slag phases wuestite, srebrodolskite and C2S/larnite are faithfully reproduced in our analysis of the altered grains and turn out nearly identical in the two analysed grains (Table 2 and 3). Chemically identified alteration phases (PARC), which
could not be confirmed with XRD are: Gibbsite, Ferrite_remnant, Fe_oxide_Goethite,
Carbonate-Apatite and C2S_Si_depleted. The names we used for these phases are
provisional and are based only on their averaged chemical composition as derived
with PARC. Point analyses on these phases show a range of compositions. Nonetheless, the PARC compositions (sum spectra of all assigned pixels) give very similar
results for the two grains.
1,
4
0,
4
3,
4
2
3
4
1a
1a
2
3/4
1b
3,
0
Figure 5: Phase distribution (PARC) in
reacted grain at inlet (left) and outlet (right) after leaching with 5%CO295%H2O fluid (7 days, 150ºC, 3ml/min)
283
Table 2: Abundance (area percentage) and phase composition of primary phases and reaction products for a slag grain
at the inlet of the reactor tube.
Area percentage
phase
density
Phase compositions
Zone 1
Zone 1b
Zone2
Zone3/4
fresh
Na2O
Al2O3
0.0
0.4%
0.2%
1.4%
2.1%
0%
empty_spectra
0.0
36.0%
33.0%
23.2%
2.1%
0%
embedding
0.0
13.0%
12.0%
11.3%
1.4%
0%
.51
Gibbsite
2.5
17.0%
16.8%
3.8%
0.4%
0%
.037
Wuestite
5.0
14.8%
19.0%
18.7%
21.0%
19%
Ferrite_remnant
3.0
8.7%
9.5%
7.8%
3.3%
0%
.16
Fe_oxide_Goethite
2.5
9.2%
8.3%
1.5%
0.3%
0%
.052
CaSilicate
3.3
0.0%
0.0%
0.0%
6.8%
55%
.16
.064
CaFerrite
3.3
0.2%
0.4%
5.1%
18.2%
20%
.016
.84
Carbonate_Apatite
3.2
0.7%
0.6%
27.0%
34.3%
0%
.19
.48
8.3
Lime
3.4
0.0%
0.0%
0.3%
9.5%
2%
.092
.11
1.6
3.3
0.0%
0.0%
0.0%
0.6%
0%
.058
.98
bulk 
1.68
1.90
2.34
3.43
3.48
C2S_Si_depleted
.18
MgO
unclassified
1.3
.
5.1
2.3
1.9
.028
6.4
18.
21.
8.2
P2O5
3.1
3.1
3.5
4.9
1.3
2.3
18.
1.5
4.2
.66
10.
.18
.42
2.8
.65
30.
1.2
2.2
13.
.049
18.
4.7
13.
.097
CaO
4.5
55.
30.
.018
SiO2
.011
.36
2.1
3.4
1.1
2.7
.5
2.9
.14
21.
8.5
.079
62.
45.
TiO2
V2O5
MnO
4.8
1.6
4.7
35.
6.2
1.7
5.
48.
2.5
3.8
58.
1.3
1.3
8.1
11.
.048
.23
6.
3.8
3.2
3.8
.94
1.2
15.
.76
3.7
.011
Fe2O3
28.
52.
64.
80.
1.4
5.2
1.4
1.8
35.
46.
4.9
1.1
2.
14.
72.
1.5
.42
74.
.1
.21
.85
12.
2.6
13.
Table 3: Abundance (area percentage) and phase composition of primary phases and reaction products for a slag grain
at the outlet of the reactor tube.
Area percentage
phase
density
Phase compositions
Zone 1
Zone 2
Zone3
Zone4
fresh
Na2O
Al2O3
0.0
0%
1%
1%
3%
0%
empty_spectra
0.0
26%
18%
5%
0%
0%
embedding
0.0
15%
11%
0%
.88
Gibbsite
2.5
17%
3%
0%
0%
0%
.093
Wuestite
5.0
22%
21%
20%
17%
19%
Ferrite_remnant
3.0
11%
7%
2%
0%
0%
.12
Fe_oxide_Goethite
2.5
7%
1%
0%
0%
0%
.077
CaSilicate
3.3
0%
0%
0%
40%
55%
.2
.073
CaFerrite
3.3
1%
11%
21%
20%
20%
.028
.95
Carbonate_Apatite
3.2
1%
27%
45%
12%
0%
.11
.37
8.6
Lime
3.4
0%
0%
0%
0%
2%
.17
.05
1.4
C2S_Si_depleted
2%
0%
3.3
0%
0%
2%
8%
0%
bulk 
2.09
2.59
3.30
3.47
3.48
.16
MgO
unclassified
2.6
.
4.4
2.2
2.2
.026
7.7
17.
P2O5
CaO
15.
2.
23.
9.1
5.5
6.4
3.6
14.
3.5
5.5
56.
1.3
2.5
27.
.019
SiO2
.074
17.
1.7
4.1
.69
10.
.43
2.3
.61
30.
1.2
2.2
15.
.011
3.7
2.6
3.2
.15
21.
8.4
.23
2.
1.1
.47
62.
TiO2
V2O5
5.1
1.7
5.
37.
5.9
1.8
4.4
45.
4.
56.
1.3
27.
7.6
10.
.07
2.5
1.1
.19
6.9
3.7
2.7
3.9
.96
1.4
MnO
16.
.67
3.9
.13
Fe2O3
54.
64.
80.
1.4
43.
5.6
1.5
1.9
36.
47.
4.6
1.3
1.8
13.
69.
1.5
.73
.77
2.9
Reconstruction of bulk residue compositions
Using known densities for phases, or estimated values if not available, area proportions together with phase compositions were recalculated into bulk compositions for
each zone (see Table 4). Pixels which belong to the categories “unclassified”, “empty
pixels” and “embedding” do not contribute to the calculated bulk composition, since
their densities have been set at zero. This approach is only questionable for the “unclassified pixels” which likely represent solid matter. However, “empty pixels” and
“embedding” represent voids in the altered slag and their composition can be safely
excluded in the bulk. No assumption is involved in the reconstruction of the zone
compositions other than the use of estimated densities for alteration phases of which
a crystalline nature could not be confirmed with XRD.
3
Slag alteration
SEM-images faithfully reveal the structure of the slag grain remnant, including all
spatial relations of minerals and voids. Area percentages of phases in 2D sections
are directly proportional to volume percentages in the bulk slag. This also applies to
the void fraction which is measured as empty pixels and embedding and which increases from zero in the core to around 40% in the most altered surface zone. It ap284
pears that wuestite behaves as an inert phase, i.e., one that did not react in the experiment. Such an inert mineral in the slag should show a constant volume fraction in
each zone. This is observed for wuestite, with area (thus volume) fractions being
constant in all zones at a value of 20 +/- 2% vol.%, with the exception of Zone 1a in
the inlet grain, which contains only 14% wuestite. The wuestite content of the starting
material (Table 1) is 19±2.4 vol.%.
Table 4: Bulk chemical compositions of leached zones for grains at the inlet and outlet of the reactor after 7 days leaching with a 5%CO2-95%H2O fluid (3ml/min) at
150ºC. Original slag composition (fresh) included for reference.
Area compositions (outlet)
Area compositions (inlet)
Zone 1
Zone 2
Zone3
Zone4
fresh
Na2O
0.0
0.1
0.1
0.1
0.1
8.5
MgO
14.7
11.2
8.7
7.0
8.5
5.4
2.3
Al2O3
14.9
7.7
6.5
3.3
2.3
4.3
15.6
SiO2
1.0
1.3
1.9
13.0
15.6
1.5
P2O5
1.7
7.4
9.3
4.2
1.5
2.5
22.7
32.0
42.3
42.9
2.1
1.6
1.2
Zone 1
Zone 1b
Zone2
Zone3/4
fresh
Na2O
0.0
0.0
0.1
0.1
0.1
MgO
13.6
15.4
12.4
9.7
Al2O3
17.5
15.6
7.9
SiO2
1.1
1.0
1.4
P2O5
1.8
1.5
8.3
7.8
CaO
2.0
2.2
21.4
34.3
42.9
CaO
TiO2
4.2
3.8
3.3
2.9
1.6
TiO2
3.6
3.3
3.4
V2O5
1.6
1.4
1.1
0.9
1.2
V2O5
1.3
1.1
1.0
1.1
MnO
7.5
8.3
7.0
5.7
4.8
MnO
9.1
7.3
6.1
4.6
4.8
21.7
Fe2O3
51.1
37.8
31.1
22.4
21.7
Fe2O3
50.7
50.7
37.1
29.0
The overall mineralogical changes in the slag resemble a "lateritization" process,
wherein the end-stage is represented by only alumina, and Fe-(hydr)oxides with the
inert marker phase wuestite. Nearer to the grain core, a phosphate-phase develops
as an amorphous apatite-like compound.
The PARC data can be used directly to calculate the bulk compositions, if we make
the assumption that the measured grains are representative. The PARC bulk compositions of the inlet and outlet grain are compared in Table 5. In fact, the PARC data
represent a 3 dimensional structure in which individual zones represent shells as in
an onion. We have reconstructed the grain composition using the radii of these shells
(Fig. 6) derived from the exposed zone widths in the polished sample. The dimensions of the unexposed core are speculative, but the core itself is likely to exist,
based on the grain size fraction used for the experiments (2-3.3 mm) and the exposed cross sectional diameter being only 1.9 and 1.65 mm. Reconstructed bulk
composition from planar data and a 3D structure assumption are different because
285
the inner shells contribute less to the bulk composition in 3D than their area% in 2D.
These bulk compositions for the altered slag cannot be used to infer directly how
much material was leached, simply by comparison to the fresh slag. In order to calculate true mass loss by leaching, the density difference between fresh and altered
grains needs also to be taken into account. This is done in the next section.
W1
W2
W3
W1b
W4
sectioning
level
Table 5: Bulk Compositions derived with PARC from area view
and recalculated assuming 3D shell-structure (see Fig. 4)
r1
area data
Core
r3
r4
Figure 6: Slag grain alteration zones of
various observed widths (Fig 5)
recalculated to shell radii, assuming a concentric alteration pattern
3D interpreted
fresh
inlet
outlet
inlet
Na2O
.1
.074
.059
.1
.1
MgO
8.5
11.1
10.7
r2
12.
11.
outlet
Al2O3
2.3
9.8
9.3
7.9
8.2
SiO2
15.6
2.2
2.4
7.9
6.
P2O5
1.5
6.
CaO
42.9
19.
TiO2
1.6
3.4
V2O5
1.2
1.2
MnO
4.8
6.9
Fe2O3
21.7
39.
5.7
3.1
4.2
26.1
24.3
3.3
2.7
2.9
1.1
1.2
1.1
7.4
6.2
6.8
33.8
35.7
20.
39.
Table 6: Shell radii calculated from exposed zone widths (Fig. 5) assuming a total
grain radius and size for the unseen core.
inlet
observed width, respective of centre line of grain
width zone
w1
w1b
w2
w3
w4
micron
321
321
214
268
32
unseen
r2
r3
r4
core
r1
radius shell
1150
1150
943
845
791
delta r
207
207
98
54
1
Vol (mm3)
2.86
2.86
0.98
0.46
0.01
total w
790
836
depth of
total r
sectioning
1150
360
total Vol.
2.07
6.37
core & total radius are fitting parameters
outlet
observed width, respective of centre line of grain
width zone
w1
w2
w3
w4
micron
300
300
150
200
radius shell
r1
r2
r3
r4
1250
1040
885
837
delta r
210
156
48
337
Vol (mm3)
3.46
1.82
0.45
1.93
total w
unseen
core
500
950
depth of
total r
sectioning
1250
438
total Vol.
0.52
8.18
core & total radius are fitting parameters
4
Mass balance between solids and fluid
Using measured area fractions for phases (respective of void area), phasecompositions and phase-densities (Table 2, 3) and measured reaction zone widths in
286
the exposed grain surface, mass loss during the experiments was calculated. The
fluid compositions (Fig. 3, 4) were also used to calculate the extracted mass of solids. Based on the sieve grain-size (2.0-3.3 mm) and measured reaction zone width in
the exposed grain surface (Fig. 5), an idealized spherical shell volume for each reaction zone was derived (Table 6). Only two parameters are unconstrained, 1) the true
grain diameter and 2) the size of a more-or-less exposed, unreacted core. We have
chosen to use these two unknown parameters to fit a mass balance with the extracted by fluid.
For soluble compounds (MgO, CaO) with no tendency to sorption, a good massbalance with the leachate fluid composition permits the two radii parameters to be
fitted. The mass balance for the fluid composition is fitted equally well for the inlet as
the outlet grain (Table 7), however, concentrations of some fluid components (notably Si, Fe, P and less so Mn) are too high to be consistent with the analyzed slag residue composition. For the calculation, we assumed that all grains in the experiment
behaved either as at the inlet or as at the outlet grain, and the mass lost from the individual grains (difference between altered and fresh grains of equal radius) was extrapolated to a lost mass for the total mass of starting slag. From this mass balance
exercise, we were able to establish an order of mobility for the slag with silica
~approx. 63-70% extracted, followed by lime - 56%, sodium - 50%, vanadium - 25%,
and magnesia 5% extracted.. Manganese was depleted at the inlet and enriched at
the outlet. In addition, several elements, namely Al, P, Ti and Fe accumulated in the
slag residue compared to the starting slag. All these elements are slightly more enriched at the outlet than at the inlet, consistent with an expected mobility in the flow
direction.
The source of alumina and phosphate is unclear. If they were introduced with the
leaching fluid they would need to have been present at the 1-2 mg/l level (the experiment had a duration of 30 l / 7 days). The iron- and titanium-oxide may have been
derived from the tubing or the reactor (stainless steel and titanium alloy) in agreement with their presence in the fluid as well.
287
Table 7: Mass balance between solid and fluid
experiment mass balance (in mg)
mass
initial
outlet
inlet
mass lost in run
outlet
inlet
Na2O
2.51
1.23
1.36
-1.28
-1.15
MgO
182.17
176.85
174.15
-5.31
-8.02
Al2O3
50.25
136.25
123.98
85.99
73.72
SiO2
335.02
99.58
123.75
-235.44
-211.27
mass in
fluid
7.70
%remaining
outlet
inlet
49%
54%
97%
96%
271%
247%
307.83
30%
37%
P2O5
31.41
69.11
48.79
37.70
17.38
10.05
220%
155%
CaO
921.31
403.30
409.78
-518.01
-511.53
518.64
44%
44%
TiO2
33.50
47.68
42.41
14.18
8.91
142%
127%
V2O5
25.13
18.81
19.33
-6.32
-5.80
75%
77%
MnO
102.60
112.38
97.09
9.78
-5.51
15.10
110%
95%
466.10
591.79
530.52
125.69
64.42
53.08
127%
114%
2150.0
1657.0
1571.2
-493.0
-578.8
912.40
Fe2O3
total
5
remaining mass
Fluid composition in relation to slag
The fluid composition of run 250110 changed from an initial Ca/Si = 2-3 to Ca/Si = 1
after 50-60 hrs dropping to Ca/Si = 0.2-0.1 after 170 hrs. The overall extracted fluid
has a Ca/Si = 1.8 using a fluid blank corrected Si-content. For the C2S phase, the
Ca/Si = 2.2. We therefore conclude that throughout the entire run, Si-extraction ran
ahead of Ca, implying incongruent dissolution of C2S with Ca staying behind. A similar argument can be based on the slag residue composition. A Ca/Si = 2.4-2.6 was
calculated for the lost mass from the grains, which could be consistent with congruent dissolution of C2S. However, knowing that the C2F phase also reacted, if C2S
and C2F were to have reacted at the same rate, which they did not, the Ca/Si = 3.2
for the overall bulk slag is probably the more relevant figure for comparison. Our
PARC observations indicate that C2S disappeared, leaving a Ca-phosphatecarbonate residue. The steady state end-value for Si in the fluid of 0.14 mmol/kg
needs to have been accompanied by 0.3 mmol/kg of Ca. Even if externally supplied
phosphate locally formed Ca-phosphate-carbonate with Ca from released C2S, this
would be insufficient to explain the Ca/Si imbalance in the fluid. It can only account
for fixation of 0.03 mmol/kg of Ca. The only explanation we can propose is that the
applied blank correction for Si is too low, and that less Si was released. Compared to
the maximum solubility calculated for fluids saturated with C2S ± calcite ± silica using
the HCH-model, the values in experiment 250110 were initially similar but subsequently lower by a factor of 5-10.
The dissolution rates based on the Haenchen-model (section 2) were used to estimate the fluid composition for a slag-fluid contact time of 1 minute, which is comparable to the time needed for the passage of the fluid from the inlet to the outlet of the
reactor. Assuming a grain-size of 3 mm, 2.15 gram of slag in the reactor would consist of 43 grains with a total surface area of 12 cm2 (assuming perfect spherical
shapes). The calculated upper limit for the dissolution rate at pH=5 and 150ºC of 10-7
mol/cm2/s yields a mass transfer to the fluid of 2 mmol/l and the lower limit at pH=5,
90ºC of 10-9mol/cm2/s a mass transfer to the fluid of 0.02 mmol/l. These values agree
quite well with the measured concentration range for Ca and Si in the fluid. For longer contact times, e.g., 5 minutes or more, it is likely that saturation with calcite and
silica would have been reached within the reactor. If element mobility were estimated
only on the basis of fluid composition, a conclusion that 42% of the starting slag dissolved would be warranted. Based only on the changes in the nature and composition of the slag, the mass loss is estimated to have been around 25%, although it is
difficult to account for the observed gains for some elements using these data. The
enrichments in Al, P, Ti and Fe of the grains could potentially be attributed to densification of the outer shells, combined with differences in solubility. In such a model,
alumina would be least soluble and most densified, by a factor of 6-7 (Table 4), between the outer shell and the fresh slag. However, the apparent behaviour of
wuestite as an inert marker in the microstructure argues against the densification
model. Neither the wuestite grain-size nor its volume proportion points towards a
densification. More complex models can possibly be developed involving element
migration into specific enrichment shells, driven by chemical gradients between the
fresh slag in the grain core and the surrounding fluid. Such models still need to account for an overall net accumulation in the grains without regions of net loss.
6
-
Conclusions
Model solubilities for Ca and Si (HCH) and model dissolution kinetics (following
Haenchen/Westrich) developed as part of this study have been shown to be consistent with the experimental data on slag dissolution, and can be of further use
for the development of a CO2-slag carbonation process.
-
In the CO2-H2O leached slag, the void fraction increases from zero to around 40%
in most altered surface zones and wuestite serves as an inert marker, as indicated by its area (thus volume) fraction remaining constant in all alteration zones.
-
Mineralogical changes in the slag reflect "lateritization", wherein only alumina, and
Fe-(hydr)oxides with inert marker wuestite remain at the grain surface. Nearer the
grain core, phosphate becomes enriched in an amorphous apatite-like compound,
demonstrating that phosphate removal is limited in this carbonic-acid leaching
treatment.
-
The environmental impact of slag leaching for the critical element, vanadium,
might need re-evaluation, because vanadium seems to have been largely (75%)
immobilized in the solid slag residue.
-
Using measured phase-area fractions, phase-compositions and phase-densities,
the various bulk zone densities and compositions can be reconstructed.
-
Soluble compounds with no tendency to sorption yield good mass-balance with
the leachate fluid composition (MgO, CaO).
-
Alumina and phosphate became enriched in the leached slag sample, and more
so at the inlet than at the outlet of the reactor tube. The source of these elements
has not been identified.
-
Iron and titanium oxide also became enriched in the experimental sample, and
may have been derived from the tubing or the reactor (stainless steel and titanium
alloy), consistent with their presence in the fluid
7
References
[1]
Bonenfant, D, L Kharoune, S Sauvé, R Hausler, P Niquette, M Mimeault, M
Kharoune: CO2 Sequestration Potential of Steel Slags at Ambient Presssure
and Temperature. Industrial Eng. Chem. Res. V47 (2008), pp. 7610-7616
[2]
Gerdemann, SJ, DC Dahlin, WK O’Connor, LR Penner, and GE Rush: Factors
Affecting Ex-situ Minral Carbonation Using Calcium and Magnesium Silicate
Minerals. DOE/ARC 2004-032
[3]
Huijgen, WJJ, G Witkamp, RNJ Comans: Mineral CO2 sequestration by steel
slag carbonation. Environ. Sci. Techn. 39 (2005), pp. 9676-9682
[4]
Huijgen, WJJ, RNJ Comans: Carbonation of steel Slag for CO2 sequestration:
leaching of products and reaction mechanisms. Environ. Sci. Technol. 40
(2006), pp. 2790-2796.
[5]
Rawlins, CH: Geological sequestration of CO2 by hydrous carbonate formation
in steelmaking slag. PhD Thesis. Missouri University of Science and Technology. (2008)
[6]
Westrich, HR, ST Cygan, WH Casey, C Zemitis, GW Arnold: The dissolution
kinetics of mixed-cation orthosilicate minerals. American Journal of Science
V293 (1993), pp. 869-893
[7]
Haenchen, M, V Prigiobbe, G Storti, T.M Seward, and M Mazzotti: Dissolution
kinetics of forsteric olivine at 90-150ºC including effects of presence CO2.
Geochim. Cosmochim. Acta, 70, (2006) pp. 4403-4416.
[8]
Berryman EJ, AE Williams-Jones, AA Migdisov, SR van der Laan, Carbonation
of Steel Slag I, Goldschmidt Conference Abstracts (2011)
[9]
Van der Laan, SR , C Liebske, HJB Kobesen, EJ Berryman, AE Williams-Jones
AA Migdisov, Carbonation of Steel Slag II. Goldschmidt Abstracts (2011) p2063
[10] Berryman EJ, AE Williams-Jones, and AA Migdisov. Steel slag: a medium for
CO2 sequestration by carbonation. Environmental Science and Technology (in
review).
[11] Van Hoek, CJG, M. de Roo, G van der Veer, SR van der Laan: A SEM-EDS
study of cultural heritage objects with interpretation of constituents and their distribution using PARC data analysis Microsc. Microanal. 17 (2011) pp. 656-660
I. Sohn* , J.I. Hwang**, H.S.Kim*** , J.S Choi****, Y.S. Jeong***** and H.C. Lee******
Development of ECO Slag Processing Technology for Iron
Recovery and Value-Added Products in Steelmaking
* Associate Professor, Dept. of Materials Science & Engineering, Seoul Korea
** Manager, HYUNDAI STEEL, Incheon Korea
*** Principal Researcher, Korea Institute of Geoscience and Mineral Resources,
Daejeon KOREA
**** Principal Research Engineer, Korea Conformity Laboratories, Seoul Korea
***** Senior Researcher, Research Institute of Industrial Science & Technology,
Pohang Korea
****** Senior Vice President, HYUNDAI STEEL, Incheon Korea
Abstract
The production of crude steel in South Korea surpassed 70 million tons in 2012 and
consequently enormous amounts of slag (24 million tons) has been produced. Even
though granulated blast furnace slag is widely used in cement, steelmaking slags,
e.g. electric arc furnace (EAF) slags and basic oxygen furnace (BOF) slags containing FeO higher than 20wt%, are still not fully utilized for highly value-added resources
and optimal iron recovery technology from steel slags applicable to the industry has
yet to be fully realized. As environmental concerns become a critical issue in the sustainability of the steel industry, manufacturing of various kinds of glass-ceramics and
related functional materials recycled from post-mortem slags has become an important topic of research. In this study, the results of a joint research collaboration
between government-industry-academia-research institutes on various technical approaches for eco-friendly processing of EAF slags from primary steelmaking to cooling are introduced and discussed. Reduction of bulk iron content in EAF slags during
primary steelmaking utilizing a newly designed closed-type slag door, which provided
a relatively reducing atmosphere in the furnace before tapping, is presented. Further
reduction methods of the iron oxide in hot slag tapped from the furnace by adding
various reducing agents such as Al-dross and carbon is also introduced. The slag
composition, which was reduced and cooled using this secondary treatment is affect-
ed by the operating parameters and thermo-physical properties of the slag, and thus
can provide the potential for utilization in highly value-added products such as Portland cement and high-alumina cement. To increase the applications of the reduced
slag product, a new cement material as well as slag aggregates from the EAF slags
were developed by controlling the composition and morphology through novel methods. Furthermore, light weight aggregates using BOF slags and stabilization of the
steel slag was also investigated. From these trials, technical breakthroughs and
countermeasures to treat the increasing environmental problems originating from
steel slags are proposed.
1. Introduction
The blast furnace slags have found widespread use in the cement industry in the
form of ground-granulated blast furnace slags (GGBS) obtained from quenching the
slag in water or steam to produce an amorphous fine powder, which is blended with
Portland cement.[1-3] Unlike the blast furnace slags, the electric arc furnace and ladle
slags produced from steelmaking has had limited applications in higher-value added
by products due to its high content of FeO and increased density, which make it unsuitable for use in cement applications. Thus much of the slags in the steelmaking
spectrum have been utilized mostly in the lower-value added product chain.[4-6]
Utilization of steelmaking by-products beyond the aggregates for road construction,
armour stones, or fertilizers is needed to provide additional revenue for both the steel
and cement industry. As CO2 regulations begin to be widely adopted through carbon
taxes, it is imperative that these two industries collaborate to ensure a common synergy.[7-8] Calcination of dolomite or limestone to produce clinkers may not be a long
term viable solution for the cement industry with approximately 5% (1.88 Gt) of anthropogenic global CO2 emissions annually. Thus, application of steelmaking slags to
higher-value added products such as cements is needed.
In this study, various technical approaches for eco-friendly processing of EAF slags
from primary steelmaking to cooling are introduced and discussed. Reduction of excessive EAF slag discharge utilizing a dual slag door hermetic design is introduced.
Near the end of the primary steelmaking process, reduction methods of the iron oxide
before tapping the furnace by adding various reducing agents such as Al-dross and
carbon to lower the total Fe content in the slag and increase yield has been studied.
To increase the applications of the reduced slag product and further lower the total
Fe content, separate lab-scale experiments were conducted in a 10 kg resistance
furnace with a rotating carbon rod to ascertain the degree the FeO reduction in the
slag and the compositional variation in the reduced slag. To identify possible cooling
conditions and effects of the slag composition to obtain amorphous calciumaluminate phases, fundamental work on the crystallization behavior of slags were
further conducted using the confocal laser scanning microscope. In addition, light
weight aggregates using BOF slags and the stability of the steel slag was also investigated.
2. Experimental
2.1 Methods and procedure of EAF slag reduction lab scale tests
A rotating carbon rod was used in a 10 kg resistance furnace to reduce the steelmaking slag. Figure 1 is a schematic of the apparatus. The carbon rods were made from
coke with coal tar and pitch as the adhesive and heated to 120~150oC and poured
into a cylindrical mold and pressed into rods.
Figure 1. Schematic of the carbon rod
Figure 2. Schematic of the confocal laser
agitated EAF slag reducing equipment
scanning microscope
2.2 Crystallization behaviour using the confocal laser scanning microscope
Calcium-aluminate (CA) and calcium-silicate (CS) slag samples were prepared by
mixing reagent-grade chemicals. The mixture was then pre-melted in the box furnace
at 1823 K (1550 oC) for approximately 5 hours under Ar atmosphere and poured onto
the copper plate to be solidified. Composition of the pre-melted samples was confirmed using x-ray fluorescence (XRF, S4 Explorer; Bruker AXS GmbH, Karlsruhe,
Germany) as shown in Table 1.
No.
Pre-experiment
CaO Al2O3 MgO FeO
Post-experiment
No.
CaO SiO2 MgO FeO
CA0
48.6 46.1 4.9
0
-
-
-
-
-
CA5
47.0 42.8 4.5
5.6
CS5
42.1
42.8
9.5
5.3
CA10 42.4 41.9 4.7 11.0 CS10 40.2
39.5
9.9
10.1
CA15 39.3 41.1 4.4 15.3 CS15 37.7
36.8
9.6
15.4
-
-
-
CA20 38.7 35.6 4.6 21.2
-
-
Table 1. Chemical composition of the pre-melted slag analyzed by XRF.
Crystallization behavior of CA and CS samples were observed in-situ using confocal
laser scanning microscope (CLSM; SVF-SP, Yonekura MFG. Co. LTD, Japan). The
experimental setting of CLSM was schematically described in Figure 2. A Pt-10Rh
crucible (H: 5.0 mm, ID: 4.95 mm, OD: 5.0 mm) including the pre-melted sample was
placed on the R-type TC plate in the confocal chamber. The sample was heated up
to 1823 K (1550 oC) and maintained for 5 min for complete homogeneity of the slag
melt, followed by continuous cooling at cooling rates from 25 to 800 K/min or isothermal cooling, where the slag melt is undercooled at a fixed temperature until the
primary phase is observed.
Composition (wt.-%)
Classification
Blast
furnace
slag
CaO
slag
T.Fe
Al2O3
Granulated
Aircooled
BOF slag
Steelmaking
SiO2
Cement materials
35~45 30~40
<1
10~20
Lower-value
35~45 10~20 20~30
<5
EAF Acidic 20~30 10~20 20~30
<10
slag
Applications
Basic 50~60 10~20
<1
10~20
added
slags for aggregates
- Road bed
- Fertilizer
Table 2. Major composition and application of the typical iron and steelmaking slags.
3. Results and Discussions
3.1 Reclamation of iron from EAF steelmaking process wastes
Slag produced from the EAF steelmaking operations is tapped through the slag door.
However, the design limitations in existing furnaces allows the discharge of high
FeO-containing slags during the initial foaming stages resulting in excessive iron loss
through the slag and in addition restricts the application of the EAF slags from highervalue added slag by-products. Typical compositions of iron and steelmaking slags
are provided in Table 2. To overcome these limitations and inhibit excessive slag discharge during EAF steelmaking processes, an innovative dual door hermetic slag
door was design in this study to control these issues. Through these design innovations applied to typical furnaces, effective control of C and Al additions to reduce the
FeO in the slag could be implemented and increase the final product yield beyond
existing EAF processes. Figure 3 shows the digital images of the existing single door
design and the modified dual door hermetic design.
Figure 3. Digital image of existing EAF single slag door and the modified dual door.
Figure 4. Comparison of the total Fe content in the slag and yield before and after
implementation of the EAF dual door.
Through this study, implementation of the modified slag door inhibited the premature
initial discharge of high FeO-containing slag and allowed a subsequent reduction
process of the FeO with reducing agents such as Al and C to decrease the Total Fe
content in the slag from 20% to 17% as shown in Figure 4. A fully sealed EAF operation promotes lower FeO turndown slag improving yield and lowering overall power
consumption. Typical yield improvements resulted in more than 0.3% return during
steel tapping.
3.2 Development of alternative cement mixtures from reclaimed EAF process slag
Blast furnace slags have been well-documented to be a stable substitute in the production of slag cements and have been expanded in use globally. However, EAF oxidizing acidic slags have yet to be implemented into the higher-value added stream
due to its high Fe content and density, which make it suitable for use in only lowervalue added products such as heavy aggregates or road bed ash. In order to lower
the Fe content in the molten EAF oxidizing slags, fundamental work on reducing the
slags with a rotating carbon rod to reduce the FeO, while providing forced convection
for increased kinetics, have been done. Increased reaction time with the carbon rod
resulted in lower Fe content in the slag producing increased CaO and Al2O3 content,
which closely resembled a CaO/Al2O3 ratio similar to the blast furnace cement slags.
In addition, for application in the higher-value added cement stream the reduced oxidizing slags needs to be amorphous and rapid quenching of the reduced slag in water produced the required phase conditions comparable to the blast furnace
quenched slags.
Figure 5. Comparison of the compressive strength of the granulated blast furnace
slags and the reduced EAF slags developed in the present study.
Figure 5 shows a comparison of the GGBFS and REAFS (reduced EAF slags) composites at various times. GGBFS is a mixture of 50% cement and 50% blast furnace
slags. REAFS is a mixture of 50% cement and 50% reduced EAF aggregate composite. Compressive strength of the developed EAF slag composite was found to be
comparable to existing blast furnace slag cements currently in use. Therefore, the
REAFS composites are expected to be applicable in the cement mixtures for achieving the latent hydraulic properties needed for commercial use. Currently, a semi-pilot
plant scale for simultaneous reducing and water-quenching for production of amorphous REAFS is under development.
3.3 Crystallization behaviour of synthesized EAF slags
Continuous cooling transformation (CCT) and time-temperature-transformation (TTT)
diagrams were plotted by recording the crystallization time and temperature of the
CA and CS samples during in-situ observation through CLSM. Overall, crystallization
temperature decreased with higher FeO content during continuous cooling of CA and
CS as shown in Figure 6. In Figure 6(a), crystallization temperature of low FeO samples decreased dramatically with FeO addition while the crystallization temperature of
FeO-rich samples were almost independent of cooling rate.[9-10] It seems that FeO
stimulates crystallization, which allows crystallization during rapid cooling to occur at
a temperature similar to the crystallization during slow cooling.
Figure 6. CCT of (a) calcium-aluminate based and (b) calcium-silicate based slags
with various FeO contents continuously cooled at cooling rates from 25 to 800 K/min.
In Figure 6(b), crystallization temperature increased when FeO content increases
from 5 to 10 % and decreased from 10 to 15 % as the liquidus temperature decreases with FeO additions in the CS system. This unexpected increase of the crystallization temperature at 10 % FeO stems from the accelerated crystallization via the addition of FeO, which causes the crystallization at relatively higher temperatures.
In Figure 7, crystallization time and temperature of both CA and CS decreased with
FeO addition. CS with 5 % FeO showed a “C”-shaped TTT curve contrary to other
samples, which indicate that a rapid cooling of this slag could result in the formation
of amorphous phase especially when the cooling rate is beyond the critical cooling
rate at which the cooling path passes through the TTT nose. The TTT curve became
a half “C”-shaped with higher FeO contents making it less likely for the formation of
an amorphous phase. This agrees well with the CCT curve in Figure 6(b), where the
crystallization of 5% FeO containing CS based slags was not observed during rapid
cooling at 800 K/min.
Figure 7. TTT of (a) calcium-aluminate based and (b) calcium-silicate based slags
with various FeO contents isothermally cooled below the liquidus temperature.
3.4 EAF basic ladle slags as an activator for blast furnace cement slag
Chemical composition (wt%)
Raw material
SiO2
Al2O3 Fe2O3
CaO
MgO
Na2O
K2O
SO3
LOI
4.71
26.10
2.74
51.45
5.94
0.02
0.01
1.34
6.40
Fly ash
39.70 21.50
6.95
16.80
2.45
1.04
1.08
2.34
7.95
Gypsum
0.89
0.14
44.60
0.07
-
0.07
53.80
-
EAF ladle
slag
0.38
Table 3. Approximate chemical composition of the raw materials.
Modulus
LSF
SM
IM
66.0
0.5
6.4
CM
P
97.9
3.3
Mixing ratio (wt.-%)
Slag
Fly ash
Gypsum
66
24
10
LSF: Lime Saturation Factor, SM: Silica Modulus, IM: Iron Modulus
CM=[CaO/{0.7(Fe2O3+SO3)+(0.55×Al2O3)+(1.87×SiO2)}]×100, P=Al2O3/SO3
Table 4.Raw materials mixture ratio and the modulus values.
Blast furnace slags and acidic EAF slags are recycled in cement admixture and aggregates. However, EAF basic ladle slags contain significant amorphous lime that
can expand and are mostly used in earth filling materials. Thus, to apply the EAF
basic ladle slags beyond its current use, fly ash and plaster is added in the CaOAl2O3 based slag to form CSA(Calcium sulfo aluminate) materials, which acts as a
activator for blast furnace cements and increase the initial low compressive strength.
The chemical composition of the raw materials utilizing the EAF basic ladle slag is
given in Table 3. The mixture ratio of the components and the modulus values are
provided in Table 4.
The CSA activators produced from the recycled EAF ladle slags results in needle-like
ettringite phases during the hydration process expressed in reaction (1) and (2).
Ladle-furnace Slag Activator (12CaO•7Al2O3(CaF)
3CaO•Al2O3•3CaSO4•32H2O + C-S-H (Aluminate gel)
+
CaSO4•nH2O)
→
(1)
3CaO•Al2O3 + 3(CaSO4•2H2O) + 26H2O ⇒ 3CaO•Al2O3•3CaSO4•32H2O (Ettringite)
(2)
Figure 8. Compressive strength of the slag cement with the addition of the activator.
These ettringite phases are produced in the initial stages of the hydration process
and have a long-term stability that increases the coagulation and initial strength of
the typical slag cements. The effect of the compositional mixture variations of the
OPC, slag, and CSA activators on the compressive strength has been measured, as
shown in Figure 8. Laboratory results indicate 7% of the CSA activator in the mixture
by weight is optimal for increasing the compressive strength. Moreover, slag cements
after 3 days of aging with the CSA activator in slag cements, where blast furnace
slags comprise more than 70%, showed compressive strengths higher than 10MPa
applicable to building concretes. Currently, control of the coagulation time and increased stability of the activator is being addressed.
3.5 Application of blast furnace slags for light weight aggregates
Aggregates are divided into various categories according to the classification method. The details of the aggregate classification provided in Table 5 are complex and
broad.
Condition
Classification
Remark
By production meth- Natural aggregates
Sands from rivers and
od
seas
Artificial aggregates
Crushed
stones
and
sands from stone quarries
By-product aggregates
By particle size
By usage
Coarse aggregates
Particle size > 5mm
Fine aggregates
Particle size < 5mm
Structural aggregates
Non-structural aggregates
By purpose
Concrete aggregates
Road aggregates
By weight
Heavy aggregates
Specific gravity >3.0
Ordinary aggregates
Specific gravity 2.0~3.0
Light aggregates
Specific gravity 1.0~2.0
Ultra-light aggregates
Specific gravity < 1.0
Table 5. Classification of the various aggregates.
Aggregates classified according to the weight or specific gravity are heavy aggregates, ordinary, light, and ultra-light aggregates, as shown in Figure 9. Recently,
stone quarry development and sand collection, which is detrimental to the environment are mostly restricted and recent trends to use manufacturing by-product aggregates like iron and steelmaking slags, copper slags, and recycled aggregates such
as recycled concrete has gained increasing popularity.
Figure 9. Classification of aggregates according to weight.[12]
Light weight aggregates should have low density, low absorption rate, and consequently have the capabilities to maintain a certain level of formability while having a
resistance to cyclic freezing and thawing.
Table 6. Characteristics of light-weight test aggregates compared to vendor samples.
Table 6 shows the characteristics properties of the developed light weight aggregates
using the blast furnace slag. A comparison of the results shows that the characteristics of the developed light-weight aggregates are identical to better than those of imported aggregates from four different vendors. The specific gravity is approximately
1.17, which is better than most of the compared samples except for vendor D. The
absorption rate is 14.5 %, which is slightly lower than the sample from vendor C.
Overall, considering both the absorption rate and specific gravity, the aggregate developed from the present study seems to show a highly competitive advantage.
4. Conclusions
1. Premature slag discharge in the EAF process was prevented by redesigning
the typical un-sealed single slag door to a novel dual door hermetic design.
This allowed controlled feeding of reducing agents including C and Al to
reduce the high FeO-containing before tapping and lower the total Fe 20%in
slag to an average of 17% resulting in increased yield, lower power
consumption, and lower FeO-containing acidic slags from the EAF process.
2. An increase in the CaO and Al2O3 content could be observed during reduction
of the FeO in the molten EAF acidic slag, which became comparable to the latent hydraulic activating materials in blast furnace slags.
3. CCT and TTT diagrams indicate that the EAF slag could become more likely
to be crystallized during cooling when alternated into calcium-aluminate based
slag from calcium silicate slag, especially at low FeO contents. Both systems
are easily crystallized at relatively higher FeO contents.
4. CSA activator produced from basic EAF ladle slags produced needle-like
ettringite phases during the hydration process for application in stimulating
blast furnace slags.
5. It can be concluded that investigated slag, when mixed with a blowing agent
for making spiracle and a fusing agent for lowering of the softening
temperature, and sintering at targeted temperatures. The light weight made
this way has the characteristics with density 1.17g/cm2 and absorption rate
14.5%.
Acknowledgements
This study was funded by the Ministry of Trade, Industry and Energy.
5. References
[1]
G.S. Osborne : Durability of Portland blast furnace slag cement concrete, Cement and Con-
crete Composites, Vol. 21, 1999, pp.11-21.
[2]
D.M. Roy and G.M Idorn : Hydration, Structure, and Properties of Blast Furnace Slag Ce-
ments, Mortars, and Concrete, ACI Journal Proceedings, Vol. 79, 1982, pp. 444-457.
[3]
S. Kumar, R. Kumar, A. Bandopadhyay, T.C. Alex, B. R. Kumar, S.K. Das, and S.P. Mehrotra :
Mechanical activation of granulated blast furnace slag and its effect on the properties and structure of
Portland slag cement, Cement and Concrete Composites, 2008, Vol. 30, pp.679-685.
[4]
J.M. Manso, J.A. Polanco, M. Losanez, and J.J. Gonzalez : Durability of concrete made with
EAF slag aggregate, Cement and Concrete Composites, 2006, Vol. 28, pp.528-534.
[5]
J. Geiseler : Use of steelworks slag in Europe, Waste Management, 1996, Vol.16, pp.59-63.
[6]
H. Motz and J. Geiseler : Products of steel slags an opportunity to save natural resources,
Waste Management, 2001, Vol.21, pp.285-293.
[7]
United Nations Framework Convention on Climate Change, Copenhagen Accord. December,
2009.
[8]
United Nations Framework Convention on Climate Change, Kyoto Protocol. December, 1997.
[9]
S. S. Jung and I. Sohn, “Crystallization Behavior of the CaO-Al2O3-MgO System Studied with
a Confocal Laser Scanning Microscope,” Metall. Mater. Trans. B., 2012, Vol. 43B, pp.1530-1539.
[10]
S. S. Jung and I. Sohn, “Effect of FeO Concentration on the Crystallization of High-
Temperature CaO-Al2O3-MgO-FeO Melts,” J. Am. Cer. Soc., 2013, Vol. 96, pp.1309-1316.
[11]
Y. Kashiwaya, C. Cicutti, A. Cramb and K. Ishii, “Development of double and single hot ther-
mocouple technique for in situ observation and measurement of mold slag crystallization”, ISIJ Int.,
1998, Vol. 38, pp.348-356.
[12]
Concrete structures, Korea Concrete Institute Pub., 1992, ISBN: 89-7086-091-693540.
M.C. Provance-Bowley and S.R. Miranda
GLOBAL OPPORTUNITIES OF STEEL-MAKING SLAG MATERIALS AS A
SOURCE OF SILICON-BASED FERTILIZERS
Harsco Metals and Minerals North America, Office of the CTO, 359 North Pike Road,
Saver, Pennsylvania, USA
Abstract
Acid soils represent a major crop production constraint contributing to nutrient deficiencies, metal toxicities, lower microbial populations, poor soil tilth and unfavorable
air-to-water ratios. Acidic soils adversely affect crop production worldwide impacting
nearly 50% of the world’s potentially arable soils.
Liming of soils for crop production is common practice in the Northeastern United
States where highly leached, acidic soils formed under forest vegetation. Within the
past 16 years farmers and turf managers in this region, have been substituting calcium and magnesium silicates, by-products of the area’s steel production industry, for
lime.
Research comparing calcium and magnesium silicates with either calcitic or dolomitic
lime have confirmed similar effects in neutralizing soil acidity. However, increased
yields and increased resistance to both abiotic and biotic stresses have also resulted
from increased plant silicon uptake with silicate additions. Positive crop responses to
calcium and magnesium silicate additions have also been seen in other regions of
the world where the dominant soil types are not only acidic, but highly weathered and
desilified.
The recent designation of silicon as a “plant beneficial substance” by mainstream
science provides an opportunity for beneficial uses of slag in the fertilizer industry,
not only in the United States, but globally.
Introduction
Acidic soils adversely affect crop production impacting approximately 50% of the
world’s potentially arable soils [1]. Acidic soils can limit crop growth and production
by reducing nutrient availability, Calcium (Ca), Magnesium (Mg), Phosphorus (P),
and Molybdenum (Mo); increasing toxic metal solubility, Aluminum (Al), Manganese
(Mn), Zinc (Zn), Copper (Cu), Cadmium (Cd), Nickel (Ni), and Lead (Pb); decreasing
beneficial microbial populations, e.g. rhizobia; limiting overall root growth, destabilizing soil aggregates, and reducing water infiltration rates [2,3]. Soil pH is often referred to as the master variable due to its effects on chemical, biological, and physical processes of the soil [4].
Soil Acidification
Acidification processes in soils are highly complex, a product of cropping, rainfall,
(leaching and erosion) removal of bases, acidic parent material and organic matter
decay where cations are removed more rapidly than parent material weathering replacement [2, 5, 6]. Overtime, when water events (rainfall or irrigation) exceed evapotranspiration (ET), basic nutrient cations [e.g. (Ca), (Mg), sodium (Na) and potassium (K)] that help in maintaining soil pH, leach from the soil profile [7]. Exchange sites
emptied and leached of nutrient cations are then preferentially occupied by free H+
and Al3+ ions released from the internal soil matrix and solubilized from parent materials by the acidification processes [8]. In separate reactions, the level of hydrogen
ions (H+) increase as water reacts with carbon dioxide (CO2) trapped in the soil, depicted in the following reaction [7]…
CO2 + H2O → H2CO3 → HCO3 - + H+
(1)
In addition to Carbon (C), nitrogen (N) and sulfur (S) cycling and transformation are
considered to be the main causes of acidification in agricultural soils with low buffering capacity [7].
Free H+ and Al3+ ions bind with reactive sites on soil particles and organic matter creating either neutral or positively charged surfaces [9]. This prevents base cations
(positively charged) from attaching to these surface locations, thus reducing the base
saturation of CEC sites in the soil profile [7].
Soil pH is actually an index of active acidity -- the acidity present in the soil water [4,
6]. Of greater importance in many cases is the reserve (exchangeable) acidity, the
amount of H+ and Al3+ cations occupying exchange sites [4, 6, 10]. Reserve acidity
impacts active acidity, and acts to replenish it when liming materials neutralize H+
ions and precipitate Al from the soil solution [11].
A good management program to
address soil acidity, therefore, must ensure that the soil conditioner that is applied to
the acidic soil is capable of neutralizing both active and reserve acidity [4].
In the Eastern United States, a majority of the region’s soils are mildly to strongly
acidic [12]. Crop production within this region, however, is not limited to acid loving
crops (e.g. blueberries and cranberries) but includes a wide variety of tree fruits,
grains, legumes, vegetables, and turf [13, 14]. As many of these soils formed under
forest vegetation and are highly leached in base cations, liming is common practice
every 2 to 4 years to maintain soil pH levels within a range 0f 6.0 to 7.0 conducive to
the growth of most crops [14].
Silicates as Liming Materials
Within the past 16 years farmers and turf managers within this region, substituting
Harsco’s calcium and magnesium silicates for lime, have seen increased yields and
healthier crops in addition to reductions in soil acidity [Harsco reports]. However,
using slags as liming materials in the United States is not new, being reported as far
back a ~1925 [15]. Research comparing calcium and magnesium silicates with both
calcitic and dolomitic lime have confirmed that both increased yields and disease
suppression can result from increased plant silicon (Si) uptake [16]. A positive crop
response to calcium and magnesium silicates is expected to be seen in this region as
the dominant soil type, Ultisols, are acidic, highly weathered, and desillified [14, 16].
Desilication can result in reductions in soil solution Si levels to the point of deficiency
for crops grown in areas of high rainfall [17].
Over time, desilication predicts that
highly weathered soils will be linked with infertility [17]. It is estimated that between
210 and 224 million tons of Si are removed from the soil globally each year from crop
removal only [18].
The calcium cation constituent of calcium silicate functions similarly to the calcium
cation found in liming materials [15], depicted in reactions (2 and 3) below [19]. They
both act on active acidity, disassociating (desorbing) H+ and Al3+ cations from the surfaces of soil particles and soil organic matter [4]. Both silicate and carbonate liming
materials have been shown to alter soil pH similarly when applied at the same calcium carbonate equivalence (CCE) rate [20]. However, the silicate and carbonate anions function in different ways.
CaCO3+ 2H+ → H2O + CO2+ Ca+2
(2)
CaSiO3 + 2H+ → H2O + SiO2 + Ca+2
(3)
In addition to having its Ca cation displace Al3+ and other metals from the surface of
soil particles and soil organic matter, the silicate anion forms monosilicic acid [21]
which can complex with Al and other metal cations on exchange sites — fixing potentially toxic metals [18]. This represents a significant addition to the performance
spectrum of calcium silicate and is a process not found in calcium carbonate liming
materials.
In acid soil, phosphorus in soil solution becomes adsorbed to surfaces of iron and
aluminum hydrous oxides and clay minerals [21]. When this occurs, the phosphorus
is considered “fixed,” immobilized and unavailable to the plant. Investigations of
monosilicic acids on soil properties confirm that they can increase the quantity of
mobile phosphates in the soil and soil solution [18, 21].
Monosilicic acids are reported to complex with aluminum hydroxide, creating hydroxyaluminosilicates (HAS) [18, 22]. Polymers form structural “bridges” between soil
particles and between soil particles and organic matter [18].
Silicon is released from calcium silicate into the soil’s bulk solution and is absorbed
by plants as Si(OH)4 where it is involved in the diverse structural and dynamic aspects of plant life and the performance of crops [23, 24]. Although not considered an
essential element for plant growth and development, silicon is considered a beneficial
element and is reported as being very useful where plants are subjected to abiotic or
biotic stresses [23, 25].
Harsco’s Global Presence
A calcium and magnesium silicate based processed slag, known as Reclime, was
first sold in the United States as a liming material in 1993. Following Harsco’s acquisition of Excel Minerals, this demetalized stainless steel slag co-product was renamed AgrowSil™ and continues to be sold. Harsco has been producing a similar
product (AgroSilcio®) in Brazil since 2001. Although Harsco processed slag is also
sold in South Africa, the product known as Calmasil™ is sold by another supplier.
Recently Harsco began producing pelletized calcium and magnesium silicate based
fertilizer products (Crossover™) in the United States and is set to begin production
and sales of similar silicon based fertilizer products in China and India in 2014.
Conclusion
The recent designation of silicon as a “plant beneficial substance” by mainstream
science [25] should provide additional opportunities for the beneficial uses of slag in
the fertilizer industry, not only in the United States, but globally.
References
[1]
von Uexküll H.R., and E. Mutert. 1995. Global extent, development and
eco-
nomic impact of
acid soils. pp. 5–19. In: R.A. Date, N.J. Grundon, G.E.
Raymet, and
M.E. Probert (eds.) Plant–Soil interactions at low pH: Prin-
ciples and
management. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
[2]
Foy, C.D. 1984. Physiological effects of hydrogen, aluminum, and manganese
toxicities in acid soil. pp. 57-97. In: F. Adams (ed.) Soil acidity and liming. 2nd
ed. Agronomy Monograph #12. ASA-CSSA-SSSA. Madison, WI. USA.
[3]
Haynes, R.J., and R. Naidu. 1998. Influence of lime, fertilizer and manure
applications on soil organic matter content and soil physical conditions: a review. Nutr. Cycl. Agroecosystems. 51(2):123-137.
[4]
Uchida, R.S., and N.V. Hue. Soil acidity and liming. pp.101-111. In: J.A. Silva
and R.S. Uchida (eds.) Plant nutrient management in Hawaii’s soils: Approaches for tropical and subtropical agriculture. College of Tropical Agriculture and
[5]
Human Resources, Univ. Hawaii Manoa, Honolulu, HI USA.
Johnson, G.V., and H. Zhang. 2004. Cause and effects of soil acidity. Oklahoma Coop. Ext. Svc. F-2239. Oklahoma State Univ. Stillwater, OK USA.
[6]
Mclean, E.O. 1982. Soil pH and lime requirement. pp. 199-224. In: A. L. Page
(ed.) Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties. Agronomy Monograph #9.2. ASA-CSSA-SSSA. Madison, WI. USA.
[7]
Bolan, N. S., D. C. Adriano and D. Curtin. 2003. Soil acidification and liming
interactions with nutrient and heavy metal transformation and bioavailability.
Advances in Agronomy 78:215-272.
[8]
Obura, P.A. 2008. Effect of soil properties on bioavailability of aluminum and
phosphorus in selected Kenyan and Brazilian acid soils. Doctoral Dissertation.
Purdue University. West Lafayette, Indiana USA.
[9]
Crozier, C., and D.H. Hardy. 2003. SoilFacts: Soil acidity and liming for
agri-
cultural soils. AG-439-50. NC Coop. Ext. Svc. Raleigh, NC USA.
[10]
Wutscher, H. K. 1997. Soil acidity and citrus blight. Comm. Soil Sci. Plant
Anal. 28(6-8):603-612.
[11]
Farina, M. P. W., P. Channon, and G.R. Thibaud. 2000. A comparison of
strategies for ameliorating subsoil acidity II. Long-term soil effects. Soil Sci.
Soc. Amer. J. 64(2): 652-658.
[12]
The Board of Regents of the University of Wisconsin System. Atlas of the
Biosphere. 2002. Center for Sustainability and the Global Environment
(SAGE), Nelson Inst. Environ. Stud. Univ. Wisconsin, Madison, WI USA. available.
online:
<http://www.sage.wisc.edu/atlas/maps.php?datasetid=20&includerelatedlinks=
1&
[13]
dataset=20> accessed 22 Sept. 2013.
Abaye, A.O., T.J. Basden, D. Beegle, G.D. Binford, W.L. Daniels, S.W. Duiker,
G.K. Evanylo, K.C. Haering, D.J. Hansen, G. Mullins, and R.W. Taylor. 2006.
Mid-Atlantic Nutrient Management Handbook. Mid-Atlantic Regional Water
Program.
[14]
Lathwell, D. R. (1984). Crop response to lime in the northeastern United
States. In: F. Adams, Soil Acidity and Liming 2nd ed. (pp. 305-332). Madison,
Wi: ASA-
[15]
CSSA-SSSA.
Barber, S.A. 1984. Liming materials and practices. pp. 200-209. In: F. Adams
(ed.) Soil acidity and liming. 2nd ed. Agronomy Monograph #12. ASA-CSSASSSA. Madison, WI. USA.
[16]
Heckman, J.R. 2012. Silicon and soil fertility. The Soil Profile. v. 20. Rutgers
Coop. Ext. Plant Biol. Pathol. Dept. Rutgers Univ. New Brunswick, NJ USA.
[17]
Fox, R.L., N.V. Hue, R.C. Jones, and R.S. Yost. 1991. Plant-soil interactions
associated with acid, weathered soils. Plant-Soil Interactions at Low pH
Developments in Plant and Soil Sciences. 45:197-204.
[18]
Matichenkov, V.V., and E.A. Bocharnikova. 2001. The relationship between
silicon and soil physical and chemical properties. Studies in Plant Sci. 8:209219.
[19]
Schererm H.W., and Mengel, K. (2007). Fertilizers. pp. 19. In Ullmann's Agrochemicals Vol. 1. John Wiley and Sons. Weinheim.
[20]
Heckman, J., Johnston, S., & and Cowgill, W. (2003). Pumpkin yield and dis
ease response to amending soil with silicon. HortScience, 38, 552-554.
[21]
Owino-Gerroh, C., and G.J. Gascho. 2004. Effect of silion on low pH soil
phosphorous sorption and on uptake and growth of maize. Comm. Soil Sci.
35:2369-2378.
[22]
Panov, N.P., N.A. Goncharova, and L.P. Rodionova. 1982. The role of
amorphous silicic acid in solonetz soils processes. Vestnik Agric. Sci. 11:18.
[23]
Epstein E., 1999. Silicon. Annu. Rev. Plant Physiol. Plant Mol. Biol. 50:641664.
[24]
Ma, F.A., and N. Yamaji. 2006. Silicon uptake and accumulation in higher
plants. Trends Plant Sci. 11(8):392-397.
[25]
AAPFCO. 2012. Uniform State Fertilizer Bill, Rules and Regulations (2)(f). Of
ficial Terms T-73.
Theme 5
Environmental Affairs E Poultney(1), N Ghazireh(2), N Jones(3), C Laskey(4), M Davies(5), P Redfern(5) and
J Barritt(6)
Developing a Quality Protocol for Steel Slags
1.
Tata Steel R,D&T, Teesside Technology Centre, Middlesbrough, TS6
6US, UK.
2.
Lafarge Tarmac, Millfields Rd, Ettingshall, Wolverhampton, WV4 6JP,
UK.
3.
Harsco Metals Group Ltd., Bradmarsh Way, Rotherham, S60 1BW, UK.
4.
Celsa Steel UK Ltd., East Moor Road, Cardiff, CF24 5NN, UK.
5.
Environment Agency, Horizon House, Deanery Rd, Bristol BS1 5AH,
UK.
6.
WRAP, The Old Academy, 21. Horse Fair, Banbury, OX16 0AH, UK.
Abstract
Quality Protocols are currently being developed in the UK for certain selected materials including steel slags, incinerator bottom ash (IBA) and biomethane. These documents are being developed jointly between the Environment Agency, Waste & Resources Action Programme and the related industries. They set out end of waste
criteria so that secondary materials can be fully recovered as products and be used
without undermining the effectiveness of the Waste Framework Directive. The slag
industry in the UK agreed to support, without prejudice to its assertions that steel
slags are by-products, the development of the quality protocol for steel slags so as to
demonstrate that steel slag products are fit for purpose.
Before it can be determined whether a quality protocol can be developed for the steel
slags selected in this study, a considerable amount of evidence needs to be gathered. The Quality Protocol process begins with a detailed Financial Impact Assessment. This defines the market certainty and financial benefits of a material being fully
utilised in specific identified applications. A full risk assessment is then undertaken
for each end use, which includes chemical analyses and leachate testing to ascertain
any risks to health and the environment. If successful this stage is then followed by
the development of a Technical Report and Quality Protocol which are put out for
public consultation in the UK. Following the consultation and subsequent revision,
the Quality Protocol document is then sent to the EU Commission for further consultation before final revision and publication.
This paper outlines the assessment process that selected steelmaking slags have
gone through to date towards the development of a Quality Protocol. The paper also
presents a summary of the outcome from the risk assessment carried out on steel
slag.
1. Introduction
The Waste Protocols Project (WPP) is a joint Environment Agency and WRAP
(Waste and Resources Action Programme) initiative, funded by the Department for
Environment Food and Rural Affairs (Defra), the Welsh Assembly Government and
the Northern Ireland Environment Agency as a business resource efficiency activity.
The aim of the project is to provide clarity on when specific waste streams cease to
be waste and where sufficient evidence exists to produce an end of waste Quality
Protocol (QP). Quality Protocols define the point of full recovery from a waste into a
product or material that can be either reused by the business or industry or sold into
other markets.
Part of the evidence needed for the QP process and to make an end of waste decision is an assessment of the potential risk of harm to human health, animal health
(where relevant) and the environment posed by the handling and use of the waste
derived product.
2. Financial Impact Assessment
In the UK, steel slags have been used in construction products and agriculture application as fertilisers for over half a century. In 2006, the Environment Agency and the
Waste and Resources Action Programme (WRAP) initiated the project on developing
the Quality Protocol for steel slags. As a first stage a Financial Impact Assessment
was undertaken to identify cost benefits associated with a Quality Protocol.
The quality protocol is necessary to avoid government intervention in the form of
waste regulations on steel slag that the Environment Agency may or may not enforce. It will allow the free operation of the market for steel slag by falling outside of
scope of waste regulations in which steel slag was never intended to be waste in the
first place.
The FIA identified the costs and benefits including carbon savings of introducing a
Quality Protocol by comparing two options:
 Option A: Business as usual for the Slag Industry and the Environment
Agency would clarify the current regulatory position specific to steel slag.
 Option B: Introduce a Quality Protocol.
Option A included the projected costs of Regulation including costs associated with
steel slag being marketed as waste and not by product
Option B will provide increased assurance to customers and ensure that steel slags
are recovered if produced in accordance with the Quality Protocol.
Over the period to beyond 2020, cost savings associated with the recovery of materials to the Slag Industry of £3-8 Million per year were identified. Similarly cost savings
to end users of £0.8-2.5 Million per year mainly from lower embodied carbon content.
3. Quality Protocol Project
The Quality Protocol is a formalised quality control procedure. It has three main purposes:
i. Providing users with confidence that the steel slag they purchase conforms to
agreed quality and product standards comparable with materials of a primary source;
ii. Protecting the environment and human health – by setting standards for the use of
steel slag in market applications.
iii. Easing the regulatory burden on industry.
During the early stages of the Project, the Technical Advisory Group developed the
sampling and testing regimes for the Risk Assessment. It was identified early on in
the project that steel slags in bound applications including asphalt, concrete and hydraulically Bound Mixtures (HBM) have very limited potential to leach. Therefore it
was accepted by the Environment Agency in the UK that the use of steel slags in
bound applications that meet the relevant engineering standards and specifications
are recovered once they are mixed with the binders. Consequently the Risk Assessment concentrated mainly on the unbound applications for Steel Slag but also included the use of steel slag in surface dressing (a bound application). The use of steel
slags in agriculture as fertiliser, known as Agslag will now be assessed at a later
stage. Samples of Steel Slag in size gradings similar to their end use were collected
for Upflow Percolation Leachate Testing at DHI in Denmark. Anonymous Limestone
samples were also tested as comparators. Compositional Analysis of samples had
also been done by aqua regia digestion.
The aim of the Risk Assessment was to determine whether the properties of steel
slags could present potentially unacceptable risks to the environment or human
health through a range of unbound applications. The outcomes of the risk assessment will be used to inform a decision about whether a Quality Protocol can be developed for steel slags.
4. Risk Assessment
The risk assessment [1] outlined in this paper has identified reasonable use scenarios in which steel slags, including Basic Oxygen Steelmaking slag (BOS), and Electric
Arc Furnace slag (EAF) are used. Chemical data for steel slags from UK sources
have been collected since 2006. These data were synthesised for use in assessing
potential chemical risks to groundwater, surface water and human health from reasonable worst case use scenarios.
For groundwater, the assessment, including modelling, indicates compliance to 100
years for all substances of interest. The assessment suggests that an impact on
groundwater quality from steel slag used as unbound capping layer or sub-base in
the modelled scenario is unlikely. Further assessment of chromium(VI), antimony,
cadmium and selenium indicate similar leaching behaviour of BOS to natural limestone aggregate. The capping scenario afforded the lowest dilution factor in the Level
2 assessment. As the pathway in the unsaturated zone will be the same for all scenarios this was modelled as the most sensitive scenario. It is therefore concluded
that all scenarios examined would pose little risk to groundwater up to 100 years.
For Level 3 the data are filtered to include only material types that would be used in a
road sub-base / capping scenario. A Level 3 assessment was done for the following
elements; Al, V, Sb, As, Ba, Cd, Cr(VI), Co, Pb, Hg, Ni, Se and Tl. A constant source
model is set up for aluminium and vanadium as declining behaviour is not demonstrated to LS 2 (about 60 years, and LS 10 in validation samples). Both are nonhazardous substances and compliance is assessed 50 metres downgradient of the
source. A declining source model is set up for the other substances listed above.
Compliance was assessed 50 metres downgradient of the source for non-hazardous
substance (barium) and at the base of the unsaturated zone and one metre downgradient for hazardous substances (cadmium and mercury) and interim hazardous
substances (antimony, arsenic, chromium(VI), cobalt, lead, nickel, selenium and thallium). ConSim software was used to run Monte Carlo simulations to model potential
leaching to Groundwater. Table 1 shows a summary of the ConSim Level 3 output for
the selected elements.
GWS
Substance
10 yrs
(μg l-1)
30 yrs
100 yrs
500 yrs
1000 yrs
Concentration at compliance point (μg l-1)
Non-hazardous substances
Aluminium
200
Barium
700
Vanadium
20
0
0
0
151
187
(0)
(0)
(309)
(664)
(664)
0
0
0
0
0
(0)
(0)
(0)
(0)
(0)
0
0
0
0
0
(0)
(0)
(0)
(45.6)
(145)
(Interim) hazardous substances
Antimony
0.5
Arsenic
1.0
Cadmium
0.1
Chromium(VI)
2.0
Cobalt
0.3
Lead
2.5
Mercury
0.01
Nickel
2.0
0
0
0.
0.006
0.13
(0)
(0)
(1.0)
(1.2)
(1.2)
0
0
0
0
0
(0)
(0)
(0)
(0.70)
(0.35)
0
0
0
0
0
(0)
(0)
(0)
(0.25)
(0.89)
0
0
0
15.7
16.9
(0)
(131)
(107)
(216)
(374)
0
0
0
0
0
(0)
(0)
(0)
(0)
(0)
0
0
0
0
0
(0)
(0)
(0)
(0)
(10.7)
0
0
0
0
0
(0)
(0)
(0)
(0)
(0)
0
0
0
0
0
(0)
(0)
(0)
(0)
(0)
GWS
Substance
10 yrs
(μg l-1)
Selenium
1.0
Thallium
0.96
30 yrs
100 yrs
500 yrs
1000 yrs
Concentration at compliance point (μg l-1)
0
0
0.003
0.36
0.36
(0.99)
(2.7)
(3.9)
(3.5)
(4.8)
0
0
0
0
0.11
(0)
(0.5)
(0.9)
(0.8)
(0.77)
GWS = Groundwater Standard
Table 1: ConSim Level 3 output summary
The results are output over 5 timeslices: 10, 30, 100, 500 and 1,000 years. These
timeslices represent:
10 & 30 years – within the range of the useful life of a road sub-base or capping layer
100 years – emission limits for construction products under the Dutch Soil Quality
Decree derived from maximum concentration in groundwater 1 metre into the water
table beneath a source.
500 years – timescale for permanent applications under the German Recycling Decree.
1,000 years – maximum end-point for assessment, particularly for substances modelled as a constant source.
For a temporary structure such as a road, the recommended timescale for evaluation
is taken as 100 years; consistent with the Dutch Soil Quality Decree. The results indicated that 90% of the iterations are compliant with the groundwater standard for all
substances up to 100 years.
Maximum concentrations (highly conservative) only exceed groundwater standards
at 100 years for aluminium, antimony, chromium(VI) and selenium. At 500 and 1,000
years, only the 90th percentile values for chromium(VI) exceeds its iMRV, as highlighted by the shaded cells in Table 1. The iMRV is exceeded at just over 200 years.
The only other substances that may merit further discussion due to sensitivity analysis on the basis of maximum values at 100 years are aluminium, antimony and selenium, with cadmium added due to its status as a hazardous substance and exceedence of the MRV at 500 years (maximum output concentration).
For surface water, the generic risk assessment initially identified potential risks when
10% of the Environmental Quality Standards was used in a face-value comparison at
Level 1 for numerous metals. When scenario-specific surface water dilution is accounted for in the assessment at Level 2 there are no potential risks identified from
the use of steel slags for road embankments, unbound capping, unbound sub-base
materials and pipe bedding.
The human health workplace scenario risk assessment for use of steel slags indicates an absence of risk for the majority of contaminants provided that inhalable dust
is kept below the legally permissible limit. A further detailed assessment undertaken
by the Environment Agency air pollution team on the potential risks posed to the
general public by the use of steel slags in construction activities, especially from fugitive dust emissions, noted very low levels of risk. The conclusions from this risk assessment are applicable to the use of steel slag as unbound materials in groundwork
and construction activities.
The risk assessment identified reasonable use scenarios in which steel slags including Basic Oxygen Steelmaking slag (BOS) and Electric Arc Furnace slag (EAF) are
used. The main conclusions from the assessment are listed below, along with recommendations made.
5. Groundwater Risk Assessment Conclusions & Recommendations
For groundwater the assessment, including modelling, when both dilution and attenuation are accounted for and a declining source term used, compliance with the relevant groundwater standard at 100 years is indicated for all potential substances of
concern using 90th percentile output concentrations. For the maximum output concentrations, groundwater standards are exceeded for aluminium, antimony, chromium(VI) and selenium but this is considered to be conservative and provided only for
sensitivity analysis.
The other scenarios were re-evaluated having regard to the type of materials used
and it was concluded that the level 3 outputs may not be conservative for the pipe
bedding scenario for antimony, cadmium, chromium(VI), cobalt and nickel. Additional
modelling using a source term defined by the 4/10 mm BOS and EAF data showed
compliance of the 90th percentile output value with the relevant groundwater standards (iMRVs).
Additional modelling work for chromium(VI) involved filtering the data set and comparison of impact with a non-waste aggregate, limestone. 90% of the outputs are below the iMRV across all timeslices for all but CS EAF (above the iMRV at 500 and
1,000 years). Although compliant at 100 years, there is some sensitivity in using CS
EAF on the basis of both 500 year concentration and the 100 year maximum concentration at the receptor. However, any risk is considered to be low and weathering carried out to improve geotechnical properties (swelling) would reduce the risk even further.
For the non-waste comparator (limestone), the 90th percentile concentrations are below the iMRV at all timeslices. The limestone shows broadly comparable results to
BOS and with slightly lower maximum concentrations than weathered CS EAF. It is
therefore concluded that none of the materials modelled are likely to impact groundwater at 100 years from the leaching of chromium(VI).
The summary statistics for leached concentrations from limestone were compared
with the steel slag materials for aluminium, antimony, cadmium and selenium. This
comparison shows slightly higher leached concentrations of aluminium from steel
slags, equivalent levels for cadmium and slightly higher leaching of antimony and
selenium from limestone.
The scenario modelled at Level 3 afforded the lowest dilution factor in the Level 2
assessment and represents the most sensitive scenario with respect to groundwater
impacts. The results presented here suggest that based on 100 years, an impact on
groundwater quality from steel slags used in the developed scenarios is unlikely.
6. Surface Water Conclusions
For surface water the generic risk assessment initially identified potential risks when
10% of the EQS was used in a face-value comparison at Level 1 for numerous metals. When scenario-specific surface water dilution is accounted for in the assessment
at Level 2 there are no potential risks identified from the use of steel slags for road
embankments, road capping, pipe bedding and surface dressing.
7. Human Health
The workplace scenario risk assessment for use of steel slags indicates an absence
of risk for the majority of contaminants provided that inhalable dust is kept below the
legally permissible limit. However, based on 90th percentile composition data, the
hazard quotient for manganese is 1.0 which indicates a potential risk due to high end
concentrations of this contaminant in steel slag that may require further consideration. The conclusions from this risk assessment are applicable to all workplace scenarios including the use of steel slag in groundwork and construction activities.
Assessment of the risks posed to nearby residents during use of steel slag in
groundworks and construction projects calculates that 1% of the long term Environmental Assessment Level (EAL) is exceeded for the majority of contaminants present
in steel slag, but the short term is considered more appropriate as construction activities will generally only take place over a limited time period, that high levels of dust
will be intermittent, and that dust will not always be blown in the same direction away
from the site using steel slag. A predicted airborne concentration of 10% of the short
term EAL is taken to be indicative of an insignificant contribution from the process
under consideration. For those contaminants for which a short term EAL is available,
only vanadium is highlighted as posing a potential risk although this conclusion is
questionable as the short term EAL is lower than that recommended for long term
assessment. Significant though, is the fact that there are no longer short term EALs
available for the more hazardous contaminants present in steel slag, including arsenic, cadmium, cobalt, chromium (VI), lead and nickel (short term EALs are no longer
recommended for these compounds in the latest version of the H1 guidance on environmental risk assessment; EA, 2010). An additional assessment undertaken by the
air pollution team at the Environment Agency specifically looking at exposures via
dust identified no potential human health risks.
8. Conclusions
Following the Risk Assessment, a Technical Report on Steelmaking Slags together
with the Quality Protocol will be produced for public consultation within the UK. The
UK consultation should take place during 2013 and if no amendments are required
the Quality Protocol will be passed to the European Commission for further consultation in 2014.
9. References
[1] Steel slag quality protocol: Chemical risk assessment on BOS and EAF, WCA
Environment Ltd., Environment Agency Document 2013
Jérémie DOMAS
Sustainable reuse of iron and steel slags in road applications Technical requirements for environmental acceptance in France
C.T.P.L. – Technical and Promotional Centre for Iron and Steel Slags
Site ArcelorMittal – Aile 1, Bureau 120 – 13776 Fos-sur-Mer cedex – France
Abstract
In France, production of iron and steel slags represents about 5 millions tons per
year (2012). This production can be separated in three main categories: blast furnace
slags, basic oxygen furnace steel slags and stainless steel and carbon steel electric
arc furnace slags. 75% of iron and steel slags are reused mostly in construction
works, other applications as industrial raw materials, internal site recycling, storage
or landfilling being more marginal.
This situation and the lack of explicit national or European regulatory framework governing their uses have led the iron and steel industry, under the coordination of the
Technical and Promotional Centre for iron and steel slags (CTPL), together with the
French Administration (ADEME and Ministry in charge of Environment), to launch in
2006 a harmonised approach to reuse slags in road applications in safe conditions.
This national initiative was based on a comprehensive technical study for testing different reuse scenarios with the various categories of slags produced all around the
French territory.
Since 2010, and in close cooperation with the French administration, the French iron
and steel industry drafted practical guidelines to define environmental acceptance for
the reuse of slags in road works. These technical guidelines are specifically dedicated for end-users, and can be related with the criteria and the general methodology
formerly developed by the French administration for valorisation of materials from
secondary sources, under waste status (SETRA guidelines, March 2011).
Practical and dedicated guidelines for iron and steel slags were published in France
in October 2012.
Keywords: steel slags, reuse, road, environmental acceptance, specifications
1. Introduction
In France, production of iron and steel slags represents about 5 millions tons per
year (2012). This production can be separated in three main categories: blast furnace
slags, basic oxygen furnace steel slags and stainless steel and carbon steel electric
arc furnace slags. 75% of iron and steel slags are reused mostly in construction
works, other applications as industrial raw materials, internal site recycling, storage
or landfilling being more marginal [1].
Technical requirements are given by national and European standards for either hydraulic binders (EN 13282), concrete (EN 206-1, EN 15167), aggregates
(EN 12620, EN 13043, or EN 13242) or for road materials (EN 14227-2), and provide
specific elements on “how to use” iron and steel slags in public works and construction fields in order to reach suitable technical performance for the intended purpose.
However, neither standards, nor national or European regulatory framework give
clear and explicit indications concerning chemical and/or environmental specifications.
This situation and the lack of explicit national or European regulatory framework governing their uses have led the iron and steel industry, under the coordination of the
Technical and Promotional Centre for iron and steel slags (CTPL), together with the
French Administration (ADEME and Ministry in charge of Environment), to launch in
2006 a harmonised approach to reuse slags in road applications in safe conditions.
This national initiative was based on a comprehensive technical study for testing different reuse scenarios with the various categories of slags produced all around the
French territory [2] [3]. Batch and up-flow percolation leaching tests were also completed in 2011 to propose environmental thresholds in compliance with the French
national reference Guidelines for acceptability of alternative materials in road construction (SETRA, 2011) [4].
The present paper presents contents of the dedicated SETRA guidelines for environmental acceptance of iron and steel slags in road applications, published in October 2012 [5].
2. Aim and structure of the document
The aim of these practical application guidelines is to provide a solid reference frame on
which technical personnel could base project design or appraise alternatives proposed within a call for tender. In this way, it
is mainly intended for civil engineering professionals.
This reference document details also the
respective duties of the different road project
stakeholders to ensure the memorisation of
road construction projects, which have resorted to the use of iron and steel slags.
The structure of these guidelines is divided in five main parts:
o definitions and terminology,
o a brief description of various slags suitable for use in public works: origin,
steelmaking process and main figures,
o a brief description of the manufacturing process of the alternative material, as
well as of the road material,
o prescribed road applications and associated technical specifications and limitations,
o a quality insurance procedure, including field compliance verification and
memorisation of road construction projects
3. Terminology and main descriptions
Main definitions about “alternative material”, “road material” and “road usage” are
defined in the French general guidelines (SETRA, 2011) [4]. Complemented specific
definitions about “iron and steel slags”, “manufacturing”, “producer”, “manufacturer”
and “seller” (i.e. main stakeholders) are also given to help the reader having a better
understanding of the document [5].
The different types of slags are then briefly described, giving to the reader general
information about steelmaking processes (Figures 1 and 2), quantities and geographic repartition, and typical chemical composition.
Figure 1. Blast furnace (Arcelormittal)
Figure 2. carbon EAF slag poured from
the furnace (Ascométal)
Thus, blast furnace (BF), basic oxygen furnace (BOF), carbon and stainless electric
arc furnace (cEAF and sEAF) steelmaking processes and the resulting slags are described in the document.
Manufacturing phases, briefly describing cooling process, physical and mechanical
treatment (crushing, screening, sorting, particle size reduction …) and physicochemical treatment covered by the term “maturation” of raw slags are also given in
the document.
The nature of the main alternative materials, based on the current available practices, is given for the different types of slags: sands and gravels 0/D or d/D are the main
common practices. Once manufactured, the alternative material is used alone and
without modification, or may be used in a mix with other materials (aggregates, fillers,
hydraulic or bituminous binders …) after a so-called “formulation” stage in one of the
road applications covered by the scope of the guidelines. Main road materials for
slags are unbound road-base layer, hydraulic bound mixtures, and road materials for
bituminous mixtures and surface treatments for roads, airfields and other trafficked
areas.
4. Road applications, specifications and limitations
This chapter lists the allowed road applications together with the attached limitations
of use, in relation with the environmental characteristics of the alternative materials.
Three main types of applications are defined:
o Road applications “type 1” : road base layers or shoulder sub-layers of capped
pavement48 : subgrade fill, capping layer, sub-base course, base course, subgrade or binder courses (figure 3).
Figure 3. Road applications “type 1” (surfaced applications)
o Road applications “type 2” : road applications for covered engineering embankments associated with road infrastructure (e.g. phonic protection) or for
capped shoulders49 (figure 4).
48
Applications must be surfaced with a surfacing layer considered impervious (asphalt, bituminous
mixtures, wearing surface dressings, cement concrete, binder-jointed paving blocks) with a 1% minimum gradient.
49
Applications must be covered by at least 30cm of natural materials (including topsoil) with a 5%
minimum gradient on the top of this cover, to limit water infiltration.
Figure 4. Road applications “type 2” (capped applications)
o Road applications “type 3” : road base layers of uncapped pavement or shoulder sublayers, road applications for uncovered engineering embankments associated with road infrastructure or for uncovered shoulders sub-layers, applications for preloading fills required for road infrastructure construction, wearing courses and surface treatments, applications for drainage systems (e.g.
trench drain, pavement of rainwater reservoirs…), and applications in construction of uncovered forest access roads or rural routes.
Figure 5. Road applications “type 3” (uncapped applications)
A decision grid (Table 1) has been built to indicate to the end-user what is possible or
not, in which conditions, and with the corresponding limitations for the local boundaries of the road work, and the associated recommendations applicable during the
work phase.
Environmental
Road applications
Local boundaries
Work phase
reference
Limitations
Limitations
Limitations
Limit values
Road applications
< reference 1
“type 1” < 3m
Road
work
shall
be :
- out of flood-threatened areas
- ≥ 50 cm above the highest
level of 50 years water mark
- minimum distance ≥ 30 m of
watercourses, lakes or ponds
Road applications
Limit values
“type 1” < 3m
< reference 2
Road applications
“type 2” < 6m
- out of the close protection
1000 m3 of temporary
perimeter around drinking water
storage capacity
supply
Beyond 1000 m3
area
- out of sensitive areas in
authorization
relation to aquatic environment
and
-
water
out
of
ressources
karst
areas
- out of National Parks
Or
hydrogeologist-expert
judgment
Road applications
“type 1” without
restriction
Road applications
“type 2” without
- Out of National Parks
restriction
Limit values
< reference 3
No limitations
Road applications
“type 3” and pH < 12
Road
Road applications
“type 3” and pH > 12
work
shall
be :
- minimum distance ≥ 30 m of
watercourses, lake or ponds
- out of the close protection
perimeter around drinking water
supply
area
- out of sensitive areas in
relation to aquatic environment
and
-
water
out
of
ressources
karst
areas
- out of National Parks
Or
hydrogeologist-expert
judgment
Table 1. Decision Grid for the technical specifications
Environmental reference thresholds are also given in specific tables provided in Annex A of the guidelines (Table 2). Those criteria were subject to the validation of the
French Ministry in charge of Environment, according with general methodology given
in SETRA Guidelines (March 2011) [4].
Table 2. Environmental limit thresholds depending of road applications
Setting of environmental thresholds has been carried out based on a risk exposure
scenario assessment, taking into account impact of alternative material usage on
groundwater. The aim was to determine the release at a structure outlet (source
term) ensuring compliance with a given groundwater quality hydraulically downstream of this structure (impact term). Further details are given in Annex 5 “modelling
principles applied” of SETRA methodological guidelines (March 2011) [4]. Limit values of general guidelines (Annexes 3 and 4) were adapted according to iron and
steel slags specific properties to define criteria of Annex A [5].
In case of a specific scenario, or when the environmental characteristics do not comply with the reference thresholds, specific studies may be performed and their tech-
nical results submitted to the acceptance of the relevant French Administration services.
5. Quality Insurance Procedure
The Quality Insurance Procedure details how to proceed for the verification of the
environmental compliance of the alternative materials.
The quality procedure defines the different tasks and duties of stakeholders (producer, manufacturer, seller …) throughout the supply chain. It is one of the main steps of
the methodology, where particular care has to be paid.
The environmental compliance assessment is described to provide recommendations
as to how to check the characteristics of the fabricated alternative materials, before
delivering and using them as road materials in the various allowed applications.
Regular representative samples of alternative material production are collected, in
compliance with a specific sampling procedure for which detailed recommendations
are given in technical annex (annex C) of the guidelines [5], based upon the best
available practices in this field (European guidelines for waste sampling is given by
EN 14899).
Environmental characteristics – pH, electric conductivity, As, Ba, Cd, Cr, CrVI, Cu,
Hg, Mo, Ni, Pb, Sb, Se, Zn, fluorides, chlorides and sulphides – are determined after
compliance batch leaching test according to EN 12457-4 standard, and at a set minimum periodic frequency, as proposed in Table 3.
Results are compared with limit values given in the annex and compliance with the
willing road applications is verified prior to accepting the use of the manufactured alternative material.
Capacity* of the
manufacturing
plant
Minimum fre-
Evolution of compliance assessment
quency
for compliance
Frequency**
List of parameters
If 12 consecutive samples
If 12 consecutive samples are
are all < limit values, pos-
all < (limit value)/2, possibility
sibility
to check only compliance as-
assessment
Plant
1 sample every
< 30 000 T/year*
three months
to
check
only
1 sample / 6 months
sessment for the parameter
once a year
Plant
1 sample
> 30 000 T/year*
every month
If 12 consecutive samples
If 12 consecutive samples are
are all < limit values, pos-
all < (limit value)/2, possibility
sibility
to check only compliance as-
to
check
1 sample / 3 months
only
sessment for the parameter
once a year
* Capacity in T/year assessed on the base of year n-1 for compliance assessment performed during year n.
** Only for plants using compliance assessment with a pre-requisite analysis before delivering and
using road materials in the road applications.
Table 3. Minimum requirements for compliance assessment
For practical reasons, manufacturing site plants are not always able to store individual and numerous batches on site, manufacturing, sale and delivering being carried
out on a continuous cycle. Consequently, it was important to have a simplified compliance procedure to determine whether it is possible, and under which conditions
(i.e. extent of the historical statistical data), the environmental compliance assessment can be done after the actual delivery of the materials and use in road applications. This simplified procedure is possible only if the material can demonstrate that
12 consecutive samples are – for all parameters – comply environmental limit values
for road applications scenario “type 3”. In this case, compliance assessment can be
carried out after delivery to the clients.
The Quality Insurance Procedure also proposes practical recommendations to ensure the traceability of use of the alternative materials manufactured from slags production. Template of a compliance certificate is given in one technical annex (Annex
D) of the guidelines [5], allowing stakeholders to assess the relevance of the intended road applications with the environmental characteristics throughout the supply
chain.
6. Conclusions and future prospects
General drafting of the operational application guidelines, for environmental acceptance of iron and steel slags in road applications is now achieved, and was published by SETRA in October 2012, under validation of the French Ministry in charge
of Environment.
Operational application is currently in progress with the steel industry, to help promoting slag reuse in road applications, improving recycling of sustainable resources and
going towards economy of natural resources
With the help of these operational guidelines, French authorities will be able to implement the current outpour of environmental European regulations related to construction products: Construction Product Regulation (CPR), REACh registration dossiers for slags, and “end of waste criteria” (EoW) included in the Waste Framework
Directive (WFD), and to provide safe, factual and pragmatic recommendations for
both producers and users of slag products.
For steel slags industry, these guidelines are a major evolution and open new opportunities and positive future.
References
[1]
Internet website of CTPL : Statistics for 2012
Blast furnace slags : http://www.ctpl.info/wp-content/uploads/2013/05/CTPLFFA-laitiers-HF-2012-2011.pdf
Steel
slags
:
http://www.ctpl.info/wp-content/uploads/2013/05/CTPL-FFA-
laitiers-acierie-2012-2011.pdf
[2]
Domas J.: Feed-back from French industrial developments on steel slags in
road-base applications (2009). WASCON Conference 2009.
[3]
Legret M., Chaurand P., Bénard A., Capowiez Y., Deneele D., Reynard J.,
Lassabatere L., Yilmaz D., Rose J., Domas J., Béchet B., Richard D., Bottero
J.-Y. : Environmental assessment of a BOF steel slag used in road construction: the ECLAIR research program (2009). WASCON Conference 2009.
[4]
SETRA (Technical department for transport, roads and bridges): Acceptability
of alternative materials in road construction - Environmental assessment.
(March 2011, translated February 2012)
http://www.setra.developpementdurable.gouv.fr/IMG/pdf/Acceptabilit_GB_Web.pdf
[5]
SETRA (Technical department for transport, roads and bridges): Acceptabilité
environnementale de matériaux alternatifs en technique routière – Les laitiers
sidérurgiques (Octobre 2012), Réf. 1226, 48 pp.
Charles Ochola, Ph.D., P.E
Legislation vs. Regulation in the USA
Tube City IMS, LLC, Horsham PA, USA
Abstract
A regulation is the method used by an empowered agency to implement the requirements of legislation. It can be targeted, such as an industry specific regulation or as
a regulation that is broad based in it scope. The use of slag in the United States is
for the most part regulated under solid waste rules. These rules vary from state to
state and in some instances are even handled on a county basis. This devolution of
regulatory control has resulted in a hodgepodge of rules and guidance that makes
the commercialization and utilization of slag complex and cumbersome. Furthermore, due to the tightening regulatory control of waste materials it is becoming more
difficult to use iron and steel slags in applications that have been well established.
Slags are therefore required to get an exemption from waste classification before a
beneficial use is allowed. Due to these challenges, the iron, steel, and slag industry
has formed a coalition to proactively address the concerns within the slag industry
with the primary objective being to get iron and steel slags recognized as products.
In order to achieve this, the goal is to have a recognized National / International
Standards organization adopt a specification designating Iron and Steel Slag as
products. This will require a thorough chemical and mineralogical characterization of
these materials in order to present legislators with compelling evidence including the
successful historical use of these materials that they are products and not wastes.
The principle way to achieve this is by legislation that totally removes them from the
waste category and treats them similarly to other comparable products.
Introduction
In the steel industry, two main types of slags are produced – Blast Furnace Slag and
Steel Slag. Slag is a co-product of the iron and steel-making process. As shown in
Figure 1 below, iron cannot be made in a blast furnace without the production of its
co-product, Blast Furnace Slag. Similarly, steel cannot be produced in a basic oxygen furnace (BOF) or in an electric arc furnace (EAF) without making its co-product,
Steel Slag.
Figure 1. Depiction of Two Products from a Blast Furnace
The American Society of Testing and Materials (ASTM) specifically define iron and
steel slag as follows:
Blast Furnace Slag is “the non-metallic product, consisting essentially of silicates and
aluminosilicates of calcium and other bases that is developed in a molten condition
simultaneously with iron in a blast furnace.”
Steel Slag is “a non-metallic product, consisting essentially of calcium silicates and
ferrites combined with fused oxides of iron, aluminum, manganese, calcium and
magnesium that are developed simultaneously with steel in basic oxygen, electric arc
and open hearth furnaces.”
The utilization of iron and steel slags has been predominantly in the construction and
cement industry as an aggregate raw material that can compete and in some cases
exceed natural aggregates in quality with typically better pricing. More recently, these products have shown great potential and are being used in agricultural and environmental applications. In the United States, the U.S. Geological Survey (USGS)
has been the major source of information on the Nation’s natural resources, providing
information about the production, and consumption of minerals from U.S. companies,
mines, and mineral- processing plants [1]. For more than 100 years, the USGS has
collected, processed, analyzed, and published this data, and as far back as the late
1930’s has tracked the production and consumption of iron and steel slags as commodities.
In many states the utilization of iron and steel slag materials most likely predates the
tracking statistics developed by the USGS and therefore, given its well documented
use and applications; it is ironic that the regulation of iron and steel slags typically
falls under the Bureau of Waste Management or similar agencies. This however may
be for a myriad of reasons; one chiefly being that in many areas the abundance of
natural aggregate and lack of foresight in past decades led to large stockpiles of iron
and steel slag materials. Now these stockpiled materials are being mined, recovering valuable construction aggregate and metallic material for use in steel making.
Nevertheless the value of this commodity has never been in question and as evidenced by USGS statistics, iron and steel slag value has been on the rise. The importation of iron and steel slags which started in the early 70’s has risen to the point
whereby the United States is now importing well over one million tons of these commodities annually.
Regulatory Environment
Unfortunately or fortunately depending on what the concern is, slag issues in the
United States are currently addressed individually on a state by state basis. This devolution of regulatory control has caused a hodgepodge of rules and guidance that
makes the commercialization and utilization of slag cumbersome. Table 1 below
summarizes the current regulatory status of iron and steel slags across the various
states that have or have had a significant iron and steel industry.
State
Category
Status
Arizona
Excluded from Solid Waste definition Statute
Indiana
Slag not regulated
Michigan
Excluded from Solid Waste definition Statute
Ohio
Excluded from Solid Waste definition Statute
Utah
Industrial by-product not solid waste
Statute
Alabama
Product
Statute 2010
Washington
By-product
Regulations
West Virginia
Excluded from Solid Waste definition Regulatory interpretation
Iowa
Beneficial use with some limitations
Pennsylvania
Beneficial use standing determina- Co-product determination
Statute
Regulations
tion
Mississippi
Not a solid waste
Case-specific
North. Carolina
Recovered Material
Case-specific
South Carolina
Recovered Material
Case-specific
Texas
Co-product
Case-specific
Arkansas
Case by case
Beneficial Use
Florida
Case by case
Beneficial Use
Georgia
Case by case
Recovered Material
Louisiana
Case by case
Beneficial Use
Table 1. State Status of Iron and Steel Slag (courtesy Nucor Steel)
State
Category
Status
Louisiana
Case by case
Beneficial Use
Minnesota
Case by case
Beneficial Use
Nebraska
Slag is a Product
Statute 2013
New Jersey
Case by case
Beneficial Use
Tennessee
Case by case
Beneficial Use
Virginia
Case by case
Beneficial Use
California
Case by case
Statute and Regulation
Colorado
Case by case
Statute and Regulation
Idaho
Case by case
Statute and Regulation
Illinois
Case by case
Statute and Regulation
Kansas
Case by case
Statute and Regulation
Kentucky
Registration
Statute and Regulation
Maryland
Case by case
Regulation
Missouri
Excluded from Solid Waste definition Statute and Regulation
Table 1 Contd’. State Status of Iron and Steel Slag (courtesy Nucor Steel)
Due to the tightening regulatory control of waste materials it is becoming more difficult to use iron and steel slags in applications that have been well established since
they are in many cases first considered a waste before being allowed a beneficial
use. For example, in the State of Texas the agency charged with ensuring environmental protection, the Texas Commission on Environmental Quality (TCEQ), established that the use of EAF steel slag in road construction applications was acceptable
and the EAF steel slag was a coproduct and not a recycled material or waste [2].
Subsequently, the Texas Department of Transportation (TXDOT) issued a specification known as Departmental Materials Specification DMS 11000 [3], whose purpose
is to evaluate the environmental properties of recycled materials proposed for use as
road construction material. Under DMS 11000, materials must undergo analysis and
testing necessary to demonstrate to TXDOT that they do not present a risk to human
health, the environment, or waters in the state when applied to the land or used in
products that are applied to the land. Since TXDOT views EAF steel slag as a recycled material it is subjected to DMS 11000. Unfortunately, through the structuring of
DMS 11000, EAF steel slag does not meet the criteria established by these guidelines and as such its use is not allowed by TXDOT.
Another example that highlights the hurdles of being regulated under solid waste
rules is the current situation in the Commonwealth of Pennsylvania where a new
general permit by the Pennsylvania Department of Environmental Protection
(PADEP) is being proposed. This new approval known as WMGR 082 [4] is designed to address the processing and beneficial use of steel slag, iron slag, and refractory bricks that were co-disposed with slag (“slag”) as a construction material.
The authorized processing is limited to magnetic separation of metallics and mechanical sizing and separation. Uses of slag as a construction material under this permit
are limited to the following: as an ingredient in bituminous concrete; as aggregate; as
base course; as subbase; and as antiskid material. In the past all that has been re-
quired by PADEP was a co-product determination to establish that slag from iron and
steel making had viable commercial applications that was similar to other competing
materials and natural aggregates. Many slag processors still have these co-product
determinations; however PADEP is strongly pushing for adoption of WMGR 082 as a
replacement of these co-product determinations. The requirements within WMGR
082 in many cases restrict the utilization of slag, but perhaps the more concerning
issue is the labeling of these products as beneficial use of a waste and not a product.
Similar examples can be found within other states and even in those states with favorable legislation for iron and steel slags, the regulatory environment is controlled by
the solid waste divisions of the various state environmental agencies.
Recognizing the challenges faced by slag processors, the slag industry in an attempt
to be proactive in the marketing and dissemination of information regarding the proper use of iron and steel slag, initiated a study on the risks to human health and the
environment in the application of iron and steel slags in various applications [5]. A
coalition comprising a group of 63 companies that produce steel, process slag, or
both, undertook a comprehensive study of the chemical composition of three slag
types generated during the steelmaking process and the potential human health and
ecological risks associated with possible exposure to such slag. Risk assessments
developed during 1998 and later revised in 2011 [6] demonstrate that these "slags
pose no meaningful threat to human health or the environment when used in a variety of residential, agricultural, industrial, and construction applications”. Nevertheless
even after compelling efforts to alleviate erroneous perceptions iron and steel slags
are often considered waste materials. With ever increasing regulatory scrutiny it is
necessary for the iron and steel and slag industry to consider other options to ensure
its survival and mitigate the increased costs iron and steel producers could potentially
face if these materials cannot be marketed.
Legislative Proposals
In November 2011 a coalition of steel makers and members of the National Slag Association gathered in Charlotte, North Carolina to discuss environmental issues associated with Iron and Steel Slag. A particular focus of this meeting was the challenges faced by the industry within the current regulatory framework surrounding the
use of iron and steel slag. The State of Indiana was recognized as one of, if not the
most, slag friendly state with a long history in the successful use of slag. The follow-
ing is an excerpt of the language from the Indiana statute on solid waste overseen by
the Indiana Department of Environmental Management (IDEM) which states “…the
board may not adopt rules under section 1 of this chapter to regulate the following
activities involving the legitimate use of slag generated by the production of iron or
steel under Bureau of the Census Standard Industrial Classification 3312” [7]. As
helpful as this exemption has been in Indiana it is still under the solid waste rules and
there is a perception that this allows the beneficial use of an industrial byproduct.
The group that came to be known as the “Slag Coalition” recognized that it was time
to be proactive rather than reactive in terms of addressing slag issues and the only
viable long lasting option was to find a way to have slag designated as a product. It
was agreed that the best way to accomplish this was to have a recognized National /
International Standards organization adopt a specification designating Iron and Steel
Slag as a product. Ironically and as previously discussed, ASTM already defines
both blast furnace and steel slag as products generated during iron and steel making.
The coalition has decided to work with ASTM to strengthen these definitions in order
to be able to justify to state legislators the prudence of removing slag from the various Solid Waste Department’s jurisdiction.
The key to a specification is to have a defined set of criteria established to use for
determining acceptance. The European R.E.A.C.H. program has set out to characterize materials for end use applications. For iron and steel slag they have developed
both chemical and mineralogical guidelines for characterizing the different types of
slag. The Coalition agreed that following the R.E.A.C.H. characterization approach
would be a good place to start in the development of the specification. Under solid
waste rules the criterion typically used to characterize a waste emphasizes the total
chemical constituent within a material. For a product however this is not relevant as
long as the product will not cause any harm to humans or degrade/contaminate the
environment when used as prescribed. This is why your stainless steel cutlery that
may be composed of higher levels of chromium than slag is acceptable as a food
utensil, while slag that is used for road building faces intense regulatory scrutiny. In
order to establish the benign aspects of iron and steel slags as far as environmental
considerations are concerned, the slag coalition has also embarked on characterizing
slag from a mineralogical standpoint. This is a relatively new approach within the US
slag industry, but it explains how the mineral constituents are bound together in the
slag. The goal here is to show that due to the binding of the chemical constituents
within slag the metallic elements are not environmentally available, and are very similar to other naturally occurring minerals regularly used in the construction industry.
The slag coalition believes that these efforts coupled with the work that has been
done on the human health and ecological risk assessment of slag and the lengthy
history of successful slag use in the United States, Europe, and other parts of the
world will provide compelling evidence to regulators that slag is a product.
Conclusions
Despite the low cost of aggregates as a basic product, they are an indicator of the
economic well-being of the Nation. The economic recession of the past six years has
strongly impacted the marketing of iron and steel slags in the United States.
This
alone is not necessarily the cause of the hardships the industry has faced when
compared to the similar drop-off in sales of natural aggregates. The lack of sales
may also be attributed to customer concerns with using a product that has been designated a waste, even if it has beneficial use exemptions.
Although iron and steel slags have over time proved their comparable if not superior
qualities in construction applications, the playing field has not been fair to these
products when compared to natural aggregates. The stigma associated with these
products being regarded as beneficial wastes although long lasting can be erased.
The first step is to legislatively establish these materials as bona-fide products free of
the regulatory constraints imposed by solid waste agencies. The proposed work plan
by the Slag Coalition to develop a specification, by which slag, as produced by the
manufacturer of Iron or Steel can be designated as a Product, is a first step to
achieving this objective. Thereafter this specification will be presented to a Standards Organization for adoption and ultimately be used in the legislative efforts to remove slag from waste characterization. Recent efforts by the Slag Coalition spearheaded by Nucor Steel in Nebraska have yielded positive results that now specifically
include legislative language within their statutes that not only specifically exempt slag
from being characterized as a waste but additionally state that slag is, “a product that
is a result of the steel manufacturing process and is managed as an item of value in
a controlled manner and not as a discarded material” [8]. Until there is an avenue to
proceed on a federal level, the plan is to address this issue on a state by state basis.
References
[1]
http://minerals.usgs.gov/minerals/pubs/commodity/iron_&_steel_slag/
[2]
Communiqué 1997 from TNRCC currently TCEQ
[3]
Texas Department of Transportation, DMS-11000, Evaluating and Using NonHazardous Recyclable Materials Guidelines.
[4]
PADEP General Permit No. WMGR082
[5]
ChemRisk. 1998. Human Health and Ecological Risk Assessment for BF, BOF
and EAF Slags. Prepared for the Steel Slag Coalition. ChemRisk, Inc.
[6]
Tox Strategies 2011 Human Health Risk Assessment for Iron and Steel Slag.
Prepared for the National Slag Association.
[7]
Indiana Environmental Statute 13-19-3-8
[8]
Nebraska Legislature LB203 - Change provisions relating to solid waste under
the Environmental Protection Act – April 24, 2013
H.J.C.M. Onstenk
DEVELOPMENT OF LEACHING TESTS FOR BY-PRODUCTS AND
OTHER SECONDARY CONSTRUCTION PRODUCTS IN EUROPE
Pelt & Hooykaas BV, Bijlstraat 5, 3087 AA Rotterdam, The Netherlands
Abstract
To assess the potential risk of the release of dangerous substances from by-products
and other secondary construction products producers should supply information
about the release of dangerous substances from their product(s). To support this, two
types of horizontal leaching tests are being developed. The dynamic surface leaching
test (DSLT) for monolithic materials and the up-flow percolation test (PT) for permeable, granular products.
One of the basic assumptions is that the assessment is only necessary for products
in their end-use, meaning that aggregates for concrete, mortar and bituminous mixtures do not require testing. Mixtures and aggregates for other intended end-uses
should be tested. The choice of the reference test depends on the characteristics of
the construction product, rather than its intended use.
Practical experiences with the DSLT and PT show that these tests can be used in a
feasible and practical way. Two examples are given of the practical use of these
tests.
Introduction
On July 1st 2013 the Construction Products Regulation (CPR) has become in force.
The CPR describes seven basic requirements for construction works that should be
fulfilled in order to prove fitness for their intended use. By-products and other secondary construction products are mainly used in civil works, like road construction
and hydraulic constructions. For the uses of these construction products basic requirements no. 3 is important, which deals with hygiene, health and environment.
One of the requirements is that the release of dangerous substances into ground water, marine waters, surface waters or soil should not be a threat to humans, environment or climate. To assess the potential risk of the release of dangerous substances
from by-products and other secondary construction products producers should supply information about the release of dangerous substances from their product(s).
Harmonized product standards describe the essential characteristics of construction
productions, which are necessary to fulfil the basic requirements of the CPR. For byproducts and other secondary construction products the aggregate standards developed by TC 154 are the most important ones. Release of dangerous substances is
not yet implemented in the present generation product standards, but CEN Technical
Committees are presently (2010 onwards) instructed by the European Commission to
provide information in their product standards on the potential release of a given list
of regulated dangerous substances comprising the relevant substances listed in European or notified regulations, for which a declaration of performance is or will be required. Such information will be provided under CE-marking and used to justify compliance with regulations.
Release scenarios
In order to be able to assess the release of dangerous substances from construction
products the release mechanism must be known. In relation to this two release scenarios have been defined [1]. The method of assessment should be based on the
scenario that is most likely to occur in practice. The assessment of the release of
dangerous substances will be addressed in the product standard.
Scenario I: Impermeable product or product with low permeability
The release mechanism is controlled by water flowing over the surface of the product
or transported into the matrix by capillary forces. In the matrix water movement is
slow and dissolved substances are transported out of the matrix by advection and
diffusion. At the surface substances may dissolve and precipitate. This release
mechanism is typical for monolithic products, bound materials and unbound materials
with (very) low permeability. Examples are bricks, concrete, asphalt, (bound) road
base materials, railway ballast and armourstone.
Scenario II: Permeable product
The release mechanism is controlled by water that infiltrates into the matrix driven by
gravity. Dissolved substances are transported out of the matrix by the gravity driven
flow. Transport with any water that is redirected at the surface is considered to be
negligible. This release mechanism is typical for unbound granular materials with a
moderate to high permeability.
Reference leaching tests
Assessment of the release of dangerous substances from by-products and other
secondary materials requires the choice of a reference leaching test which should be
used to characterize the leaching properties of the construction product. Reference
leaching tests are developed by CEN Technical Committee TC 351/WG 1 from 2006
onwards. For construction products two types of horizontal reference tests will become available:
1.
dynamic surface leaching test (DSLT) for monolithic materials and a modified
procedure for granular materials with (very) low permeability [2];
2.
up-flow percolation test (PT) for granular materials [3].
In the DSLT, sometimes called tank test, a test specimen is placed in a vessel and
submerged in a leachant (demineralised water). At predefined time intervals the
leachant is renewed in order to determine the time dependent release of dangerous
substances under predefined conditions. Release is expressed in mg/m2.
In the PT a vertically placed column is filled with the aggregate and continually percolated with a leachant (demineralised water) by an upward flow to achieve maximum
saturation. At predefined volumetric intervals the leachant is collected to determine
the release of dangerous substances as a function of the liquid to solid (L/S) ratio
under predefined conditions. Release is expressed in mg/kg.
Although the intended use is very important with respect to the assessment of a construction product, it is of minor interest for the choice of the appropriate reference
test. This choice mainly depends on the question how the material behaves in contact with water. If the material is permeable enough to allow water to flow through the
material by gravity, than the leaching properties of the construction product are controlled by matrix release and the percolation test should be used to determine the
leaching properties. If, on the other hand, the material is dense and durable enough
to prevent water flowing freely through its matrix during its use in a construction, than
the leaching properties are controlled by surface release and diffusion and the DSLT
should be used to determine leaching properties.
The question whether or not the construction product is in permanent contact with
water during its life in a construction is not relevant to the choice of the test method. If
(national) regulations distinguish different environmental conditions, e.g. semi-dry
and wet use or protected areas, than customized limit values will be needed for each
condition. Therefore it may be expected that (national) regulations control the condi-
tions of the use of construction products by defining limit values related to different
environmental conditions. Then, a producer or user will be able to easily assess the
environmental conditions under which a product may be used by comparing the
leaching test results (category limit values or declared performance according to the
CE-marking) and the regulatory limit values which should be used in relation to the
intended use.
Thus the choice of the appropriate reference test is mainly determined by the physical characteristics and durability of the construction product. Among others, it is expected from Product Technical Committees to include proper guidelines for the
choice of the reference test and a qualitative description of the intended use(s) in the
test standards. General instructions on the use of these tests are given in a guidance
document [1].
Assessment of by-products and other secondary construction products
Most by-products and other secondary construction materials are used in unbound
and bound layers in road construction, aggregate or filler in concrete or bituminous
mixtures, railway ballast, armourstone/gabions and embankments. These uses are
covered by the standards mentioned in table 22. It is noted that the assessment of
the release of dangerous substances is only necessary for products in their end-use.
CEN/TC 227
CEN/TC 154
TC Standard
Intended use
Type of
use
Reference
test
EN 12620
Aggregates for concrete
B
None
EN 13139
Aggregates for mortar
B
None
EN 13043
Aggregates for bituminous mixtures and surface
treatments for roads, airfields and other
trafficked areas
B/U 1)
None or PT
EN 13242
Aggregates for unbound and hydraulically bound
materials for use in civil engineering work and
road construction
U
None or PT
EN 13450
Aggregates for railway ballast
U/M 2)
PT or DSLT
EN 13383-1
Armourstone
U/M 2)
PT or DSLT
EN 13285
Unbound mixtures
U
PT
EN 14227-1 to -5
Hydraulically bound mixtures
B
DSLT
prEN 14227-15
Hydraulically stabilized soils (will replace Parts 10, -12, -13, -14 and a part of -11)
B
DSLT
U = unbound
PT
= percolation test
B = bound
DSLT = dynamic surface leaching test
M = monolithic
1)
B for bituminous mixtures and U for surface treatments
2)
M for monolithic granular materials and U for finer gradings
Table 22: European product standards relevant to the intended uses of by-products
and other secondary construction products
Aggregates used as a constituent in mixtures do not require assessment; only the
mixture is subjected to the assessment, because the leaching properties of the mixture have no correlation with the leaching characteristics of the constituents due to
chemical interactions and changes of the physical properties of the matrix.
Basically, this means that aggregates for concrete, mortar and bituminous mixtures
do not require testing; (un)bound mixtures and aggregates for other intended enduses should be tested, see figure 41. If testing is required the choice depends on the
dimensions of the product or grain size of the aggregate.
Constituent (raw
material or halfproduct)
End-use (unknown,
mixture or end-product)
Yes
Durable
monolithic
No
Low
permeability
No
Yes
No testing required
Dynamic surface
leaching test
Modified dynamic
surface leaching test
Percolation test
Figure 41: Decision scheme leaching reference test
A product is considered monolithic if:

for 3-dimensional products: all dimensions > 40 mm and a volume > 64 cm3;

for 2-dimensional (flat) products: a surface area > 100 cm2 and one dimension
< 40 mm;

for monolithic granular products: a grain size distribution according to table 23
and a particle density ≥ 2.3 Mg/m3.
Sieve size
Percentage passing by mass
63 mm
0 – 100
40 mm
0 – 75
22.4 mm
0–7
Table 23: Grain size distribution for monolithic granular products
In principle bound granular mixtures should be tested with the DSLT after hardening,
but when the hardened mixture is mechanically too weak to survive the test or looses
too many particles during the test, than the percolation test should be used. If the
criteria for a monolithic product are not fulfilled the product is considered granular by
definition. Besides, a monolithic product must be durable and maintain its monolithic
state during the whole service life of the construction. Normally, this will be covered
by the technical requirements with respect to volume stability, frost resistance, etc.
and does not need any additional requirements with respect to choice of the test
method.
Nevertheless, for practical reasons or cost efficiency it is possible to test monolithic
products with the percolation test instead of the DSLT. This may be useful in the
case of low release levels or when the intended uses of a material comprises monolithic uses as well as granular uses, while the intrinsic leaching properties of the
monolithic material and granular material are more or less the same.
A low permeable granular material, e.g. certain fly-ashes, may cause problems when
performing the percolation test due to a high hydraulic pressure in the column causing the column to leak or burst. In this case the modified DSLT is the only available
reference test.
Presently (2013) the grading requirements for the percolation test are not yet final.
There is still discussion on whether the test should produce grain size independent
intrinsic leaching properties based on testing size-reduced material (< 4 mm) or
whether the test should be performed on the aggregate 'as produced'. Tests on sizereduced material result in a 'worst case' leaching level, but may be more flexible in
use, because in principle only one test will be needed to cover all gradings of material, including monolithic granular gradings. Tests on gradings 'as produced' will probably require testing every single grading.
Examples of practical uses of the reference leaching tests
BOF slag: armourstone
In the Netherlands two gradings BOF slag are used as armourstone in hydraulic
structures: 32/90 mm and 45/180 mm. According to the Dutch regulations (Soil Quality Decree) the grading requirements of a 45/180 cover the grading requirements for
monolithic granular products and the DSLT is appropriate for the assessment of the
release of dangerous substances. The DSLT results from over 10 years of factory
production control (FPC) are summarized in figure 42. The data only apply to the
BOF slag produced by Tata Steel IJmuiden.
Release (% of the SQD limit value)
140
120
maximum value
100
80
90 % interval
60
median
40
20
minimum value
SO4
F
Cl
Br
Zn
V
Sn
Se
Ni
Mo
Pb
Hg
Cu
Co
Cr
Cd
Ba
As
Sb
0
Figure 42: DSLT results showing the release of dangerous substances from BOF
slag 45/180 (release is expressed as a percentage of the Dutch regulatory limit
value)
Of all regulated dangerous substances only vanadium shows a significant release. A
detailed analysis of the release data of vanadium according to annex B of the draft
DSLT standard [2] shows that the release of vanadium is mainly controlled by diffusion, sometimes followed by depletion. In some cases the release mechanism cannot
be established, probably due to pH changes during the test caused by the dissolution
of free lime.
BOF slag 32/90, however, does not fulfil the Dutch requirements for monolithic
granular products, which differ from table 23. In the Netherlands BOF slag 32/90 is
considered granular and the PT should be used for the assessment. The test results
are summarized in figure 43. The data only apply to the BOF slag produced by Tata
Steel IJmuiden. To show that there is no difference in the intrinsic release between
the gradings 32/90 and 45/180 figure 44 shows PT results of BOF slag 45/180.
Release (% of the SQD limit value)
140
120
100
80
60
40
20
SO4
F
Cl
Br
Zn
V
Sn
Se
Ni
Mo
Pb
Hg
Cu
Co
Cr
Cd
Ba
As
Sb
0
Figure 43: PT results showing the release of dangerous substances from BOF slag
32/90 (release is expressed as a percentage of the Dutch regulatory limit value)
Release (% of the SQD limit value)
140
120
100
80
60
40
20
SO4
F
Cl
Br
Zn
V
Sn
Se
Ni
Mo
Pb
Hg
Cu
Co
Cr
Cd
Ba
As
Sb
0
Figure 44: PT results showing the release of dangerous substances from BOF slag
45/180 (release is expressed as a percentage of the Dutch regulatory limit value)
Because the PT has been performed on size reduced material (products crushed to
< 4 mm) it is possible to test both gradings as one product in FPC with respect to the
assessment of the release of dangerous substances, thus saving on costs of testing.
The level of release is such that it may be concluded that testing of the 45/180 grading on the basis of the DSLT is not needed.
Slag bound air-cooled blast furnace slag
After hardening slag bound air-cooled blast furnace slag (mixture of air-cooled blast
furnace slag, granulated blast furnace slag and BOF slag) becomes a monolithic
product which is strong enough to survive the DSLT. Because of the higher costs of a
DSLT, the product is mainly tested with the PT (the non-hardened material is crushed
to < 4 mm). Nevertheless, the mixture is also tested with the DSLT, but at a lower
test rate (once a year), because the leaching behaviour of the monolithic product is
considered to give a more realistic release compared to non-hardened or crushed
hardened material (see figure 45 and figure 46).
Release (% of the SQD limit value)
140
120
100
80
60
40
20
SO4
F
Cl
Br
Zn
V
Sn
Se
Ni
Mo
Pb
Hg
Cu
Co
Cr
Cd
Ba
As
Sb
0
Figure 45: PT results showing the release of dangerous substances from bound
ABFS 0/45 (release is expressed as a percentage of the Dutch regulatory limit value)
Release (% of the SQD limit value)
140
120
100
80
60
40
20
SO4
F
Cl
Br
Zn
V
Sn
Se
Ni
Mo
Pb
Hg
Cu
Co
Cr
Cd
Ba
As
Sb
0
Figure 46: DSLT results of hardened, monolithic material showing the release of
dangerous substances from bound ABFS 0/45 (release is expressed as a percentage
of the Dutch regulatory limit value)
In this case the product is assessed on the basis of the test results of the DSLT,
while the test results of the PT are used to monitor the product and trigger the use of
the DSLT.
The leaching of the slag bound ABFS mixture produced in IJmuiden is dominated by
the release of Ba, V and SO4 (figure 46). The release of these substances is controlled by diffusion, although the release mechanism of SO4, and in a lesser extent
the release of V, cannot always be established, probably due to pH and redox
changes during the test and interaction of SO4 with Ba and Ca.
Conclusion
Two types of horizontal leaching tests are being developed in order to allow producers of construction products to prove that products are fit for their intended use with
respect to the release of dangerous substances. The use of these tests is related to
the way a product behaves in contact with water. The dynamic surface leaching test
(DSLT) is meant for monolithic materials where release is mainly controlled by diffu-
sion (surface release), while the up-flow percolation test (PT) is meant for permeable,
granular products that allow water to flow through by gravity (matrix release). Thus,
the choice of the reference test depends on the physical characteristics and durability
of the construction product, rather than its intended use.
Presently (2013) the DSLT is more or less finalized and ready for round robin testing.
The PT is still under development; the main question is whether the test should be
performed on size reduced material or on the aggregate 'as produced'. A testing program has been proposed to give answers.
The assessment of the release of dangerous substances is only necessary for products in their end-use, meaning that aggregates for concrete, mortar and bituminous
mixtures do not require testing. Mixtures and aggregates for other intended end-uses
should be tested.
Practical experiences with the DSLT and PT show that these tests can be used in a
feasible and practical way, not only for characterization and type testing but also for
routine testing (FPC). Furthermore, it is possible to test monolithic products with the
PT instead of the DSLT and combine groups of products by testing a 'worst case'
grading (size reduced material) to reduce the number of tests.
References
[1]
Technical Specification prCEN/TS xxx-1:2012, Construction products – Assessment of release of dangerous substances – Part 1: Guidance for the determination of leaching tests and additional testing steps, CEN/TC 351/WG 1 N
407, January 2013
[2]
Technical Specification prCEN/TS xxx-2:2012, Construction products – Assessment of release of dangerous substances – Part 2: Horizontal dynamic surface leaching test, CEN/TC 351/WG 1 N 410, January 2013
[3]
Technical Specification prCEN/TS xxx-3:2012, Construction products – Assessment of release of dangerous substances – Part 3: Horizontal up-flow percolation test, CEN/TC 351/WG 1 N 414, January 2013
A. Schuurmans
Sustainability of construction works - European standards and implications for secondary materials
Chairman NEN commission 351281 Sustainability of Construction Works, and Rockwool International, agnes.schuurmans@rockwool.com
Abstract
Several CEN standards for the sustainability (environmental, social and economic)
assessment of construction works are published and in preparation in CEN TC350.
The environmental pillar of this suite of standards includes the environmental assessment method for buildings and the standard EN15804 for Environmental Product
Declarations (EPD), based on Life Cycle Assessment. The EN15804 provides the
rules for LCAs of construction products and the presentation in an EPD in a way that
allows aggregation for constructions.
This presentation will explain the LCA calculation rules for secondary materials and
waste in the CEN TC350 standards. It will be discussed what (potential) role the
standards have in the context of European policies with regard to environment, sustainability and resource efficiency.
1. TC350 Sustainability of Construction Works
Mandated by the European Commission [1], CEN TC350 developed voluntary
horizontal standardised methods for the assessment of the sustainability of construction works and core rules for the product category of all construction products [2]. The standards in the suite of CEN TC350 describe the harmonized
methodologies for assessment of environmental, social and economic performance of constructions and construction products over their life cycle. The standards provide the indicators, the methodologies for measuring and quantification of
the indicators, and provide the communication format. The parameters for measuring sustainability are selected and agreed indicators as used and widely accepted across Europe. There are 22 environmental parameters (indicators), 6 aspect categories for describing social performance (resulting in total of more than
140 indicators!) and 3 indicators for quantifying the economic performance of a
construction. Of course the environmental performance parameters, the social parameters and the economic parameters are interlinked and inter-depending.
The sustainability of construction works is defined according to the three wellknown pillars of sustainability: people (social aspects) – planet (environmental
impacts) – profit (economic aspects). The suite of standards is designed accordingly. See Figure 1.
A life-cycle approach is applied: a construction is assessed from cradle-to-grave.
Note that it is the construction work that is assessed. In order to do the environmental assessment, data are required from products. These data are provided by
Environmental Product Declarations (EPD).
Figure 1. TC350 standards overview [2] (note: prEN16309 has already changed
into FprEN16309; WI 017 into prEN16627)
The standards provide the assessment method and result in a set of assessed
parameters. The importance of the parameters and their values are political
choices and may depend on local circumstances. There is no absolute definition
of what is sustainable. The standards do therefor not provide absolute norms.
Figure 2. Environmental parameters for sustainability of construction works and
Environmental Product Declarations
The framework of the sustainability assessment is based on the principle that a
construction work first of all must meet functional and technical requirements.
Constructions that do not fulfill the required function are not sustainable: it is likely
that they are demolished or changed before the end of the designed service life,
consequently leading to a higher environmental impact, negative social impact
and higher costs. Only when functional and technical requirements are met, a
sustainability assessment can be carried out. See Figure 3.
Most of the sustainability standards - all environmental assessment standards
and part of the social and economic - are finalised and available now. The package of standards will be further completed in 2013 and 2014. A new work item
started for civil engineering works.
Figure 3. Sound design as basis for assessing sustainability of construction works
2. Recycling, waste and secondary materials
Modular approach
Environmental Product Declarations (EPDs) are the input for a holistic building
assessment taking into account the functional and technical performances in a
construction work’s context. For this purpose EPDs must be addable to combine
them in a construction works calculation. This implies the same system boundaries, calculation rules, parameters etc.. The functional and the technical performance of a construction product depend on the building system and it requires
scenarios in order to make a useful declaration on a quantified contribution of the
product to the sustainability. The building assessor needs knowledge on the use
of the building, the construction process, the end-of-life processes, but also on
adapting scenarios and corresponding product data given in the EPD. The modular approach in the standards is designed for this purpose: the manufacturer of a
construction product supplies data from ‘cradle-to-gate’ with optional modules for
scenarios of the other life stages, whereas the building assessor can ‘built’ the life
cycle of the construction out of cradle-to-gate EPDs combined with data from relevant scenarios for the specific building. See Figure 4.
Figure 4. Modular approach in TC350 standards
Comparing products and selecting products based on environmental performance
stated in an EPD is only possible under strict conditions. In practice a direct comparison by means of EPD is almost impossible: the scenarios in the life cycle and
the performance in the construction should be fully identical.
Recycling and secondary raw materials
The system boundaries are set according to the “polluter pays principle”: Processes of waste processing shall be assigned to the product system that generates the waste until the end-of-waste state is reached.
The end-of-life stage of the construction product starts at the end-of-life of the
building and/or when it is replaced dismantled or deconstructed from the building
or construction works and does not provide any further functionality. During the
end-of-life stage of the product or the building, all output from dismantling, deconstruction or demolition of the building, from maintenance, repair, replacement or
refurbishing processes, all debris, all construction products, materials or construction elements, etc. leaving the building, are at first considered to be waste. At
some point, this output however reaches the end-of-waste state. The system
boundaries are defined according to the end-of-waste criteria from the Waste
Framework Directive. Loads and benefits of potential recycling can be declared
on a voluntary basis in the so-called module D.
In the case of input of secondary materials or energy recovered from secondary
fuels, the system boundary between the system under study and the previous
system (providing the secondary materials) is set whereoutputs of the previous
system, e.g. materials, products, building elements or energy, reach the end-ofwaste state. The use of secondary raw materials is therefor said to be ’free of environmental burden’.
Example
Concrete that remains from demolition of a building, remains waste until the endof-waste state is achieved. All environmental loads from the waste processing are
allocated to the building’s life cycle (i.e. the concrete). If the concrete is recycled
into aggregate and if the aggregate is defined as being the end-of-waste state
(based on Waste Framework Directive), the waste processing until aggregate is
allocated to the building/concrete. The environmental loads and benefits of using
the aggregate, e.g. substitution of a stony material in new concrete, are declared
in Module D. The use of aggregate in a new life cycle is free of environmental
burden.
Example
By-products in production are either “waste” or “product” (according to the Waste
Framework Directive). If “waste” then the processing belongs to the original production; the user of the “waste” gets the material free of environmental burden. If
“product” then co-allocation is applied, i.e. that the environmental burden of the
production is divided over the products and co-products according to mass or
economic allocation principles. The user of the “product” shall take part of the environmental burden into account. Such definitions of “waste” or “product” must be
made by the manufacturer, for instance for slag as a by-product of iron production
and used in cement production.
3. The European strategic framework
The EU Roadmap to a Resource Efficient Europe [3] recognises the relevance of
the construction sector and buildings for resource efficiency in the EU. Buildings
use 42 % of our final energy consumption, the sector uses > 50 % of all extracted
materials – most of them minerals and produces 33 % of the waste in the EU. The
Roadmap concludes that existing policies, mainly linked to energy efficiency,
need to be complemented with policies for resource efficiency looking at a wider
range of resource use and environmental impacts, across the life cycle of buildings and constructions. In July 2013 the European Commission started a public
consultation on Sustainable Buildings, which will result in a Communication on
Sustainable Buildings, probably early 2014. Industrial buildings and infrastructure
are excluded.
Meanwhile, the Communication on ”Strategy for the sustainable competitiveness
of the construction sector and its enterprises” of 31st July 2012 [4] points to the
main challenges that the sector faces up to 2020 in order to grow strong and
more viable in the future. This includes improving resource efficiency, environmental performance and related business opportunities. It does not elaborate the
issues but refers to the Communication on Sustainable Buildings and, in its attached action plan, refers to an EU wide life cycle costing methodology applied to
buildings for green public procurement.
This focus on resource efficiency is reflected in new initiatives such as the new
Works Requirement no.7 in the CPR ‘Sustainable use of Natural Resources’.
Other initiatives related to sustainability and resource efficiency also pop up for
the construction sector, such as ecodesign for energy-related products (a.o. windows) and the Communication and Recommendation on the Single Market for
Green Products [5], referring to Product Environmental Footprints (PEF) and Organisational Environmental Footprints (OEF). These policies are still in discussion
and development and a final ‘picture’ of the future requirements for the construction sector in this area is not set yet.
4. Market developments
Product standards
While awaiting clarity from the European Commission on their various initiatives
and lack of info on CPR BWR7, several product TCs already decided to start with
the implementation of EPD in the product standards. The EN15804 is elaborated
and specified in more specific Product Category Rules (PCR). Liaisons with
TC350 are established to secure uniform implementation of the horizontal standard EN15804. EPDs will remain voluntary, but more uniform across Europe. This
is a step forward towards a possible link of certain EPD parameters and CPR
BWR7. But it also means a further harmonisation between the various national
product-specific PCRs of private EPD schemes in Europe. More than 6 product
TCs started the work and many others are considering to do so [6].
It can be envisaged that the CPR in future may include certain (environmental)
parameters, to be declared through the CE-marking like any other product characteristics, and to be used for environmental calculations for constructions [7]. It
will be up to member states to develop and notify any such requirements. The selection of parameters and the requirements are political choices made by the individual member states.
National requirements in member states
As of January 2013 the Netherlands introduced a requirement for new buildings to
calculate the Global Warming potential and Abiotic Depletion potential. Two environmental parameters are chosen out of 22 in the TC350 standards: a political
choice. There are no requirements (yet). The years to come will be used to gain
experience. It is not mandatory to provide EPDs of construction products. However, the default values in the database that is used for the building calculations get
an additional ‘safety’ value if no EPD is available [8].
France and Belgium prepare legislation for environmental product claims, which
must be accompanied by an EPD [9].
The role of voluntary EPD-programs and sustainability schemes
Most of the European countries have an EPD scheme available or in preparation.
These schemes are private. Up to now the main driver is the market: requests
from clients and marketing benefits. But as more and more companies introduce
CSR (Corporate Social Responsibility), EPDs become a ‘natural’ thing, being the
basis for product evaluation and transparent communication about environmental
performance.
At least as long as the EU framework for sustainability and EPDs is not settled,
there is a role to play for the EPD-programs. Mutual recognition and common
quality guidelines are elaborated in the European ECO Platform, a platform of the
main EPD schemes in Europe [10]. This Platform will certainly help to introduce
EPDs more broadly and more harmonised across Europe.
Sustainability schemes for buildings and constructions, such as BREEAM, DGNB
and HQE also adopted the CEN TC350 standards. LEED from the USA acknowledges EPDs, although not directly adopting the European standards. Similar
schemes for infrastructural works will certainly go into the same direction.
5. Conclusions
European standards for the sustainability assessment of constructions works are
available through CEN TC350. They are already applied in regulations of some
member states, in several (voluntary) European sustainable building assessment
schemes and in (voluntary) EPD schemes in many countries. Several product
TCs are preparing the implementation of EPD in product standards, to harmonise
product specific interpretations and to be prepared in case of future requirements
in the CPR.
The methodological choices for assessing the environmental impact of secondary
materials are based on ‘the polluter pays principle’ and end-of-waste definitions in
the Waste Framework Directive. These choices are made in the context of the
standards and their goal: a methodology to assess the sustainability performance
of construction works.
References
[1] Mandate M350:
http://ec.europa.eu/enterprise/standards_policy/mandates/database/index.cfm?fu
seaction=search.detail&id=228#
[2[ website TC350:
http://portailgroupe.afnor.fr/public_espacenormalisation/CENTC350/index.html
[3] COM(2011)571 of 20.9.2011
[4] COM(2012)433 of 31.7.2012
[5] COM(2013)196 of 9.4.2013
[6] CEN workshop 20.6.2013:
http://portailgroupe.afnor.fr/public_espacenormalisation/CENTC350/CEN_TC350
_seminar_EN15804_presentations.pdf
[7] Gargari, C. (University of Florence), Hamans, C. (ESC), Chiara Torricelli, M.
(University of Florence), Techne 5 (2013), The Building sector commitment to
promote the sustainability of construction products: a common European approach for the Environmental Product Performances
[8]
Dutch
environmental
database,
https://www.milieudatabase.nl/
and
http://safeandsustainablebuildings.com/netherlands-first-mover-in-sustainablebuilding-requirements/
[9] Notification 2013/301/B (Belgium)
[10] ECO Platform website: http://www.eco-platform.org/
More information on CEN TC350 standards can be found on:
http://www.hamans.com/sustainability-standards/scheme-standards
Members of EUROSLAG
Chairman: Dr.-Ing. Heribert Motz
Acciaierie Bertoli Safau S.p.A.
ACRONI, d.o.o.
Via Buttrio 28
Cesta Borisa Kidrica 44
33050 Pozzuolo del Friuli (UD)
4270 Jesenice
ITALY
SLOVENIA
AEIFOROS Metal Processing S.A.
AFOCO
12th klm Old National Road Thessaloniki-Veria
2, boulevard Henri Becquerel
570 08 Ionia, Thessaloniki
57970 Yutz
GREECE
FRANCE
Centre Technique et de Promotion
ArcelorMittal Belval & Differdange S.A.
des Laitiers Sidérurgiques (CTPL)
66, rue de Luxembourg
Immeuble le Cézanne
4009 Esch-sur-Alzette
6, rue André Campra
LUXEMBOURG
93212 - La Plaine St Denis
FRANCE
CIMALUX S.A.
Cloos S.A.
BP 146
BP 71
4002 Esch-sur-Alzette
4001 Esch/Alzette
LUXEMBOURG
LUXEMBOURG
Dalmine S.p.A.
Fachverband Eisenhüttenschlacken e.V.
Piazza Caduti 6 Luglio 1944, 1
Bliersheimer Straße 62
24044 Dalmine (BG)
47229 Duisburg
ITALY
GERMANY
FEhS - Institut für Baustoff-Forschung e.V.
Bliersheimer Straße 62
47229 Duisburg
GERMANY
Groupement de la Sidérurgie
Staalindustrie Verbond (GSV)
Boulevard de la Plaine 5
1050 Bruxelles
BELGIUM
HARSCO Metals Polska Sp. z o.o.
Jernkontoret
Ul. Piłsudskiego 82
Box 1721
42-400 Zawiercie
111 87 Stockholm
POLAND
SWEDEN
Mineral Products Association MPA Slag
Outokumpu Oyj
Gillingham House, 38 - 44 Gillingham Street
P.O. Box 140
London SW1V 1HU
02201 Espoo
UNITED KINGDOM
FINLAND
Pelt & Hooykaas B.V.
Ruukki Metals Oy
Postbus 59011
P.O. Box 93
3008 PA Rotterdam
92100 Raahe
THE NETHERLANDS
FINLAND
SSAB Merox AB
613 80 Oxelösund
SWEDEN
Tata Steel IJmuiden BV
P.O. Box 10 000
1970 CA IJmuiden
THE NETHERLANDS
Tapojärvi Oy
U. S. Steel Kosice, s.r.o.
Laivurinkatu 2-4 c 32
Vstupný areál U. S. Steel
95400 Tornio
044 54 Kosice
FINLAND
SLOVAK REPUBLIC
Unión de Empresas Siderúrgicas UNESID
voestalpine Stahl GmbH
C/Castelló, 128
voestalpine-Straße 3
28006 Madrid
4020 Linz/Donau
SPAIN
AUSTRIA
Associate Members of EUROSLAG
Colakoglu Metalurji A.S.
41455 Dilovasi - Kocaeli
TURKEY
SCB International Materials, Inc
PO Box 335
Newtown, CT 06470
USA
Phoenix Slag Services SRL
C.P. 141; Oficiul Postal nr 1
800710 Galati
ROMANIA
List of Speakers
1. Dr. G. Endeman, WV Stahl / VDEh, Germany, Head of Business Area Politics, gerhard.endemann@stahl-zentrum.de
2. Dr Y.C. Lee, China Steel Corp., Taiwan, Scientist, t621@mail.csc.com.tw
3. D. Mombelli, Politecnico Di Milano - Mechanical Dept., Italy, PHD Student,
davide.mombelli@mail.polimi.it
4. Dr. H. Schliephake, Georgsmarienhütte GmbH, Germany, Member of the
Managing Board, henning.schliephake@gmh.de
5. Dr I.McDonald, Siemens VAI, United Kingdom, Blast Furnace Innovation
Manager, ian.j.mcdonald@siemens.com
6. H.Kappes, Paul Wurth, Germany, Head of BA Energy and By-Products,
horst.kappes@paulwurth.com
7. D.Piorier, ArcelorMittal Maizieres Research, France, Research Engineer, delphine.poirier@arcelormittal.com
8. A.E. Yildizcelik, Istanbul Technical University, Turkey, MsC Student, eyildizcelik@gmail.com
9. Dr H. Epstein, RVA, Technical Consultant, hepstein@013net.net
10. Dr I. Unamuno, Gerdau Aceros Especiales I+D Europe, Spain, Senior Researcher, inigo.unamuno@gerdau.es
11. Dr
J.S. Chen, National Cheng Kung University, Taiwan, Professor,
jishchen@mail.ncku.edu.tw
12. E. Nagels, InsPyro, Belgium, Project Director, Els.Nagels@inspyro.be
13. Dr N. Ghazireh, Lafarge Tarmac, United Kingdom, Senior Manager - R&D,
Nizar.ghazireh@lafargetarmac.com
14. V. Feldrappe, FEhS - Institut für Baustoff-Forschung e.V., Germany, Research Staff Member, v.feldrappe@fehs.de
15. S.M. Choi, Kongju National University,
South Korea, Doctoral student,
최선미 smchoi@kongju.ac.kr
16. J. Roininen, Oulu University / Centre for Environment and Energy, Finland,
Project Coordinator, juha.roininen@oulu.fi
17. Dr V. Colla, Scuola Superiore SantÁnna, Italy, Technical Research Manager,
colla@sssup.it
18. Dr T.A. Branca, Scuola Superiore SantÁnna, Italy, Assistant Researcher, teresa.branca@sssup.it
87
19. Dr P. Drissen, FEhS - Institut für Baustoff-Forschung e.V., Germany, Deputy
Head of Slag Metallurgy/Engineering and Fertiliser Dep., p.drissen@fehs.de
20. Dr A. Ehrenberg, FEhS - Institut für Baustoff-Forschung e.V., Germany, Head
of Building Materials Department, a.ehrenberg@fehs.de
21. J. Yzenas, Edw.C.Levy Co, United States of America, Director of Technical
Services, jyzenas@edwclevy.net
22. Dr S. van der Laan, Tata Steel IJmuiden R&D, the Netherlands, Researcher,
sieger.van-der-laan@tatasteel.com
23. Dr I. Sohn, Yonsei University, Seoul, Korea, Associate Professor, ilsohn@yonsei.ac.kr
24. M. Provance-Bowley, Harsco Metals and Minerals, United States of America,
Senior Technical Development Specialist - Agriculture & Turf, mpbowley@harsco.com
25. Dr E. Poultney, Tata Steel Research Development & Technology, United
Kingdom, Senior Researcher, edwin.poultney@tatasteel.com
26. J. Domas, CTPL, France, Manager, jeremie.domas@ffa.fr
27. Dr C. Ochola, Tube City IMS, United States of America, Environmental Engineer, cochola@tubecityims.com
28. E. Onstenk, Pelt&Hooykaas BV, the Netherlands, Product Specialist,
e.onstenkl@pelt-hooykaas.nl
29. A. Schuurmans, NEN Committee Sustainable Building / ROCKWOOL Int.,
the Netherlands, Chair NEN Committee, agnes.schuurmans@rockwool.com
30. Dr H. van der Sloot, Hans van der Sloot Consultancy, the Netherlands,
Owner, hans@vanderslootconsultancy.nl
88
3.
List of speakers
1
G. Endeman, WV Stahl / VDEh, Head of Business Area Politics, gerhard.endemann@stahl-zentrum.de
2
Y.C. Lee, China Steel Corp., Scientist, t621@mail.csc.com.tw
[iii]
Taylor, H. F. W., Ed. (1990). Cement Chemistry. London, Academic Press.
[iv]
“Additions of Industrial Residues for Hot Stage Engineering of Stainless Steel
Slags.” Pontikes et al., Proceedings of the 2nd International Slag Valorization
Symposium, April 2011, Leuven, p314
[v]
Iacobescu, R.I., Pontikes, P., Malfliet, A., Machiels, L., Epstein, H., Jones,
P.T., and Blanpain, B., “A Secondary Alumina Source for the Stabilization of
CaO-SiO-MgO Slags.” Proceedings of the 3rd International Slag Valorization
Symposium, KU Leuven, Belgium (March 2013): 311-314.
[vi]
European Parliament Directive (2003). "2003/53/EC of the European Parliament and of the Council."
[vii]
U.S. Department of Health and Human Services, Centers for Disease Control
and Prevention, et al. (2008)."Criteria document update, Occupational Exposure to Hexavalent Chromium." External Review Draft.
[viii]
Mudersbach, D., M. Kühn, et al. (2009). Chrome immobilization in EAF-slags
from high-alloy steelmaking: tests at FEhS institute and development of an
operational slag treatment process. First International Slag Valorization Symposium, Leuven, Belgium.
The technical contribution of Inspyro Ltd. to this work is acknowledged.
Address for correspondence: hepstein@013net.net
89