Chapter 5

Strip casting technology

Abstract
Strip casting technology is the most recent innovative steel casting technology that
integrates casting and rolling; thus, re-heating the steel is avoided. In this case
study, we evaluate the effect of government R&D support on the development of strip
casting technology.
The original roots of strip casting technology go back to the 19th century. Bessemer,
one of the classical inventors of the steel industry, applied for a patent in 1857. After
renewed interest in strip casting technology arose between 1980 and 1985, a robust
and large technology network emerged. The eleven micro-networks were remarkably
homogeneous, a large steel-maker and a machine supplier or engineer. The steel
manufacturers took the lead. Six of the eleven micro-networks are still active. Three
of them operate strip casting technology on an industrial scale. They needed about
fifteen years to achieve this state.
The major argument for developing strip casting technology has been the need to
reduce the capital intensity of hot rolling. This is especially attractive for smallcapacity facilities such as mini-mills and stainless steel facilities. Bessemer was
already aware of the huge advantages of direct casting. The introduction of
conventional continuous casting (1952), maturing of this conventional technology,
the steel crises in the seventies, and the rise of stainless steel and mini-mills had to
occur before strip casting technology became the centre of casting R&D activities.
Between 1975 and 1985, technologists started looking for more compact casting
technologies.
Various national governments and the European Coal and Steel Community (ECSC)
contributed 5 to 10% of the total expenditure. In three micro-networks, R&D support
was more than 40%. These micro-networks stopped R&D activities or deliberately
continued on a pilot scale. The three micro-networks that are ahead in developing
strip casting technology did not obtain any external R&D support.
The major conclusion of this chapter is that the effect of government R&D support
on the development of strip casting technology has been minimal. The development
proved to have a strong momentum of its own. Strip casting affects the core of steel
business; its development was loosely influenced by energy-efficiency improvements
or by government R&D support.

- 129 -

5.1. Introduction
Improving the energy efficiency of manufacturing industries is seen as one of the
important options for reducing greenhouse gas emissions [IPCC, 1996; UN, 1997].
As a major consumer of energy, the iron and steel industry is always mentioned as a
sector where energy efficiency needs to be encouraged. Many steel technologies
have been investigated with regard to their techno-economic potential for bringing
about energy-efficiency improvement [De Beer et al., 1994; WEC, 1995; IWG,
1997]. There is an increasing interest in the long-term potential of innovative energyefficient technologies. De Beer (1998) has performed work in this area. He identified
several innovative energy-efficient technologies for the iron and steel industry. Longterm savings of 35 to 60% should be achievable [De Beer, 1998]. Financial R&D
support for the development of such innovative energy-efficient technologies is often
suggested as an attractive strategy for government to reduce CO2 emissions (see e.g.
[Blok et al., 1996; De Beer, 1998]). However, very little is known about the effect of
R&D support on the development of industrial energy-efficient technologies.
The aim of this chapter is to evaluate the effect of R&D support on the development
of an energy-efficient technology. For this purpose, we make a detailed investigation
of the networks within which strip casting technology, the most recent innovative
casting technology for the steel industry, is developed. Strip casting technology is an
interesting case study because both energy-analysts and policy makers consider strip
casting technology as one of the major innovative energy-efficient technologies for
the iron and steel industry.
In Section 5.2, we briefly introduce the steel industry, the steel production process
and strip casting technology. Subsequently, the historical development of strip
casting technology is mapped. Section 5.3 sketches the early days of the
technologys development up till 1985. In Section 5.4, we describe the more recent
R&D history of strip casting technology. In mapping this history we present
elements, which are of interest to our analysis of the case study. This analysis is
worked out in more detail in Section 5.5. A short discussion of the validity of the
analysis is included in Section 5.6. Section 5.7 zooms in on the role of government
R&D support. The chapter closes with conclusions about the effect of R&D support
in stimulating the development of strip casting technology.

5.2. The sector, the production process and strip casting

technology
Iron and steel industry
The production of iron and steel increased substantially after the Second World War.
In 1945 worldwide steel production was about 100 millions tons; in 1997 more than

- 130 -

790 millions tons crude steel were produced [IISI, 1998b]. Large integrated steel
mills that produced up to 5 million tons steel/year met the huge demand for steel.
From 1970 onwards, prospects deteriorated. World steel prices were forced
downwards. Over-capacity, the oil crises during the seventies, the privatisation of
former national steel companies, and the growing competition from other materials,
forced the steel market to change from a suppliers market to one that was dominated
more and more by demand [Holschuh, 1995; Berg, 1996; Vrieling, 1998].
Restructuring was needed and the key-words became efficiency, just-in-timemanagement, and competitiveness. Quality and flexibility became more important
[Birat, 1992; Hamels, 1995; Berg, 1996; IISI, 1998a; Abbel, 1999].
The changed market situation affected R&D within the steel industry. On the one
hand process R&D continued; more compact technologies with a lower capital
requirement were required. On the other hand product R&D became more important.
Co-operative developments with customers (and suppliers) were undertaken [Berg,
1996; Holschuh, 1995; Hamels, 1995; Abbel, 1999; Birat, 1999b].
Steel firms spend about 2.5 billion US dollars annually on R&D [OECD, 1997]. On
average, this represents roughly 1% of sales. Integrated steel-makers in industrialised
countries often spend more than 1% of their sales on R&D. Small firms spend less.
Firms operating mini-mills undertake almost no R&D1 [Holschuh, 1995].
Making steel
There is a wide variety of grades of steel. For the purpose of this chapter, we make a
distinction between carbon steel and stainless steel. Carbon steel is steel that relies on
the carbon content for its structure and properties. Most of the steel produced is
carbon steel. Stainless steel2, one of the specialty steels, contains more than 10.5%
chromium and sometimes other elements. Stainless steel resists corrosion and
maintains its strength at high temperatures. Worldwide stainless steel production is
about 2% of the world wide crude steel production, which was 776 million tons in
1998 [IISI, 1998b].
The production process for all these grades is to a large extent the same. Differences
occur in the last two stages of the steel production process. In total there are four
stages:
1. Production of iron
2. Crude steel production
3. Casting into semi-finished products (blooms, billets and slabs)
1

Mini-mills are small-scale steel mills that melt scrap, using electric arc furnaces, to produce
commodity products. Mini-mills differ from integrated steel mills with regard to their minimum
annual output, labour relations, product markets, and management style [AISI, 2000]. In 1997, over
33% of the annual steel production was manufactured in mini-mills [IISI, 1998b].
2
Stainless steel production has increased rapidly in the last decades of the 20th century: over 16
million tons of finished stainless steel were produced in 1997, compared to 13 million tons in 1988, an
increase of more than 23 per cent. The largest stainless steel-producing countries are Japan (nearly 4
million tons), the United States (2 million tons), and Germany, Republic of Korea, Italy and France,
each of which produced over 1 million tons of finished stainless steel in 1997 [IISI, 1998b].

- 131 -

4. Rolling and shaping final products hot rolling, cold rolling and finishing
operations
The production process is energy intensive due to the chemical reaction energy
required in the production of iron, the need for high temperatures and for several reheating and cooling steps. Such heating and cooling steps can be avoided by using
technologies that combine two or three processes. Such technologies can
substantially reduce the energy required to produce one ton of steel [De Beer, 1998;
IISI, 1998a].
Strip casting technology
This chapter focuses on strip casting technology. Strip casting does away with the
need for re-heating that is still required nowadays when the cast slabs (stage 3) are
further processed to final products (stage 4). Strip casting technology permits the
production of hot strips, also referred to as hot rolled coiled steel, directly from the
liquid crude steel. Casting and rolling are linked into one continuous operation.
Strip casting technology is the most recent innovative casting technology and (thus
far) the last technology in realising direct casting. We use the generic term direct
casting to denote casting in sizes that are as close as possible to final products. In this
chapter, we show that direct casting has been realised in a step-wise fashion: the
thickness of the sheet or strip cast has been reduced by three innovative technologies,
conventional continuous casting, thin slab casting and strip casting technology (see
Figure 1).
Liquid steel
in ladle

Hot rolled
coiled steel
Ingots

Ingot
casting

Ingot caster

Conventional
continuous
casting

Heating +
Breakdown mill

Continuous
caster
(~ 1 m/min)

Slabs,
billet,
blooms
Slabs,
billet,
blooms

Heating +
Roughing/rolling

Heating +
Roughing/rolling

1-10 mm

Thickness:
200-300 mm

Length: 500-800 m

Thin slab
casting

Thin slab
caster
(4-6 m/min)

Slabs
Thickness:
50-60 mm

Heating
+ Rolling

1-10 mm

Length: 300-400 m
Strip caster
(15-120 m/min)

Strip
casting

1-5 mm

Length: 60 m

Figure 1:

Technological development in the casting and rolling stages of the

steel production process: ingot casting, conventional continuous
casting, thin slab casting and strip casting technology.

- 132 -

Before 1960 all liquid crude steel was cast in ingots. After cooling the ingots were
reheated and processed into semi-finished products such as slabs, blooms and billets.
Continuous casting technology allowed liquid steel to solidify directly into these
semi-finished products. A continuous process replaced the traditional batch process
of pouring ingots first. Continuous casting avoided large capital expenditure on
soaking pits and breakdown mills.
The introduction of thin slab casting (1989) let the liquid steel be cast in thinner slabs
(see Figure 1). Blooms and billets cannot be produced with a thin slab caster. Thin
slab casting technology integrates casting and the finishing operations in the hot
rolling section of a steel plant. The cast slabs are charged directly into a reheater and
subsequently fed into the hot rolling mill. The reheating capacity for hot rolling the
slabs is thus reduced.
Strip casting technology makes it possible to cast the liquid steel in sizes that are
even closer to final products. Note that this limits the range of final products that can
be produced [Schors, 1996].
In a strip caster, a ladle is filled with liquid crude steel. The crude steel flows into the
tundish, in which the steel is buffered and the temperature is controlled. The tundish
feeds the liquid steel through a nozzle to the mould. The mould, which has a capacity
of about one hundred litres, is composed of one or two casting rolls. The rolls are
water-cooled. To achieve high productivity, high casting speeds are required (see
Figure 1).
There are three technical difficulties in developing strip casting technology, namely
edge containment, liquid steel feeding to the mould and properties of the rolls.
The first difficulty is related to the edges or the lateral sides of the mould. These
should form a perfect seal with the rotating rolls to prevent leakage. Premature
solidification at the edges should be avoided. This causes a poor quality at the edges
of the strip (leading to yield losses) and inhibit continuous casting.
A second difficulty is that the liquid steel should be fed to the mould in such a way
that the level of steel in the mould remains constant. A perfectly even level has to be
maintained to prevent surface defects and irregularities in the thickness of the strip.
Thirdly, the rolls that form the mould need specific attention. On the one hand they
are extreme heat exchangers (8-15 MW/m2) that should be able to handle enormous
temperature differences (200-300 C/cm). On the other hand they need to produce
crack-free strips that vary in thickness by less than 30 m.
The major advantage of strip casting technology is that the capital expenditure for
producing hot rolled coils will be considerably lower than for a conventional casting
and rolling mill: the capital-intensive hot strip mill is no longer needed3. Total capital
expenditure will be a factor 4 to 10 lower [Birat et al., 1995; Hendricks, 1995;
Lindorfer et al., 1993; Schors, 1996]. The capital costs per ton steel are estimated to
3

Flemming et al. (1988) indicate that in the case of conventional continuous casting about 70% of the
capital expenditure is spent on the hot rolling mill.

- 133 -

be about 40% lower than for conventional casting and rolling operations [Hendricks,
1995; Schors, 1996]. Experts indicate that the operating costs per ton steel are subject
to a larger uncertainty primarily due to uncertainties in material costs due to
requirements to the roll materials and the problems in edge containment [Lindorfer et
al., 1993]; Birat et al., 1995].
Energy costs will decrease. Strip casting technology considerably reduces the energy
needed to cast one ton of steel4. Because energy costs are only a small part of the
operating costs of the casting and rolling stages (see also [Schors, 1996]), the energy
cost saving is modest [Senk, 2000; Cramb, 2000]. We estimate that the energy cost
saving is an order of magnitude lower than total cost saving5.
Strip casting technology leads to considerable strategic advantages for relatively
small steel firms, i.e. 0.5 million ton steel / year. Most stainless steel firms are
relatively small. Such firms do not have a hot rolling mill at their sites: the capital
costs per ton steel are too high to be competitive. The lower capital expenditure of
strip casting allow such firms to process the cast steel into final products themselves.
Outsourcing is no longer required. Storage facilities are no longer needed. Semifinished products do not have to be transported to a hot rolling mill. Throughput
times and costs for planning and transport are considerably reduced. Strip casting
technology also opens up the cold rolled market for mini-mills [Millbank, 1995;
Hendricks, 1995; Birat, 1992].
Finally, some people argue that strip casting technology creates opportunities for the
development of new products [Senk, 2000; Robson and Thompson, 1995; Millbank,
1995; Anonymous, 1994].
4

IISI (1998a) estimates the reference primary energy consumption in the casting and rolling stages
(using a conventional continuous caster, a re-heater and a rolling mill) to be 2.3 GJp/ton steel [IISI,
1998a, pp. 74, 100-102]. De Beer (1998) estimates that the primary energy needed for casting and
rolling is 2.42.7 GJp/ton steel. This is roughly 15% of the total primary energy requirement of an
integrated mill, i.e. 19 GJp/ton [De Beer, 1998, p. 150]. IISI (1998a) estimates that thin slab casting
technology consumes roughly 1.0 GJp / ton and that strip casting technology may consume 100 MJp /
ton [IISI, 1998a, pp. 196-197]. Thus, applying strip casting technology saves about 900 MJp per ton
steel compared to the thin slab casting route (based on [IISI, 1998a]). About 60 kWh less electricity is
needed. No fuel is required for reheating the furnaces, which is about 350 MJp / ton [IISI, 1998a].
Applying a strip saves about 2.0-2.5 GJ p per ton steel compared to conventional continuous casting
(based on [IISI, 1998a]). Recent performance data from the Krefeld caster indicate that strip casting
technology requires about 400 MJp per ton steel. At the Krefeld site, the reference energy for
conventional casting and rolling is 3.2 GJp [Walter et al., 2000].
5
It is difficult to make a precise calculation of the savings in energy costs as a percentage of total cost
savings. First of all, there are differences in the cost structure of carbon steel and stainless steel.
Secondly, there are differences in cost structure of steel manufactured in an integrated mill or in a
mini-mill. Sources indicate that the total cost savings achieved when a strip caster is introduced at a
stainless steel site are 2 to 5 times higher than when a strip caster is introduced at a carbon site
[Bagsarian, 1998]. The costs of producing one ton of carbon steel vary between 225-325 US$ / ton
(see e.g. [Hamels, 1995; Schors, 1996; Faurre, 1993; Daniels and Moll, 1998]). Variation is explained
by differences in the route to produce steel (integrated or mini-mill) and differences among
geographic regions [Faurre, 1993; Schors, 1996]. We estimate that casting and rolling in an integrated
mill cost around 100 US$ per ton steel [Schors, 1996; Lindorfer et al., 1993]. Casting and rolling in a
mini-mill is estimated to cost 50 to 60 US$ per ton. Around 40% of the costs are capital costs [Schors,
1996; Lindorfer et al., 1993]. Energy costs for casting and rolling in a mini-mill are less than 10% of
the operational costs [Schors, 1996].

- 134 -

5.3. Historical development of strip casting technology

In this section, the historical development of strip casting technology is mapped up
till 1985. We start at the moment when none of the known forms of direct casting conventional continuous casting, thin slab casting, and strip casting technology had
yet been applied commercially. The history is described in three steps: the history up
till 1940; the introduction of conventional continuous casting technology (19401975); and the renewed R&D interest in casting technologies (from 1975 onwards).
Development of strip casting technology up till 1940
The idea of using a twin roll caster for casting strips directly from liquid steel
without any reheating attracted Sir Henry Bessemers attention as long ago as 1846
[Bessemer, 1891]. Bessemer was one of the important inventors in the steel industry
who made the production of cheap steel possible by the invention of the steel
converter [Jewkes et al., 1960]. With respect to strip casting technology, he started
from an existing patent describing twin roll casting of tin foil and lead. After doing
some experiments, he was so impressed with the importance of his invention, that he
immediately applied for a patent in 1857 [Bessemer, 1891]. Because Bessemers
invention of the steel converter was at that time judged as an absolute failure.
One need not, therefore, be greatly surprised that the production of continuous
sheets direct from fluid iron did not excite a great amount of enthusiasm in the minds
of tin plate manufacturers of that day; in fact, the whole scheme was simply poohpoohed and laid aside, without any serious consideration of its merits [Bessemer,
1891, p. 27]. Bessemer noted that there were difficulties in feeding the liquid steel to
the mould, with edge containment and properties of the cast strip. Since he felt his
idea was before its time and he was being pressed to make his steel converter work,
he did not continue with the development of strip casting technology [Cramb, 1989].
A sketch of Bessemers twin roll caster is shown in Figure 2.

Figure 2:

Bessemers twin roll strip caster.

- 135 -

Bessemer was aware of the activities of the American Edwin Norton. Norton had
sent him a parcel containing a small sample of a steel sheet that he claimed had been
cast in accordance with Bessemers idea [Bessemer, 1891]. In 1890, Norton
conducted experiments on twin roll casting using a pilot machine producing 3-5 mm
thick sheets [Cramb, 1989; Norton, 1940]. A fire in Nortons factory and the
mechanical difficulties he encountered made Norton decide that further development
which required substantial investment was not justified [Norton, 1940].
It was not until the 1920s that any new research was done on Bessemers idea
[Hazelett, 1966]. Only a few projects are described in literature; of these Hazeletts
activities were the most conspicuous6 [Cramb, 1989]. Hazelett picked up the idea and
successfully built a number of twin roll casters for the production of lead, aluminium
and brass strips [Hazelett, 1966; Cramb, 1989; Wolf, 1992]. Hazelett also produced
some cast steel, although problems with properties of the cast strip and with the roll
materials forced him to redesign his twin roll caster [Lippert, 1940; Hazelett, 1966;
Cramb, 1989]. First he tried to overcome these problems by switching to a single roll
caster [Lippert, 1940]. Again he encountered problems, like variability in strip
thickness, properties and edge containment. As a result Hazelett abandoned roll
casting processes in the 1940s. He concluded that a mould consisting of two moving
belts was the preferred casting method. Such a caster is nowadays known as the
Hazelett caster and is commercially for all kinds of metals but not steel [Hazelett,
1966].
1940-1975: Development of continuous casting
By 1940, all R&D activities on casting using a mould with rotating rolls were
terminated [Cramb, 1989]. By that time, the casting R&D efforts were focused on the
conventional continuous caster as suggested by Siegfried Junghans in 19277 [Jewkes
et al., 1960; Lippert, 1940; Wolf, 1992]. Junghans used a stationary mould for direct
casting, not rotating rolls as in a strip caster. The technology was first applied to a
brass plant in 1937. In 1952, Mannesman, a German machinery supplier, installed
the first continuous caster for steel. For a number of years continuous casting was
confined to small steel plants due to technical limitations. Only from 1970 onwards,
continuous casting was started to be used in integrated steel mills casting slabs
[Nasbeth and Ray, 1974; Nilles and Etienne, 1991]. With continuous casting it
became possible to cast semi-finished products directly. Substantial savings in both
capital and energy costs and a considerable increase in product yield were achieved
[Wolf, 1992].
From 1975 onwards: Renewed R&D interest beyond continuous casting
Continuous casting (was) a strong paradigm of process innovation in the steel
industry from the 1950s to the 1980s [Birat, 1999b, p. 1389]. However, from 1975
6
7

In the same time as Hazeletts work, a strip casting effort was underway in the USSR [Cramb, 1989].
For an extensive description of the history of conventional continuous casting see [Wolf, 1992].

- 136 -

onwards R&D activities concentrated on casting liquid steel in sizes that are even
closer to final products [Wolf, 1992; Birat, 1999b]. Steel-makers started to look for
innovative technologies that could extend the advantages of continuous casting. They
investigated many different moulds that would allow thinner casting. Not only
Hazelett belt casters, but also Bessemers suggestion of two rotating rolls were
picked up again. At least a hundred R&D efforts were started in the period 1975
1985 [Kubel, 1988; Tony, 1990; Birat et al., 1995; Cramb, 2000; Birat, 2000].
The Hazelett belt caster was a favourite mould among many steel firms8 [Senk, 2000;
Cramb, 2000]. This caster was already used in R&D during the 1960s [Cramb,
1989]. The technical problems that hampered the R&D activities in the 1960s were
not inherent to caster and renewed R&D efforts were initiated. However, again no
commercially viable technology emerged because machine suppliers succeeded in
developing and commercialising thin slab casting technology [Birat, 1999b]. The
German supplier SMS was the first to install a thin slab caster in 1989 at Nucor, an
American mini-mill operator. The funnel-shaped mould was new, the rest of the
design was basically the same as that of the conventional continuous caster [Cramb,
1989; Birat et al., 1995]. It was only a matter of time before other suppliers
commercialised their thin slab casters9. Steel firms stopped their R&D activities on
Hazelett casters; the Hazelett casters promised casting in the same thickness range
[Birat et al., 1995; Thompson, 2000; Senk, 2000].
The success of thin slab casting did not stop steel-makers from searching for
technologies realising even thinner casting. The R&D efforts that focused on
Bessemers strip casting technology were continued [Anonymous, 1993].

5.4. Recent development of strip casting technology

In this section, we sketch the more recent development of strip casting technology by
focusing in detail on the eleven micro-networks that can be distinguished since 1985.
The majority the strip casting R&D efforts that started between 1980 and 1985 were
terminated.
We use the terms hot model, pilot caster and industrial scale caster to denote the
different scales of the equipment used. A hot model denotes a small pilot plant.
Liquid supply is only a few hundred kilograms; strip width is typically 1025 cm.
The term pilot caster is used to denote a caster with restricted liquid supply, but a

Krupp Stahl (Germany), Mannesman (Germany) , British Steel (UK), Bethlehem Steel and US Steel
(US), Nucor (US), Nippon Steel Corporation and Mitsubishi, Kawasaki Steel, Kobe Steel, Sumitomo
Metals, and Hitachi Zosen (all Japan) (see e.g. [Kubel, 1988; Evans et al., 1988; Cramb, 1989; Tony,
1990; Birat and Steffen, 1991; Birat et al., 1995]).
9
Mannesman (Germany), Voest Alpine Industrieanlagebau (Austria), Danieli (Italy) and Sumitomo
Heavy Industries (Japan) [Steffen and Tacke, 1999]. The introduction of thin slab casting facilitated
the increase of steel manufacturing in mini-mills, first in the US and later also in other countries [Birat
et al., 1995].

- 137 -

larger strip width, i.e. 80-160 cm. An industrial scale caster has both a large liquid
supply, larger than 60 ton, and a larger strip width 80-160 cm.
Allegheny Ludlum (US) and Voest Alpine Industrieanlagebau (Austria)
During the early 1980s, the US steel-maker Allegheny Ludlum started to develop a
single roll caster for speciality steels. The major arguments were to do away with
expensive hot strip mills and to produce new grades of steel10 [McManus and Berry,
1993; Hess, 1991; McAloon, 1991; Isenberg-OLoughlin, 1988]. Allegheny
preferred a single roll caster, because of earlier R&D experience with rapid
solidification [Birat et al., 1995]. In 1984, Allegheny Ludlum surprised the steel
industry by claiming that they had been successful in strip casting. Alleghenys claim
placed strip casting technology on the agenda of many boards of directors of steelmaking firms [Birat et al., 1995; Thompson, 2000; Cramb, 2000; Birat, 2000].
Voest Alpine Industrieanlagebau, an Austrian machine supplier, had been involved
in the development of thin both slab casting and strip casting technology since the
mid-1980s. The argument for developing both innovative technologies was the
potential capital savings [IISI, 1993]. Thin slab casting was not seen as the ultimate
direct casting technology and therefore Voest also performed single and twin roll
casting experiments [Hohenbichler, 2000]. Twin roll casting was found to be more
difficult than single roll casting. R&D activities focused on carbon steel [Birat and
Steffen, 1991; Kubel, 1988].
For continuing R&D activities, Allegheny needed a machine supplier that could
build a pilot caster. Voest was approached. Allegheny and Voest signed an
agreement in 1988. Voests experience with casting carbon steel complemented
Alleghenys activities in the field of stainless steel. Aim of the joint effort was to
bring strip casting technology towards commercialisation. In 1988, the firms
expected that the process could be commercialised in the mid-1990s for stainless
steel [Ritt, 1998; McAloon, 1991; Lindorfer et al., 1993; Hess, 1993]. In 1990, a
decision was made to build a pilot caster at Alleghenys Lockport site. The pilot
caster was called Coilcast [McAloon, 1991]. The firms claimed that capital costs and
operating costs could be reduced by 75% compared to the costs of conventional
casting and rolling [Hess, 1991; Lindorfer et al., 1993].
Coilcast had been tested for two years before it was suddenly shut down in 1994
[Schriefer, 1997]. Allegheny dropped all its R&D activities without any public
statement. There was no longer confidence that steel could be cast successfully. The
development had already taken more than 13 years [Williams, 1999; Schriefer, 1997;
Birat, 2000; Cramb, 2000].

10

Allegheny Ludlum was driven by competition with Allied Signal. Allied Signal has a commercial
twin roll caster at its disposal (see [McManus, 1997]). Allied Signal aimed at producing a product that
could compete with a market that was served by Allegheny [Cramb, 2000].

- 138 -

Armco (US)
In 1983, the US Department of Energy (DOE) decided to support feasibility studies
on strip casting technology. Allegheny Ludlum brought the technology to the
attention of DOE. DOEs was interested, because of the substantial energy savings
expected. Allegheny Ludlum was not interested in government support. Their R&D
activities were beyond the stage of a feasibility study [Williams, 1999; Cramb,
1989]. Four teams did acquire R&D support for performing such a feasibility study.
One team consisted of Westinghouse, a producer of electronic consumer products,
and the US steel-maker US Steel. Westinghouse studied a single roll caster. In 1984,
Westinghouse was selected for a second phase study. DOE preferred this project
because it was targeted at carbon steels that promised the largest energy savings
potential. US Steel withdrew their contribution when they decided to invest in the
development of thin slab casting (also supported by DOE)11 [Kubel, 1988; Hess,
1991; Cramb, 1989]. Westinghouse approached Armco. Armco had attempted
casting R&D using a Hazelett caster and was interested to pursue the development of
a single roll caster [Cramb, 1989]. The contract was extended to a third (1988) and a
fourth contract (1991). Armco became the prime contractor [Isenberg-OLoughlin,
1988]. DOE contributed around 12 million US$, 70% of the total expenditure.
Westinghouse and Armco retained the right to patent their findings [Willimas, 1999;
Isenberg-OLoughlin, 1994].
In late 1993, when Armco was half-way through the fourth contract, it suddenly
terminated its participation. The triggering event was that Armcos carbon steelmaking division was sold. Armco became a firm producing specialty steels
[Isenberg-OLoughlin, 1994; Ritt, 1998; Sheridan, 1997]. Furthermore, R&D results
from other micro-networks made Armco believe that a twin roll caster was the
preferred solution [Cramb, 2000; Williams, 1999; Ritt, 1998].
Other R&D activities in the US
Not only Allegheny and Armco, but also other US steel firms explored strip casting
technology during the early eighties. Bethlehem Steel, for instance, started twin roll
casting R&D in 1981. Around 1986, they formed a partnership with Armco, as well
as with Weirton, and Inland Steel. Argonne National Laboratory financed by DOE
was involved in this effort to deal with the edge containment problem [Cramb, 1989;
Hess, 1991; Isenberg-OLoughlin, 1994; Isenberg-OLoughlin, 1988]. The effort
ended because they could not find a solution for this difficulty and could not cast
reproducible grades of steel [Birat et al., 1995; Kuster, 1996]. By the year 1994, all
strip casting R&D activities in the US were terminated.
In April 1998, US strip casting R&D was restarted. Carnegie Mellon University
(CMU) and the American Iron and Steel Institute (AISI) has started a 3-year multi11

In 1984, DOE awarded US Steel and Bethlehem Steel a 30 million US$ grant to develop a thin slab
caster using a Hazelett caster [Kubel, 1988; Cramb, 1989; Hess, 1991].

- 139 -

partner R&D project within the Industries of the Future Initiative of the US
Department of Energy. Research institutes, steel-manufacturers and machine
suppliers are involved12. Aim of the project is to analyse the properties of strip cast
steel in order to assess the possibilities of producing new grades of steel and other
advantages of strip casting technology to the US steel industry [Cramb, 1998].

Usinor Sacilor (France) and Thyssen (Germany)

The French integrated steel-maker Usinor Sacilor started experimenting with strip
casting in the mid-80s [Grosjean et al., 1993]. Both single and twin roll casters were
investigated. After some numerical simulations and design concepts, Usinor decided
to build a hot model at the research institute IRSID in 198613. A twin roll caster was
found to be more promising [Birat and Steffen, 1991]. Between 1986 and 1990,
IRSID received some R&D support, 10 to 15%, via the research programme of the
European Community on Steel and Coal (ECSC) [Birat, 2000].
The German integrated steel-maker Thyssen started research on strip casting
technology around the same time as Usinor. Thyssen worked together with the
Institut fr Bildsame Formgebung (RWTH) in Aachen. A hot model was available in
1988 [Grosjean et al., 1993]. Experiments were done on twin and single roll casters.
Thyssen also considered twin roll casting more promising [Simon et al., 1997].
Thyssen received some support by the German Bundeslnder. Several projects were
supported by the ECSC though they did not specifically focus on strip casting, but
were also valuable for thin slab casting technology [Senk, 2000].
Thyssen and Usinor had co-operated in R&D before and in 1989 it was decided to
merge their R&D activities in strip casting technology14 [Grosjean et al., 1993;
Hendricks, 1995]. The argument was to share the investment costs of the pilot caster
[Birat, 1999a; Senk, 2000]. The joint venture is known as Myosotis. The pilot caster
was located at Isbergues, a stainless steel site of Usinor. The plant was built by the
French machine-builder Clecim. After a planning and construction period of 20
months, the first experiments started in June 1991 at a ladle capacity of 10 tons. The
ladle capacity was gradually increased. In October 1995, a complete ladle of 92 tons
was cast. The main interest was in the casting of stainless steel, but carbon and
silicon steel casting experiments were also performed.
During 1997 several articles on the results of the Myosotis project were published
(see e.g. [Legrand et al., 1997a; Legrand et al., 1997b; Simon et al., 1997]). By then,
the researchers were convinced that it was no longer the question whether the
12

R&D institutes: CMU (US), Max Planck Institute (Germany). Associate organisation: AISI. Steelmakers: AK Steel Corporation (US), British Steel R&D lab (UK), Dofasco (Canada), LTV Steel
Company (US), National Steel Corporation (US), US Steel / US X (US). Suppliers: SMS Steel (US),
Kvaerner Metals (US) [Cramb, 1998].
13
IRSID used to be the R&D institute of the French national steel firms. It became Usinors corporate
R&D lab in 1988 [Birat, 1999a].
14
Thyssen and Usinor also co-operated in the area of thin slab casting technology. Together with
SMS, a German machine supplier, they developed the Casting-Pressing-Rolling process [Birat et al.,
1995; IISI, 1998a].

- 140 -

technology would have an industrial future, but only when it would become
operational [Legrand et al., 1997b].
Centro Sviluppo Materiali and Acciai Speciali Terni (Italy)
The steel research institute Centro Sviluppo Materiali15 (CSM) and the Italian steelmaker Acciai Speciali Terni (AST) started strip casting R&D in 1985 [IISI, 1993;
CSM, 2000]. First, the feasibility of the process was studied. In a second phase
(1987-1988) a hot model was designed and constructed in Rome. There was then a
third phase in which a pilot caster was designed and built at ASTs steelworks in
Terni in 1989 [Shin, 1990; Birat and Steffen, 1991; Macci and Mollo, 1995]. The
Italian engineering firm INNSE was involved in constructing the pilot caster. In
1993, industrial operation was foreseen by the end of 1996 [IISI, 1993]. Some results
of the pilot caster were reported in 1995. However, surface quality did not meet the
requirements of its conventional equivalent [Macci and Mollo, 1995]. Until 1995,
AST and INNSE paid for the research done by CSM. The research at CSM was
partly supported by the ECSC, i.e. 20 to 30% of the budget [CSM, 2000].
In 1995, the Austrian machine-builder Voest Alpine joined the CSM/AST micronetwork as an equal partner. Voest became involved in upgrading the Terni facility
to the size of an industrial scale caster. Voests R&D experience in the former
partnership with Allegheny made them an attractive partner for CSM/AST [CSM,
2000; Hohenbichler, 2000]. During the summer of 1997, the Terni plant was supplied
with two new 30 ton ladles [Bagsarian, 1998]. Expenditure between 1995 and 1997
was considerably higher than in the period 1985 to 1995. In this phase, no further
application was made for ECSC support [CSM, 2000]. One did obtain financial
support by the European Commission (EC) for building a new industrial scale caster
at Terni in September 1996. Improving energy efficiency was the main rationale
behind the ECs decision to support the demonstration facility16.
Krupp Stahl (Germany) and Nippon Metal Industry (Japan)
The Japanese Nippon Metal Industry had been developing strip casting since 1980.
Joint R&D activities between the German steel-making firm Krupp Stahl and Nippon
Metal Industry started in 1986. The activities focused on a mould using rolls of
unequal diameter [Shin, 1990]. In 1990, two pilot casters were operational, one in
Germany and one in Japan. Krupp acquired support by the ECSC and the German
Bundeslnder [Pfeifer et al., 1993; Birat and Steffen, 1991]. After the merger of
Krupp with Hoesch in 1992, changes in the organisation of Krupp and the idea that
considerable extra R&D was still needed made Krupp decide to stop [Millbank,
15

Up till 1994, CSM was the national research institute of the Italian steel industry. In 2000, CSM is a
privatised research institute in the field of steel and metallurgy. CSM has different shareholders. The
stainless steel firm Acciai Speciali Terni (AST) owns 15%.
16
THERMIE programme / 4th FWP / IN/00124/96. Title: Abatement of energy consumption in the
steel hot strip production through a near net shape casting process. Prime contractor: AST, Terni.
Duration: September 1996-November 1999 [Cordis, 2000].

- 141 -

1995; Cramb, 2000; Senk, 2000]. Nippon Metal stopped its strip casting activities as
well [Birat et al., 1995].
Eurostrip: Merged micro-networks
In 1994 the Italian steel firm AST was taken over by the German steel-maker Krupp.
CSM and AST continued their strip casting R&D activities. In 1997, Thyssen
merged with Krupp. The newly formed Thyssen Krupp Steel had a stake in the
Italian micro-network and the French German Myosotis project. The activities in
these two micro-networks were formerly merged in 1999. Experienced researchers
and engineers from AST and CSM, Voest, Thyssen Krupp Steel, and Usinor and
IRSID were brought together.
Due to the merger, stainless steel producer Krupp Thyssen Nirosta (KTN) became a
sister company of the Italian AST. Whereas, KTN had some strip casting R&D
experience, they carried out two ECSC projects between 1991 and 1995, they were
rather sceptical about the possibilities of strip casting technology. After KTN visited
the Terni caster, interest in the Italian strip caster grew. In December 1998, it was
decided to build an industrial caster at KTNs Krefeld site instead of at ASTs Terni
site. The Krefeld site was particularly advantageous because it did not have a hot
rolling mill [Walter et al., 2000]. The European Commission permitted the financial
support to be used to build an industrial scale caster in Krefeld instead of in Terni
[CSM, 2000; Senk, 2000]. In September 1999, Thyssen Krupp Steel, Usinor and
Voest formed a firm called Eurostrip. Voest is the exclusive supplier of Eurostrip
strip casting technology. In December 1999, the first stainless steel was cast. The
Krefeld casters annual capacity will be enlarged from 100,000 tons/year to around
400,000 tons/year within 2 or 3 years [Anonymous, 1999; Thyssen Krupp Steel,
1999]. The Myosotis project was formerly stopped in 1998. R&D activities in Terni
are continued for developing carbon steel strip catsing. The Eurostrip partners are
involved in ECSC supported co-operative multi-partner projects on carbon steel.
Europstrip partners expect that the carbon steel market for the exploitation of strip
casting will develop rapidly from 2001 [Hohenbichler et al., 2000].

Nippon Steel Corporation and Mitsubishi Heavy Industries (Japan)

The consortium of Nippon Steel Corporation (not to be confused with Nippon Metal
mentioned earlier) and the machine-builder Mitsubishi Heavy Industries was formed
in 1985 [IISI, 1993; Millbank, 1995]. They concentrated on stainless steel.
Mitsubishi and Nippon Steel were financially equal partners [Tacke, 1999; Birat,
1999a; Cramb, 2000]. Their first hot model was built in 1986. In 1988, they extended
this model to a pilot caster at Hikari. In 1993, some promising results were reported.
Nippon Steel and MHI announced that the focus would shift towards studying the
economic feasibility of the technology and the building of an industrial scale caster
[IISI, 1993]. In September 1997 it was reported that the industrial scale caster had
reached the design capacity of 400,000 ton/year. In October 1998, Nippon Steel was

- 142 -

selling strip cast stainless steel. Nippon Steel reported that the greatest advantage of
the strip caster was that the production time was reduced to less than an hour.
Casting and finishing operations were linked by the strip caster and thus transport
was no longer required [Furukawa, 1997; Furukawa, 1998].

Broken Hill Proprietary Company (Australia) and Ishikawajima-Harima

Heavy Industries (Japan)
The Australian Broken Hill Proprietary Company (BHP) and the Japanese machinebuilder Ishikawajima-Harima Heavy Industries (IHI) had an established business
relationship, when they started co-operative R&D activities relating to strip casting
in 1985. In 1990, a pilot caster at Unanderra was realised [Furukawa, 1994]. At first,
BHP investigated casting stainless steel, like most of the other micro-networks.
Focus shifted to carbon steel when the project moved on towards the realisation of an
industrial scale caster [Opalka, 1998]. In December 1994, the industrial scale caster,
called Project M, was constructed at Port Kembla. The main purposes of the plant
were to proof the durability of the rolls, the consistency of product quality and
improve the accuracy of process control [Furukawa, 1994]. BHP achieved its
breakthrough in December 1997 with the production of its first saleable carbon steel.
Up till November 1998, BHP and IHI were secretive about Project M [Opalka, 1998;
Birat et al., 1995]. Then, they announced that Project M was very close to
commercialisation [Abbel and Moonen, 1998]. They claimed that the advantage of
Project M is that the strip caster shortens the production process, both physically and
in time, and integrates casting with rolling operations. BHP was determined to
exploit this advantage in new mini-mills in the East-Asian region [BHP, 1998;
MacLeay, 1998; Millbank, 1995]. Confronted with the Asia crises, however, BHP
decided not to build mini-mills. Early in 2000 it was announced that Nucor, the US
mini-mill steel firm that was also the first to build a thin slab caster, had signed a
letter of intent with BHP and IHI for the establishment of a joint venture for the
worldwide licensing of BHPs strip casting technology. Start-up of the first
commercial facility is expected in 2001 [Anonymous, 2000; Nucor, 2000; BHP,
1999].
Hitachi Zosen and Pacific Metals (Japan)
Hitachi Zosen, a Japanese machine-builder, experimented with twin roll casting
between 1981 and 1985 on an irregular basis. More serious activities started in 1985.
As a result, Hitachi wanted to build a pilot caster in 1987. The firm started looking
for a steel partner. Pacific Metals, a relatively small stainless steel company (and
subsidiary of Nippon Steel), joined [Birat et al., 1995]. The pilot caster, called
Hachinoe, became operational in 1990 [Freuhan, 1994; Parodi, 1993; Shin et al.,
1995]. Around 1994, the US steel-maker Inland Steel became involved [Anonymous,
1994]. Inland had developed an electro-magnetic edge containment system during
the 1980s. Inland Steel had obtained some experience with the technology in their

- 143 -

partnership with Bethlehem Steel, Armco and Weirton (1986) [Cramb, 1989;
Isenberg-OLoughlin, 1994]. An agreement was signed with Hitachi Zosen to
determine the feasibility of the edge dam on Hitachis pilot caster [IsenbergOLoughlin, 1994]. The agreement however did not result in any adaptations to the
pilot caster. Hitachi and Pacifics current intentions remain unclear.
British Steel (UK)
British Steel started research in 1986 at its Teeside Technology Centre. A small hot
model was used for casting carbon and stainless steel. In parallel with this
investigation, a marketing study identified target markets for the technology [IISI,
1993]. In 1990, the existing hot model was scaled up to a larger one. Davy, a UK
engineering firm, built the machine. British Steel experimented with some innovative
technologies which were all abandoned [Birat et al., 1995; Kuster, 1996].
British Steel Stainless was interested in the possibilities of strip casting, because they
wanted to integrate casting with cold rolling. At that moment hot rolling was outsourced. When British Steel Stainless merged with Avesta Sheffield, a Swedish
stainless steel manufacturer in 1992, a hot rolling facility was provided and there was
less need for strip casting technology [Wilkinson, 1998].
British Steel however continued their R&D. Their objective changed: they merely
wanted to maintain their expertise. British Steels wants to be an educated
customer, so that it can act as an informed buyer once strip casting technology
becomes available [Thompson, 1999; Wilkinson, 1998; Millbank, 1995].
Most of the strip casting R&D carried out at British Steel was funded by the ECSC.
British Steel was involved in three individual projects. It has also been involved in
two ECSC supported multi-partner projects, together with the Eurostrip partners, on
strip casting carbon steel [Birat and Steffen, 1991; Millbank, 1995; Thompson,
2000].

Project Bessemer and the Industrial Materials Institute (Canada)

Between 1986 and 1989 the Canadian steel industry identified strip casting as an
interesting technology for co-operative research. In a thorough survey of world-wide
casting R&D activities, the enormous potential for capital and operating cost
reductions by strip casting technology was highlighted [IISI, 1993]. The Canadian
Industrial Materials Institute (IMI) had already started strip casting R&D in 1987
driven by the feeling that there was strong interest among Canadian steel-makers
[Hamel, 1999]. And indeed, a consortium of six Canadian steel firms was formed in
198917. The consortium is called Project Bessemer. The consortium asked IMI to
carry out R&D on strip casting carbon steels [IISI, 1993; Birat et al., 1995]. The
Canadian steel firms preferred a twin roll caster because of its potential for good strip
surface quality. Furthermore, carbon steel was chosen as the main grade, simply
because none of the steel firms produced stainless steel. A hot model was installed
17

Dofasco, Stelco, Ipsco, Ivaco, Ispat-Sidbec and Algoma (80% of the Canadian steel production).

- 144 -

during the summer of 1992 by Hatch et Associs, a Canadian engineering firm [IISI,
1993; Hamel, 1999].
Between 1989 and 1992, the consortium developed a research programme, that was
financially supported by the Canadian National Research Council. Goal of the
programme was to develop a strip casting process that could produce low cost, high
quality carbon steel. In January 1997, an agreement was signed with Hatch et
Associs. They got the right to design future strip casting plants [Project Bessemer,
1997; Hamel, 1999]. However, the research programme was stopped in 1998
[Steffen and Tacke, 1999]. There are no signs that any of the firms in the Bessemer
Consortium is engaged in taking the next step. This may be due to a number of
factors: BHPs progressed in casting carbon steel; three of the Canadian steel-makers
invested in thin slab casters; there is no machine-builder involved; and finally
substantial financial resources are required. Investments in an industrial caster would
require an additional 65 million US$ [Bagsarian, 1998; Hamel, 1999; Millbank,
1995].
The major argument for the steel-makers to start the co-operative R&D programme
in the first place was to keep abreast of strip casting technology. The steel-making
firms are pleased with the dividends that the programme has paid to date. Each firm
has spent a fairly restricted amount of resources and the knowledge coming back to
them has been considerable [Isenberg-OLoughlin, 1994; Millbank, 1995; Hamel,
1999].
Pohang Iron and Steel Company (South Korea)
The Korean Pohang Iron and Steel Company (POSCO) and the Research Institute of
Industrial Science & Technology (RIST) started work on strip casting in 198918
[Shin et al., 1995]. POSCOs efforts in strip casting have started relatively late,
because it is only since 1987 that POSCO has pursued an active R&D strategy. After
a period in which POSCO concentrated on expanding its steel production capacities,
POSCO wanted to reinforce its competitiveness by strengthening its R&D activities.
One of the long-term projects started was a strip casting project [Kang, 2000].
POSCOs aim is to develop an industrial scale strip caster, that is less capital
intensive than conventional casting and rolling [Shin et al., 1995; Kang, 2000]. The
focus is primarily on stainless steel. Davy International, a British engineering firm,
was responsible for the design of the hot model [Anonymous, 1993]. In 1994, a pilot
caster was constructed [Millbank, 1995]. At the moment no steps have been taken
towards the design and construction of an industrial scale caster. Sources at RIST
state that the technical problems have been solved on the pilot scale, however, the
uncertainties about the techno-economic performance make POSCO refrain from
investing in an industrial scale caster [Kang, 2000]. Others speculate that the crises in
Asia are discouraging POSCO from engaging in large capital expenditure [Birat,
1999a; Cramb, 2000].
18

POSCO established RIST in 1987. RIST is a legally independent R&D institute. POSCO supplies
60 to 70% of RISTs total budget [Kang, 2000].

- 145 -

5.5. Analysing the development

In mapping this history of strip casting technology we presented elements, which are
of interest to our analysis of the case study. First we address the questions formulated
about the technology network, then we focus on the various micro-networks and
subsequently we discuss the results of the R&D activities.
Technology network
What is the composition of the technology network?
Before 1975, localised R&D activities were performed by certain individuals the
three main ones being Bessemer, Norton and Hazelett. Between 1975 and 1985 a
large number of R&D efforts were launched which all tried to extend the advantages
of conventional continuous casting to even thinner casting. Different types of casting
moulds were explored. There were also a growing number of efforts weighing up the
pros and cons of single or twin roll strip casting. By 1985, strip casting R&D efforts
were announced in almost each major steel-making country [Cramb, 1989; Cramb,
2000]. By 1990, the strip casting technology network had acquired a clear status. The
majority of the R&D efforts were stopped when the decision had to be made to
continue R&D and thus to invest in a hot model. The efforts that continued were
veiled in secrecy because of the strategic importance of the R&D activities [Senk,
2000]. The eleven micro-networks described in Section 5.4 resulted.
Figure 3 gives the technology network of strip casting technology from 1982
onwards. Six of the eleven micro-networks are still active. Four of the eleven micronetworks stopped their activities. Two micro-networks merged their R&D activities.
Eurostrip resulted. Three of the six active micro-networks are operating an industrial
scale caster and claim to be ready to launch the innovative process technology.
The eleven micro-networks cover a wide geographic area. Steel firms from the US,
Canada, Western Europe, South-Korea and Australia are involved. In four of the
eleven micro-networks, co-operating actors within a micro-network come from
different countries. The three most advanced micro-networks are located in Japan,
Western Europe and Australia.
Japanese and US steel firms were the first to start strip casting R&D efforts during
the late seventies [Isenberg-OLoughlin, 1988; Kubel, 1988; Senk, 2000]. Although
Japanese firms are still active in various micro-networks, a substantial number of
early Japanese efforts were stopped [Birat et al., 1995; Anonymous, 1993]. They
passed the job on to Nippon Steel Corporation, which is by far the largest steel firm
in Japan [Cramb, 2000; Birat, 2000]. The US lost its original leading position. Only
recently US actors have become active again. A co-operative R&D programme was
started by Carnegie Mellon University [Cramb, 1998]. In 2000, the US mini-mill

- 146 -

steel firm Nucor formed a joint venture with the BHP and IHI to commercialise BHP
/ IHIs strip casting technology.
EUROSTRIP
Armco

Also carbon
steel
Carbon

Stop

Voest Alpine *
Allegheny *

Stainless

Stop

+ Clecim
Usinor * / IRSID

Stainless

Thyssen * / RWTH
+ Voest
CSM / AST

Stainless

Stainless

Stop

Krupp * / Nippon *
Nippon Steel
Corp.*/ Mitsubishi*

Stainless

+ Nucor
BHP * / IHI

Carbon

+ Pacific Metals
Hitachi *

Stainless

British Steel *

Stainless

+ Hatch et Associs
Bessemer C * / IMI

Stop

Carbon

+ Davy
Stainless

POSCO * / RIST

1982
Preliminary
phase

Figure 3:

1986

1990

Hot model

Pilot caster

1994
Industrial scale
caster

1998
Sale of steel cast on
industrial scale caster

Technology network of strip casting technology. The size of the R&D

equipment used hot model, pilot caster, and industrial scale caster
is included. A preliminary stage and a commercial stage (the steel
cast is sold) have been added. The R&D activities of Allegheny and
Voest and the co-operation of Thyssen and Usinor are each counted
as one micro-network. Actors in normal letters are steel-making firms.
Actors in italics are machine suppliers or engineering firms. Actors
with underlining are R&D institutes. Asterisks (*) indicate the driving
actors within micro-networks.

To what extent and how often do micro-networks exchange knowledge/information?

In spite of the fact that R&D activities were strategic, there has been exchange and
communication between the various micro-networks. In addition to the exchanges at
international conferences and meetings there have been more intensive contacts.

- 147 -

Clearly, technical details were generally not disclosed. However, the exchange of
knowledge and R&D experience have played a role in the continuation of R&D
activities within the various micro-networks [Thompson, 2000; Senk, 2000; Birat,
2000; Cramb, 2000]. It is important to know whether others have succeeded in
solving technical difficulties. Then you do not need the details to do the work itself.
Furthermore, it is important to know the position of a firms micro-network among
the competing micro-networks for instance to create firm-internal support for
continuing R&D activities.
Are there dominant micro-networks in the technology network?
None of the micro-networks had a continuing decisive influence on the development
of strip casting technology. There was some early leadership by Allegheny. Their
claim of success in 1984 was instrumental in drawing the attention of other steel
firms (and their boards of directors) [Birat et al., 1995; Cramb, 2000; Thompson,
2000; Senk, 2000]. For a long time, Allegheny was considered to be the most
advanced micro-network (see e.g. [Kubel, 1988; Hess, 1991; Robson and Thompson,
1995]). By the early 1990s, it had become clear that a number of steel firms had
invested seriously in developing strip casting technology, although none of them was
dominant. In the last two years, it has become clear that there are three front-running
micro-networks, Eurostrip, Nippon Steel and Mitsubishi, and BHP and IHI. These
three micro-networks may prove the feasibility and advantages of strip casting
technology in the coming two or three years. The other micro-networks and the steel
industry in general are waiting to see how their casters will perform [Tacke, 1999;
Bagsarian, 1998; Cramb, 2000; Senk, 2000].
Micro-networks
How are the various micro-networks made up?
Figure 3 shows the type of actors that have been involved in the eleven micronetworks. It also denotes who took the initiative in the separate micro-networks.
A first observation is that steel-making firms generally took the lead [Birat et al.,
1995]. The machine suppliers, Hitachi, Voest and Mitsubishi, are exceptions. In eight
of the eleven micro-networks, a machine-builder or engineering firm became
involved in later stages of the R&D activities. In the early eighties it was clear to
most actors that it would take time for strip casting technology to be developed and
that considerable steel-making expertise was required. This explains why machine
suppliers were not pressing for the development of this technology in the way they
had pressed for the development of thin slab casting technology [Birat, 1999b].
In developing strip casting technology, control in the various micro-networks is
typically exerted by the steel manufacturers. Machine suppliers or engineers were
involved when steel manufacturers had the plan to build a pilot or industrial scale
caster. However, steel firms own the crucial patents. The suppliers involved are thus

- 148 -

dependent on the steel firms for selling strip casting technology [Birat, 2000;
Thompson, 2000; Senk, 2000].
A second observation is that there are no universities in the technology network. The
role of research institutes that are financed with public money is also limited; the
only ones are the German RWTH and the Canadian IMI. The Italian CSM is a
specialised private research firm.
What motivates actors to start and / or stop R&D activities?
Steel firms are interested in strip casting technology mainly because they want to
reduce the capital intensity of the steel plant. They want to lower the costs of
converting liquid steel to the coiler (see also Figure 1) [Cramb, 1989; Birat et al.,
1995; Tacke, 1999; Senk, 2000; Thompson, 2000; Cramb, 2000]. This advantage had
been clear to Bessemer. Still, it took more than a century before the technology
network emerged.
It took time for steel-makers to recognise the possibilities and advantages of direct
casting. This recognition became possible because the industry evolved itself. The
steel industry today is not comparable to the steel industry in Bessemers days.
Conventional continuous casting was then not possible. Mini-mills and grades such
as stainless steel did not yet exist. When conventional continuous casting matured
during the seventies19, the advantages of skipping even more steps in casting and
rolling became apparent, especially to relatively small firms, e.g. stainless steel firms
and mini-mills.
Between 1975 and 1985, the search for innovative casting technologies emerged.
Technologists and researchers agreed internationally that this was an interesting
route [Birat, 1999b]. A process emerged in which the possibilities of such casting
technologies were explored. Expectations were confirmed and articulated. The
number of R&D efforts increased during the 1980s. A number of factors induced this
process: the established need for more compact technologies [Herbertson et al., 1992;
Birat 1992]; the availability of research money within steel firms [Birat, 1999b;
Cramb, 2000; Senk, 2000], some government R&D support in the US and Europe;
and last but not least Alleghenys claim of success in 1984 [Birat et al., 1995]. The
production of new grades of steel based on R&D experience in rapid solidification
was also feeding the interest in strip casting [McManus and Berry, 1993; Senk,
2000].

19

Experts indicate that the improvement of conventional continuous casting is continuing. The R&D
activities between 1950 and 1980 were so numerous that a wealth of ideas are still available that can
improve the costs or quality of conventionally cast steel [ Birat, 1999b; Cramb, 2000; Senk, 2000].

- 149 -

Table 1:

Current status of the eleven micro-networks.

Steel firm

Scale
achieved

Status

Comments

Armco

Machine
supplier or
engineer
-

n.a.

Stop 1991

Allegheny

Voest

Pilot

Stop 1995

Usinor /
Thyssen

Clecim

Industrial
scale

Continuation
Eurostrip

CSM / AST

Voest

Industrial
scale

Continuation
Eurostrip

Krupp /
Nippon Metal

Pilot

Stop 1992

Nippon Steel
Corporation

Mitsubishi

Industrial
scale

Continuation

BHP

IHI

Industrial
scale

Continuation

Pacific
Metals

Hitachi
Zosen

Pilot

Continuation

British Steel

British Steel
engineering

Hot model

Continuation

Bessemer
Consortium /
IMI

Hatch et
Associes

Hot model

Stop 1998

POSCO /
RIST

Davy

Pilot

Continuation

Armco stopped its single roll caster

R&D, because it altered it business
policy: carbon steel division was sold.
Coilcast was not successful enough to
justify continued R&D activities. It
had taken too long and had proved too
expensive.
After Thyssens merger with Krupp
(1997), the Myosotis project and the
Italian micro-network were merged.
Eurostrip was formed to achieve
industrial maturity and
commercialisation.
The industrial scale caster at Terni is
used by Eurostrip partners for carbon
steel strip casting.
After Krupps merger with Hoesch
and problems with the low
productivity of the caster and with the
quality of the product R&D activities
were ceased.
Current status of R&D is veiled in
secrecy. Nippon Steel seem to intend
to apply the technology themselves
and not to sell/license the technology.
Clarity is expected within a year.
BHP aimed at building mini-mills in
South-East Asia. Early in 2000, BHP
announced a joint venture with Nucor,
a US mini-mill operator, and IHI for
further commercialisation.
Problems with edge containment.
R&D continues on pilot scale. Future
plans are unclear. Some say efforts
have ceased.
After merger with Avesta, arguments
for R&D changed. British steel wants
to be an educated customer.
None of the firms wanted to invest at
the level needed for up-scaling. Three
Canadian firms invested in thin slab;
and BHP is successful in carbon steel
strip casting.
The Korean economy collapsed; it
had become difficult to finance
further up-scaling. Uncertainties about
the performance make POSCO
refraining from investments in an
industrial scale caster.

- 150 -

This combination of mutually reinforcing factors added momentum to the trend and
strengthened internal support in steel firms to perform explorative R&D. Integrated
steel-firms with a stainless steel division20 or major independent stainless steel
manufacturers21 continued R&D activities using hot models. As indicated before, a
large number of these early strip casting efforts were stopped. The technology
network resulted [Birat et al., 1995; Thompson, 2000; Senk, 2000].
In spite of the promising advantages of strip casting a lot of efforts were stopped.
The three early explorers of strip casting technology stopped. Most of the efforts that
were started in the early eighties and four of the eleven micro-networks ended their
strip casting R&D activities. We comment briefly on the arguments for stopping
R&D activities.
Bessemers concluded that his idea was before its time. He also foresaw serious
technical difficulties. Both Norton and Hazelett stopped or changed their focus in
R&D due to technical problems.
The abandonment of R&D efforts that were started during early eighties is directly
related to decision for investing in a hot model. Most firms could not or were not
willing to spend the money needed for taking this step [Senk, 2000; Birat, 2000;
Cramb, 2000; Thompson, 2000]. Some additional arguments were heard as e.g. we
cannot solve technical problems, we are waiting for others to develop, the quality of
the steel was uncertain, and finally we stop because others are also stopping [Tacke,
1999; Cramb, 2000]. However, the bottom line was that firms did not want to spend
the financial resources required.
Table 1 summarises the status of the eleven micro-networks. It shows that there are
two main arguments why actors stopped or slowed down R&D activities. First of all,
the budget needed to continue R&D on a larger scale explains why actors lost
interest. Secondly, R&D activities were overruled by all sorts of strategic decisions
that have nothing to do with the development of the technology as such. However,
continuation of R&D activities did become superfluous when for instance divisions
or firms were sold, bought or merged.
How much money is spent and by whom?
Table 2 summarises the expenditure on strip casting R&D by the eleven micronetworks as far as relevant data are available. Due to the large differences in the
degree of technical complexity of the R&D equipment used it is impossible to give a
standard indication of the ratio of the cost of a hot model and the cost of an industrial
scale caster. According to rough estimates, an industrial scale caster is 10 to 20 times
more expensive than a hot model.

20
21

E.g. Nippon Steel Corporation, Usinor, British Steel, Thyssen.

E.g. Allegheny, Nippon Metal, Pacific Metals, AST.

- 151 -

Table 2:

Expenditure on R&D and government R&D support.

Micronetwork

Scale

Armco1
Allegheny2
CMU3
Usinor /
Thyssen4

Pilot
Pilot
Pilot

CSM / AST5
Voest
Krupp Stahl6 /
Nippon Metal
Eurostrip7
Nippon Steel
/ Mitsubishi8
BHP / IHI9
Pacific Metals
/ Hitachi
British Steel10
Bessemer C /
IMI11
POSCO/
RIST12
Total
Total estimate

Total
expenditure
[M US$]
17
10-15
1
about 10

Of which
govt. support
[M US$]
12
0.7
1.4

Government support
granted by:

70
70
10-15

US Department of Energy

Industrial
Pilot
Industrial
Pilot

100
15-20
n.a.
n.a.

limited
4.1
5
0.6

Industrial
Industrial

50
110

Industrial
Pilot

200
n.a.

n.a.
-

n.a.
-

Hot model
Hot model

4
26

1.7
13

40
50

Pilot

20

570
500 700

40

20-30

US Department of Energy
ECSC (see Table 3).
Limited support German
Bundeslnd.
French government.
ECSC (see Table 3).
European Commission
ECSC (see Table 3).
Limited support German
Bundeslnd.

ECSC (see Table 3)

National Resources
Canada

Data concerning Armcos private investment in twin roll casting is unknown. 2 In an interview in
1988 Ludlums chairman said that of the order of 10 to 15 million US$ worth of investment was
paying off in a new process [Isenberg-OLoughlin, 1988]. Long running projects like Coilcast
involve of tens of million US$ [Hohenbichler, 2000]. 3 CMU = Carneggie Mellon University [Cramb,
1998; Cramb, 2000]. 4 [Birat, 1999a; Birat, 2000; Senk, 2000]. 5 Derived from [CSM, 2000; Cordis,
2000]. 6 [Pfeifer et al., 1993]. 7 [Thyssen Krupp Stahl, 1999]. 8 Only costs of the industrial scale
caster [Furukawa, 1997]. 9 [MacLeah, 1998]. 10 Derived from [Thompson, 2000; Cordis, 2000]. 11
[Bagsarian, 1998]. 12 Only pilot caster. Total expenditure is higher but is not made public [Millbank,
1995; Kang, 2000].

Table 2 clearly indicates that micro-networks having an industrial scale caster

invested the most. Strip casting experts estimate that expenditure on industrial scale
casters is of the order of magnitude of 100 to 200 million US$ (see e.g. [Cramb,
1998; Bagsarian, 1998]). In most of the micro-networks, machine-builders did not
make substantial investments, up to 10% [Birat, 1999a]. Only Mitsubishi and Voest
were equal partners.
We estimate that worldwide about 500-700 M US$ was spent, which is 35 to 50
million US$ per year. When this is compared to the annual R&D expenditure by the

- 152 -

iron and steel industry, we conclude that roughly 1 to 2% of the industrys annual
R&D expenditure goes to strip casting technology22.
What important decisions are made with regard to the direction of technological
development?
Each of the eleven micro-networks in Figure 3 had to take two important decisions:
the grade of steel and whether to use a single or twin roll mould.
Almost all micro-networks studied strip casting of stainless steel. Only two of the
eleven micro-networks focused seriously on carbon steel. Some argued that strip
casting of carbon steel was technically more difficult, although it had been primarily
an economic incentive that determined the preference for strip casting stainless steel.
The relatively small stainless steel firms have the highest capital cost advantage. In
addition, the productivity of strip casters (annual output) for large carbon steel
manufacturers was too low [Schors, 1996].
When the strip casting technology network emerged during the eighties, the single
roll caster was the first and the most seriously studied technology by the two micronetworks in the US. This preference for the single roll caster had to do with earlier
experience of rapid solidification research and the wish to develop new steel grades
[Cramb, 1989; Birat et al., 1995]. Most of the other micro-networks evaluated both
single roll and twin roll casters. The single roll caster lost ground to the twin roll
caster. Eight micro-networks preferred twin roll casters, because of their higher
productivity and of the possibility to cast steel with symmetric properties [Senk,
2000; Cramb, 2000].
Materialisation
What is the rate of development and what steps in up-scaling can be distinguished?
Roughly 140 years ago, Bessemer suggested that two rolls could be used as a mould
for casting liquid steel. Whereas Bessemer, Norton and Hazelett all had some hot
model scale equipment at their disposal, conventional continuous casting and thin
slab casting technology were commercialised, before three micro-networks
succeeded in bringing strip casting technology almost to the point of innovation.
Thin slab casting and strip casting technology both benefited from the renewed
interest in casting R&D between 1975 and 1985. When thin slab casting technology
was introduced (1989), R&D activities on Hazelett casters was stopped. Strip casting
22

The iron and steel industry in the US, Canada, Japan, Australia and the EU spent roughly 2.5 billion
US$ / year on R&D (1985 1994 average) [OECD, 1997]. Note that the budget of the micronetworks working on an industrial caster scale is not supplied by the R&D budget only. It is
interesting to compare annual spending on strip casting technology with the budget the governments
of IEA/OECD countries spent on industrial energy-efficiency R&D. Over the last 15 years,
IEA/OECD countries have spent 220 million US$ per year on industrial energy-efficiency R&D.
Average annual expenditure on strip casting is as much as 15 to 25% of the annual government budget
on industrial energy-efficiency R&D [IEA, 1997].

- 153 -

technology remained as the sole option to realise direct casting of even thinner slabs.
There has been a spurt in developing strip casting technology in the last 15 to 20
years. Figure 3 indicates the steps that various micro-network took in up-scaling strip
casting technology. It should be noticed that hot models continue to be used if larger
equipment is available. Figure 4 gives the scale of the most advanced equipment
used in the eleven micro-networks.
Industrial scale caster
100

Usinor / Thyssen

Ladle capacity (t)

80

CSM / AST

Eurostrip

NipponSteel / MHI

60

BHP / IHI
40

Pilot caster
20

Hot model
(Bessemer Consortium)
0

200

(Allegheny)
(Krupp / Nippon)
Posco
British Steel
Hitachi / Pacific

400

600

800

1000

1200

1400

1600

1800

Strip Width (mm)

Figure 4:

Cross-section of the development in 1999: showing the state-of-the-art

equipment. Three micro-networks operate on near commercial scale:
BHP and IHI, Nippon Steel and MHI, and Eurostrip (CSM, AST and
Usinor and Thyssen). Firms between brackets indicate micronetworks that stopped R&D activities.

Figure 3 illustrates that the three most advanced micro-networks all needed 15 to 17
years to reach industrial scale operation. There is some variation in the length of the
different stages, though each stage took roughly three to seven years.
Both the Hitachi/Pacific micro-network and the POSCO micro-network continued
their R&D activities on a pilot scale for more than six years. British Steel and the
Bessemer Consortium have a longer hot model stage; both micro-networks had
defensive arguments for investing in strip casting R&D.
What are the perceived performance characteristics of the technology?
Bessemer recognised the capital advantages of his idea [Bessemer, 1891]. An
interesting question is whether strip casting could have emerged earlier? On the one
hand one can argue that the proven performance of conventional continuous casting
and the introduction of thin slab casting (which induced the growth of mini-mills)
triggered the need for casting strips of even thinner thickness ranges at even lower

- 154 -

capital costs [Herbertson, 1992]. Strip casting technology became the next step to
take since the search routine was to cast as close to final requirements as possible.
On the other hand one can say that the existing paradigm of conventional
continuous casting may have kept strip casting from taking off earlier.
One should however take into account the fact that two things changed during the
last decades of the 20th century. First of all, the industry itself changed considerably:
stagnation in world steel output, rapid rise of mini-mills and the introduction of new
grades such as stainless steel. Secondly, knowledge became available that could be
used to solve the major technical problems, i.e. the liquid steel feeding, edge
containment and roll material requirements [Thyssen Krupp Steel, 1999; Birat 2000].
Without the advances in material knowledge (e.g. regarding ceramics and copper
alloys) and the availability of computers and process control, it would have been
much more difficult to make strip casting an engineering reality [Senk, 2000; Cramb,
2000; Thompson, 2000; Birat, 2000].
The eleven micro-networks did not publish very much about the performance
characteristics of strip casting technology. It is clear to all actors that strip casting
will lead to lower capital costs and will open the market for cold rolled steel for
mini-mills. It is also certain that strip casting reduces the amount of energy needed
per ton steel, although this advantage did not drive technological development. The
precise cost advantages of strip casting technology are still hidden away among
unpublished data 23.

5.6. Discussion
For data gathering we were dependent on the information publicly available and on
the willingness of experts involved. We tried to acquire a complete collection of
written material on strip casting technology and consulted both magazines, journals
and conference proceedings. The Corus Library in Ijmuiden was visited.
There are intrinsic shortcomings in the written documentation. First of all, the
attention paid to the various micro-networks was not equal. Some micro-networks
have been more intensively described in literature. Secondly, the secrecy that
surrounds micro-networks make it theoretically possible that micro-networks remain
unnoticed in public articles for a long time. An example is that BHPs R&D activities
concerned with strip casting became known in 1994, whereas they started in 1985.
Thirdly, the information given is sometimes confusing or contradictory. For example
the terms used to denote the type of equipment are not uniform, nor are they
unambiguous. Fourthly, the issues we are interested are not included in information

23

The available data on cost performance should be used with care. They reflect what the researchers
want the technology to be; crucial estimations had to be made e.g. regarding the costs and life time of
refractory materials [Cramb, 2000; Birat et al., 1995].

- 155 -

available to the public. This necessitated personal communication with experts

involved.
By interviewing experts we were able to gain a better understanding of the status of
the technology network. There are some drawback in interviewing experts like for
instance a dependency on experts willingness (time) and capability (what do they
know themselves). Their response is clearly biased by their perception of the
development history. Furthermore, experts can only tell what they are allowed to tell.
In four micro-networks, personal communication turned out to be impossible.
Experts did not respond or were not willing to respond.
However we are convinced that our analysis is robust. We have identified all major
micro-networks and have obtained insight into the current leading micro-networks,
the actors involved in the eleven micro-networks, their arguments for investing and
into the expenditure done.

5.7. The effect of government R&D support

Now that we have analysed the historical development of strip casting technology,
we focus our attention on the effect of government R&D support. Table 2 showed the
financial contribution that government made to various micro-networks. R&D
support was granted in the US, Canada and in Europe. In Japan, however, there was
no government support. Two-thirds of the European support was channelled via the
research programme of the European Coal and Steel Community (ECSC). It should
be realised that the R&D budget of the ECSC is raised by a levy on steel prices in
Europe and as such cannot be regarded as pure government R&D support.
In this section, we characterise the role and the effect of government R&D support in
the US, in Canada and in Europe.
US
We estimate that financial support by the US Department of Energy (DOE) made up
30 to 60% of the total expenditure there24. Government R&D support did not
accelerate the technologys development.
The role played by DOE R&D support in the development of strip casting
technology can be summarised thus:
- DOE was passive in identifying strip casting technology. Allegheny brought the
technology to DOEs attention.
- Allegheny was not interested in obtaining support from DOE because Allegheny
was further ahead.
- DOE had a preference for the micro-network of Armco. This effort was focused
on carbon steel (bulk grade) and thus promised larger energy savings. DOE felt
justified in granting R&D support on the grounds of improving energy efficiency.
24

Armco micro-network, Allegheny micro-network and the recent CMU project.

- 156 -

Argonne National Laboratory, financed by DOE, became involved in a

consortium of amongst others Weirton and Bethlehem Steel. Argonne had
expertise in electromagnetic casting. R&D activities did not materialise.
During the 1980s DOE supported US Steel first of all in strip casting technology
(see Armco micro-network). Having abandoned this effort, US Steel did again
succeed in acquiring DOE support, now for developing a Hazelett caster.
Armcos major argument for abandoning their strip casting project was that all
carbon steel divisions had been sold. Therefore, DOEs R&D support did not
have any chance of materialising the technology.
In 1998, Carnegie Mellon University started a multi-partner R&D programme on
strip casting technology supported by DOE. The R&D activities permit the actors
to become familiar with recent insights in strip casting R&D and the possibilities
for application in the US.

Canada
The Canadian National Research Council supported 50% of the total costs of the
multi-year research programme initiated by the Bessemer Consortium and IMI.
Government R&D support did lead to additional R&D activities although it did not
accelerate the technologys development. The Canadian steel firms aimed at
acquiring an understanding of the technology.
Europe
We estimate that R&D support from the European Coal and Steel Community
(ECSC), the European Commission (EC) and some national governments made up
less than 10% of the total expenditure on strip casting R&D in Europe25. A recent
overview of the research areas covered in the ECSC research programme (19861999) and a list of funded strip casting projects [Ball, 2000] allowed us to make a
more detailed evaluation of the effect of ECSC support on the R&D activities of the
European strip casting micro-networks26.
Strip casting and other casting technologies have been on the agenda of the ECSC
since a working group was established in 1985. The Commission and the European
steel industry agreed that both conventional continuous casting and new casting
technologies were areas of R&D which were important for the modernisation of the
European steel industry [Evans et al., 1988; Birat and Steffen, 1991]. Figure 5 gives
an overview of the budget spent on casting R&D between 1986 and 1999.

25

CSM / AST micro-network, Usinor / Thyssen micro-network, Krupp / Nippon micro-network,

Eurostrip, and the British Steel micro-network.
26
A telephonic survey was conducted among the actors involved in ECSC projects (Table 5). The aim
was to gain insight into the effect of ECSC R&D support. Claustahl was not consulted because their
strip casting projects are not linked to any of the micro-networks. Five actors were consulted. One
actor did not want to co-operate. In one case, it was impossible to get in contact with the persons
involved in the ECSC projects.

- 157 -

45,0
Strip casting 1986-1990:
- # projects: 15
- # contractors: 21
- support/project: 500 kEuro

40,0

Strip casting 1981-1995:

- # projects: 13
- # contractors: 21
- support/project: 550 kEuro

Total projects costs (million Euro)

35,0

30,0

Strip casting 1996-1999:

- # projects: 3
- # contractors: 10
- support/project: 1,150 kEuro

25,0

ECSC support
Actor's share

20,0

15,0

10,0

5,0

0,0
1986-1990
Strip casting

Figure 5:

1991-1995

1996-1999

1986-1990
Thin slab
casting

1991-1995

1996-1999

1986-1990
Other casting

1991-1995

1996-1999

ECSC support for strip casting, thin slab casting and other casting
(mainly conventional continuous casting) in three time-periods. More
specific information regarding the ECSC supported strip casting
projects is given in the three boxes [Ball, 2000].

In the period 1986-1999, roughly 2% of the ECSC budget was spent on projects
earmarked as strip casting R&D; 20% of the ECSCs casting R&D budget was
spent on strip casting technology. The ECSCs strip casting expenditure was about
18 million US$.
Figure 5 indicates that in the last time-period (1996-1999), there was a considerable
decrease in the number of strip casting projects. Since the beginning of the 1990s, the
European Commission has fostered multi-national multi-partner projects in order to
fulfil the requirements of subsidiarity [Ball, 2000]. As a result both the number of
partners in strip casting project and the average funding per project increased. These
recent multi-national projects are all devoted to carbon steel strip casting. This is
explained by the simple fact that the steel firms are less willing to co-operate in
competitive R&D on stainless steel.
Table 3 shows that ECSC support allowed European research institutes such as
Claustahl University and the Max Planck Institute (MPI) to perform strip casting
R&D. ECSC support was granted to the UK micro-network of British Steel, the
Italian micro-network, and the German / French micro-network in which Thyssen
and Usinor co-operated.

- 158 -

Table 3:
Firm /
R&D institute

Overview of ECSC support (based on [Ball, 2000; Cordis, 2000])27.

1

TU Clausthal
MPI
VDeH

2
3
1

Total
budget
[M US$]2
0.9
1.2
0.8

British Steel

4.0

1.7

1986 2000

Thyssen

1.1

0.6

1994 2000

IRSID (Usinor)

3
2

3.8
0.6

1.4
0.3

1986 1990
1994 2000

11.5

4.1

1987 1997

1.4
1.5

0.6
0.7

1985 1987
1991 1995

27.0

11.6

CSM
Krupp
KTN (Krefeld)
Total

1
2

ECSC
support
[M US$]
0.5
0.7
0.5

Time-frame

1991 1995
1989 2000
1994 1997

Remarks

Co-ordinator multi-partner
carbon steel project
Involved in 2 multi-partner
carbon steel projects
Multi-partner carbon steel
projects
Multi-partner carbon steel
projects
EC 4th framework project NOT
included

Number of projects in which the firm or research institute was involved. 2 Assuming 1 US$ 1
Euro. 3 One project was carried out by Krupp Stahl AG (1985-1887), during Krupps co-operation
with Nippon Metal Industry (Japan). Two projects were performed by Krupp Thyssen Nirosta (KTN)
(1991-1995), a sister company of AST. Some people involved in the Krupp / Nippon micro-network
were also involved in the activities at KTN.

The R&D activities at Claustahl and MPI were not related to any of the R&D
activities in the three European micro-networks. There are no signs that knowledge
from Claustahl and MPI has been of major importance for these micro-networks. The
Verein Deutsche Eisenhutteleute (VDEh) co-ordinated one of the multi-partner
carbon steel projects [Steffen, 2000].
The other actors mentioned in Table 3 were already mentioned in Section 5.4. The
Italian micro-network, CSM, received the largest contribution from the ECSC. CSM
lost interest in ECSC support when R&D activities moved towards the pilot scale. A
similar pattern occurred in the French / German micro-network. There was support
for Usinors IRSID in the early days of the development (until 1990). When the
R&D co-operation with Thyssen was started, ECSC support was no longer acquired.
British Steel received ECSC support from the start of their efforts up till today.
ECSC funds formed a considerable part of the total strip casting budget of British

27

The ECSC support for the projects in Table 3 is lower than the ECSC support for strip casting
mentioned in Figure 5. This is because projects focused on casting in a range of 10 to 40 mm are
excluded in Table 3. In Table 3 only two multi-partner projects are included. A third, even more
recent multi partner project is not included. Thyssen, IRSID, CSM and British Steel participate in this
project which was started in 1998. This project is also for carbon steels. Budget as specified in the
contract is 1.7 M US$ of which ECSC support is 1.0 M US$ [Thompson, 2000; Senk, 2000].

- 159 -

Steel, 40% (see Table 2). Without ECSC support, they would have stopped earlier
[Thompson, 2000].
We conclude that:
- In the majority of the ECSC R&D projects on strip casting, additionality has been
limited. We estimate that 60% of the projects would have been performed
anyway.
- Furthermore, the projects that are supported by the ECSC are generally only
loosely related to the core of private R&D expenditure on strip casting. ECSCsupported projects cover side-lines of R&D activities. There is a minor chance
that spill-over will occur to the core R&D activities.
- For one micro-network, British Steel, it is highly probable that the projects would
not have been performed without ECSC support. Activities would also have
stopped earlier.
- Steel firms apply for ECSC support in early stages of a technologys
development. The steel firms become less interested in ECSC support when
development moves towards a larger scale. They then decide to refrain from
ECSC support.
- Taking into account the operating procedures of the ECSC research programme
and the fact that steel firms determine what kind of R&D projects are supported,
we conclude that the majority of the ECSC casting R&D resources are used to
optimise conventional continuous casting technologies.
- It is highly likely that the EC demonstration project (Krefeld caster) would have
been realised even without EC support of 5 million US$.
- The ECSC supported two multi-partner projects on carbon steel, in which all the
major European steel firms involved in strip casting technology micro-networks
co-operated. The major gain is that experienced researchers co-operate and have
the possibility to learn from each other and share expertise in a pre-competitive
area of R&D.

5.8. Conclusion
Strip casting technology is at the threshold of becoming a commercialised
technology. In this chapter we evaluated the effect of government R&D support on
the development of strip casting technology by making a detailed analysis of the
networks involved.
Government has supported 5 to 10% of the total expenditure by eleven micronetworks since 1980. Total expenditure by all actors is estimated to be about 500 700 million US$. Six of the eleven micro-networks received government R&D
support. In three micro-networks R&D support was substantial, more than 40% of
total expenditure within that micro-network. Two of these micro-networks stopped
R&D activities. The third micro-network continued R&D although intentionally at a

- 160 -

modest scale. Three micro-networks may prove the commercial feasibility of strip
casting technology within the next two or three years. These three micro-networks
did not receive (or only very minor) government R&D support.
All in all, the effect of government R&D support on the development of strip casting
has been minimal. The technology did not become available any earlier as a result of
R&D support. Government R&D support did not influence the direction of
technological development.
The case study generated convincing evidence for this conclusion. First, R&D
support has been minimal or absent in micro-networks that are ahead in developing
the technology. R&D support has been the largest in micro-networks that were not
operating at the frontier of developing the technology. Steel firms became less
interested in government R&D support when activities reached a stage that was more
sensitive to competition. Secondly, an analysis of ECSC support to strip casting
R&D indicated that about 60% of the projects would have been carried out without
that support. Thirdly, government supported R&D projects that addressed side lines
or that covered pre-competitive co-operative R&D. These were, thus, only loosely
related to the core R&D activities. Finally, R&D support did not have a decisive
effect in directing technological development to improved energy efficiency. From
an energy efficiency point of view, government should have supported the
development of carbon steel strip casting technology because of much larger
production volumes. Only three micro-networks studied carbon steel intensively.
Two of these micro-networks received government R&D support. However, both
these micro-networks ceased their R&D activities.
The effect of government R&D support has been minimal mainly because the
development of the technology proved to have a strong momentum of its own. After
renewed interest in strip casting technology developed between 1980 and 1985, a
large number of R&D efforts were started. Strip casting technology promised to
extend the advantages of conventional continuous casting. A process emerged in
which the possibilities of the technology were explored in a large number of efforts.
Expectations were confirmed and articulated. Government R&D support was one of
the factors in this process, though it was not decisive. Eleven micro-networks
continued their R&D activities. A robust technology network emerged. All micronetworks were initiated and controlled by steel manufacturers. The contribution of
public research institutes was very limited. All the steel firms and machine suppliers
involved considered strip casting technology as a key-technology for future
steelmaking. Strip casting technology affects the core of the steel production process.
The dominant advantage, especially for stainless steel producers, is that the
technology should make it possible to further reduce the capital investment in hot
rolling facilities. Engaging or not engaging in its development is, thus, only loosely
related to energy-efficiency improvements or government R&D support.

- 161 -

In our analysis we included not only technical, cost and energy performance
characteristics, but we broadened our analysis to include networks of actors. As a
result, we could evaluate the effect of R&D support quite accurately. It allowed us to
weigh up the effect of R&D support against other incentives that influenced steel
firms and machine suppliers to be involved, or to continue or stop. By studying the
networks of actors, we were able to sense what might have happened without
government R&D support.

Acknowledgements
Leopold Stratenus is thanked for his valuable preliminary analysis of the historical R&D development
of strip casting technology (see [Stratenus, 1999]). The authors like to express their gratitude for a
number of experts who have been important in providing information, comments and suggestions: Mr.
J.-P. Birat (IRSID, Usinor, France), Mr. G. Thompson (Corus, UK), Mr. D. Senk (Thyssen Krupp
Stahl, Germany), Mr. A.W. Cramb (Carneggie Mellon University, US), Mr. J. Ball (ECSC, Belgium),
and someone from CSM (Italy). Research for this chapter was financially supported by the Dutch
National Research Programme on Global Air Pollution and Climate Change (NRP / RIVM, Bilthoven,
The Netherlands). The authors are grateful to Sheila McNab for linguistic assistance.

References
Abbel, G. (1999), personal communication, Hoogovens Research and Development, Ijmuiden, the
Netherlands, January 21st 1999.
Abbel, G., W.H.L Moonen (1998), Current developments in continuous casting in Europe, in:
Proceedings 3rd European Conference on Continuous Casting, October 20-23rd 1998, pp. 925940.
AISI (2000), glossary, American Iron and Steel Institute (AISI), Washington D.C., US,
www.steel.org/learning, March 23rd 2000.
Anonymous (2000),
Nucor
plans
strip-casting joint
venture,
in:
New Steel,
www.newsteel.com/news/000207.htm, February 22nd 2000.
Anonymous (1999), Thyssen Krupp, Usinor, Voest Alpine form strip casting partnership, in: New
Steel, November 1999, p. 10.
Anonymous (1994), Research on direct casting of thin strip continues, in: Steel Times, June 1994, vol.
222, p. 229.
Anonymous (1993), Strip casting of stainless steel: the results of four pilot strip caster systems, in:
Steel Times, vol. 221, no. 11, pp. 479-480.
Ball, J. (2000), personal communication, ECSC RTD Programme, European Commission, Brussels,
Belgium, February 23rd / March 10th 2000.
Bagsarian, T. (1998), Unveiling project M, in: Iron Age New Steel, vol. 14, no. 12, pp.56-58.
Beer, J. de, E. Worrell, K. Blok (1998), Future technologies for energy-efficient iron and steel
making, in: Annual Review of Energy and Environment, vol. 23, pp. 123-205.
Beer, J.G. de, M.T. van Wees, E. Worrell, K. Blok (1994), ICARUS-3. The potential of energy
efficiency improvement in the Netherlands up to 2000 and 2015, Utrecht University,
Department of Science, Technology and Society, report no. 94013, Utrecht.
Berg, J.C. van den (1996), The forces that drive research and development in the steel industry, in:
Ironmaking and Steelmaking, vol. 23, no. 4, pp. 309-316.
Bessemer, H (1891), On the manufacture of continuous sheets of malleable iron and steel, direct from
the fluid metal, in: Journal of the Iron and Steel Institute, pp. 23-41.

- 162 -

BHP (1999), Project M to be commercialised, press release, January 28th 1999, Melbourne,
Australia.
BHP (1998), Strip casting current status, Melbourne, Australia, www.bhp.com.au/steel/project_m,
September 7th 1998.
Birat, J.P. (2000), personal communication, Institut de Recherches de la Siderurgie (IRSID / Usinor),
Maizires-les-Metz, France, March 20th / April 29th 2000.
Birat, J.P. (1999a), personal communication, Institut de Recherches de la Siderurgie (IRSID / Usinor),
Maizires-les-Metz, France, January 7th / March 23rd 1999.
Birat, J.P. (1999b), Innovation in steel continuous casting: past, present and future, in: La Revue de la
Metallurgie-CIT, vol. 96, no. 11, pp.1389-1399.
Birat, J.P., R. Steffen, S. Wilmotte (1995), State of the art and developments in near-net-shape casting
of flat steel products, final ECSC report / EUR 67671, Brussels / Luxembourg.
Birat, J.P. (1992), Direct casting of thin strip, in: Endeavour, vol. 16, no. 3, pp. 110-116.
Birat, J.P, R. Steffen (1991), Current R&D work on near-net-shape continuous casting technologies in
Europe, in: La Revue de la Metallurgie CIT, vol. 88, no. 6, pp. 543-556.
Blok, K, W.C. Turkenburg, W. Eichhammer, U. Farinelli, T.B. Johansson (eds) (1996), Overview of
energy RD&D options for a sustainable future, Report of phase II of EC project JOU2-CT
93-0280, EUR 16829 EN, Office for Official Publications of the European Communities,
Brussels / Luxembourg.
Cordis (2000), Cordis database on EC supported R&D projects, European Commission, Brussels,
Belgium, www.cordis.lu, March 15th 2000.
Cramb, A.W. (2000), personal communication, department of Material Science and Engineering,
Carnegie Mellon University, Pittsburgh, US, April 27th / April 28th / July 3rd 2000.
Cramb, A.W. (1998), Strip casting: anticipating new routes to steel sheet, Project proposal to AISI,
Carneggie
Mellon
University
(CMU),
Pittsburgh,
US,
www.neon.mems.cmu.edu/cramb/aisi/proposal.html, September 7th 1998.
Cramb, A.W. (1996), Current trends in steel casting for strip, in: Proceedings of the symposium on
melt spinning, strip casting and slab casting, TMS Annual meeting, Anaheim, US, February
4-8th 1996.
Cramb, A.W. (1989), New steel casting processes for thin slabs and strip. A historical perspective, in:
Transactions of the Iron and Steel Society, vol. 10, pp. 61-76.
CSM (2000), personal communication (anonymous), CSM, Rome, Italy, June 28th 2000.
Daniels, B.W., H.C. Moll (1997), The base metal industry: Technological descriptions of processes
and production routes; Status quo and prospects, University of Groningen, Centre for Energy
and Environmental Studies (IVEM), report no. 92, Groningen.
Evans, P.R.V., G. Dehalu, J. Ferron, Ch. Roedrer (1988), Progress on direct casting of thin steel
products in the European Community, in: Proceedings of the 4th International Conference on
Continuous Casting, Brussels, May 17-19th 1988, Verlag Stahleisen mbH, Dsseldorf, pp.
697-715.
Faurre, H.A. (1993), Development, state of the art and future of steel making, in: Metallurgical Plant
and Technology International, no. 3, pp.32-41.
Flemming, G., P. Kappes, W. Rohde, L. Vogtman (1988), Rolling of continuously cast strips and the
technical consequences with respect to the design of hot strip production plants, in:
Metallurgical Plant and Technology International, vol. 11, no. 1, pp. 16-35.
Fruehan, R.J. (1994), Effect of emerging technologies on the competitiveness in the steel industry, in:
Iron and Steelmaker, pp. 17-22.
Furukawa, T. (1998), Nippon Steel starts new caster, in: American Metal Market, November 10th
1998.
Furukawa, T. (1997), Asians continue aggressive push to stainless, in: American Metal Market, April
15th 1997.
Furukawa, T. (1994), A new entry in strip casting, in: New Steel, October 1994, pp. 42-43.
Grosjean, J.C., J.L. Jacquot, J.N. Damasse, H. Litterscheidt, D. Senk, W. Schmitz (1993), Thin strip
casting experiments at Irsid and Thyssen Stahl AG, in: Iron & Steelmaker, August 1993, pp.
27-32.

- 163 -

Hamel, F.G. (1999), personal communication, group leader strip casting, Industrial Materials Institute,
National Research Council, Boucherville, Canada, February 12th 1999.
Hamels, D. (1995), Making steel Trends in lowering energy use and reducing environmental
emissions (In Dutch: Staalfabricage Trends gericht op beperking energieverbruik en lagere
milieu-belasting), in: Bouwen met Staal, vol. 122, no. 1/2, pp. 19-24.
Hazelett, R.W. (1966), The present status of continuous casting between two moving belts, in: Iron
and Steel Engineer, pp. 105-110.
Hendricks, C. (1995), Strip casting technology A revolution in the steel industry, in: Metallurgical
Plant and Technology International, vol. 18, no. 3, pp. 42-49.
Herbertson, J., R.I. Olivares, W. Blejde (1992), Requirements of refining processes when casting thin
strips, in: Steel Times, November 1992, pp. 520-523.
Hess, G.W. (1993), Direct strip casting sputters along, in: Iron Age, vol. 9, no. 5, pp. 24-26.
Hess, G.W. (1991), Thin slab casting a fad? Fat chance!, in: Iron Age, vol. 7, no. 10, pp. 20-23.
Hohenbichler, G. (2000), personal communication, Voest Alpine Industrieanlagebau, Linz, Austria,
March 16th 2000.
Hohenbichler, G., U. Albrecht-Frh, J.M. Damasse, R. Capotosti (2000), EUROSTRIP. Future
perspectives for the strip casting process for flat steel production, presented at CCC 2000,
June 2000, Voest Alpine Industries, Linz, Austria, Innovation Session, paper no. 6.
Holschuh, L.J. (1995), International co-operation and the future of steel, in: Metallurgical Plant and
Technology International, vol.18, no. 3, pp. 28-31.
IEA (1997), IEA energy technology R&D statistics 1974-1995, International Energy Agency (IEA),
Paris.
IISI (1998a), Energy use in the steel industry, International Iron and Steel Institute / Committee on
technology, Brussels.
IISI (1998b), World steel in figures, International Iron and Steel Institute, Brussels.
IISI (1993), Near net shape casting. Technical exchange session, International Iron and Steel Institute
/ Committee on Technology, Brussels.
IPCC (1996), IPCC Second assessment. Climate Change 19995, International Panel on Climate
Change, Geneva.
Isenberg-OLoughlin, J. (1994), The wheels are turning, in: 33 Metal Producing, vol. 32, no. 6, pp.
25-30.
Isenberg-OLoughlin, J. (1988), Solidification technologies: a status report, in: 33 Metal Producing,
vol. 26, no. 9, pp.16-17, 57.
IWG (1997), Scenarios of US carbon reductions. Potential impacts of energy technologies by 2010
and beyond, Interlaboratory Working Group, prepared for US Department of Energy,
Washington D.C..
Jewkes, J., D. Sawers, R. Stillerman (1960), The sources of invention, MacMillan & Company Ltd. /
St. Martins Press Inc., London / New York.
Kang, T, (2000), personal communication, team manager strip casting project team, Research institute
of Industrial Science & Technology, Pohang, Korea, July 5th 2000.
Kubel, E.J. (1988), Direct thin strip casting, in: Metal Progress, no. 9, pp. 55-62.
Kuster, T. (1996), Advances in strip casting carbon and stainless, in: New Steel, November 1996.
Legrand, H., L. Vendeville, J.M. Damasse, R.W. Simon, D. Senk, C. Mollers (1997a), Myosotis:
coule continue de bandes minces, in: La Revue de la Metallurgie CIT, vol. 94, no. 6, pp.
769-777.
Legrand, H., L. Vendeville, J.M. Damasse, R.W. Simon, D. Senk, C. Mollers (1997b), Myosotis direct
thin strip casting Flat minis, in: Steel Times International, May 1997, p. 24.
Lindorfer, B., K. Schwaha, J. Spiess, G. Houze (1993), Economics of strip casting on future steel
production, in: Steel Times, July 1993, pp. 304-305, 307.
Lippert, T.W. (1940), Continuous casting, in: The Iron Age, vol. 11, pp. 44-47.
Macci, F., A. Mollo (1995), Status of the CSM casting process, in: La Revue de la Metallurgie - CIT,
June 1995, pp. 789-794.
McAloon, T.P. (1991), Coilcast offers second chance to Allegheny Ludlum, in: IS&M, vol. 18, no. 11,
p. 36.

- 164 -

McLeay, J. (1998), BHP nears holy grail of small-scale steelmaking, in: the Australian, October
22nd 1998.
McManus G.J. (1997), Casting strip at sixty miles an hour, in: Iron and Steel Engineer, vol. 74, no. 2,
p.69.
McManus G.J., B. Berry (1993), Nippon, Mitsubishi, Armco take the lead in strip casting, in: New
Steel/ Iron Age, vol. 9, no. 11, pp.18-20.
Millbank, P. (1995), Steel strip casting inches to commercialisation, in: MBM Metals Technology
Supplement, April 1995, pp. 8-9, 11, 13, 15, 17.
Nasbeth, L., F.R. Ray (eds) (1974), The diffusion of new industrial processes. An international study,
Cambridge University Press, Cambridge.
Nilles, P., A. Etienne (1991), Continuous casting today status and prospects, in: Metallurgical Plant
and Technology International, no. 6, pp. 56-76.
Norton, E. (1940), Notes on experiments in rolling fluid steel, in: The Iron Age, vol. 11, pp. 50-51.
Nucor (1999), Nucor signs letter of intent for new steelmaking technology, press release, January 27th
1999, Nucor Corporation, Charlotte, US.
OECD (1992), Technology and the Economy. The Key Relationships, Organisation for Economic Cooperation and Development, Paris.
OECD (1997), Research and development expenditure in industry, Organisation for Economic Cooperation and Development, Paris.
Opalka, W.A. (1998), BHP achieves breakthrough with project M, in: Iron & Steelmaker, December
1998, p. 19.
Parodi, G.G. (1993), Near net shape casting of flat products, in: Proceedings of the 27th IISI Annual
Meeting and Conference, Paris, October 3-6th 1993, IISI, Brussels, Belgium, pp. 82-88.
Project Bessemer (1997), Press release, January 16th 1997, IMI / NRCan, Boucherville, Canada.
Ritt, A (1998), Harvesting results from government steel projects, in: New Steel, October 1998.
Robson, A.L., G.L. Thompson (1995), Direct casting of thin strips, in: Materials World, May 1995,
pp. 222-224.
Pfeifer, H., M. Dubke, K-H Hanke (1993), Direct strip casting with unequal diameter rolls, in: Steel
Times International, vol. 17, no. 1, pp. 14-15.
Schors, L.L. (1996), Why mini-mills give the US huge advantages in steel, in: The McKinsey
Quarterly, no. 2, pp. 44-55.
Schriefer, J. (1997), A merger creates a global specialty metals power, in: New Steel, August 1997.
Senk, D. (2000), personal communication, Thyssen Krupp Stahl AG, Duisburg, Germany, April 28th
2000.
Sheridan, J.H. (1997), More steel productivity gains ahead, in: Industry Week, September 15, 1997,
pp. 86, 89-90, 92, 95-96.
Shin, Y.K., T. Kang, T. Reynolds, L. Wright (1995), Development of twin roll strip caster for sheet
steels, in: Ironmaking and Steelmaking, vol. 22, no. 1, pp. 35-44.
Shin, Y.K. (ed.) (1990), International Conference on new smelting reduction and near net shape
casting technologies for steel. SRNC90, Korean Institute of Metals and the Institute of
Metals, 14-19th October 1990, Pohang, Korea.
Simon, R.W., D. Senk, C. Mollers, H. Legrand, L. Vendewille, J.M. Damasse (1997), Direct strip
casting on the Myosotis industrial pilot plant, in: Metallurgical Plant and Technology
International, vol. 20, no. 3, pp.78-82.
Steffen, R (2000), personal communication, Verein Deutsche Eisenhtte (VDEh), Dsseldorf,
Germany, April 18th 2000.
Steffen R., K.-H. Tacke (1999), Stand des Bandgiessens von Stahl, in: Stahl und Eisen, vol. 119, no.
6/7, pp.129-132.
Stratenus, L.M. (1999), The historical development of strip casting. Department of Science,
Technology and Society, Utrecht University, 99023, Utrecht, The Netherlands.
Tacke, K-H. (1999), personal communication, head of the department of metallurgy, Max Planck
Institut fr Eisenforschung, Aachen, Germany, March 1st 1999.
Thompson, G.L. (2000), personal communication, Teeside Technology Centre, British Steel,
Cleveland, UK, April 6th 2000.

- 165 -

Thompson, G.L. (1999), personal communication, Teeside Technology Centre, British Steel,
Cleveland, UK, February 12th 1999.
Tony, W.A. (1990), Near-net-shape casting no longer considered and advanced technology, in: Iron &
Steelmaker, January 1990, pp. 22-26.
Thyssen Krupp Steel (1999), Press releases, April 19th 1999 and December 13th 1999, Duisburg,
Germany, www.thyssenkrupp.com/eng.information/ hm9/presse, January 25th 2000.
UN (1997), Potential and policy implications of energy and material efficiency improvement, Worrell,
E., M.D. Levine, L. Price, N. Martin, R. van den Broek, K. Blok, United Nations (UN), New
York.
Vrieling, L. (1998), The pathfinder. Strategy paths and resources in their territory through time, PhD
thesis, Groningen, The Netherlands.
Walter, M., G. Stebner, J.M. Damasse, P. Tolve, Hohenbichler, G. (2000), EUROSTRIP. Initial
operating results of the plant at Krefeld Works, Germany, presented at: Conference on
Continuous Casting 2000, June 2000, Voest Alpine Industries, Linz, Austria, Opening
Session, paper no. 4.
WEC (1995), Efficient use of energy utilising high technology: An assessment of energy use in
industry and buildings, M.D. Levine, N. Martin, L. Price, E. Worrell (eds), World Energy
Council, London.
Wilkinson, A. (1998), personal communication, Avesta Sheffield, Sweden, December 8th 1998.
Williams, R. S. (1999), personal communication, Research and Technology Department, Armco Steel,
Middletown, US, March 3rd 1999.
Wolf, M.M. (1992), History of continuous casting, in: 75th Steelmaking Conference Proceedings,
Toronto, Canada, April 5-8, 1992, Iron and Steel Society (ISS) / AIME, pp. 83-137.

- 166 -