Content uploaded by Han Gao
Author content
All content in this area was uploaded by Han Gao on Dec 18, 2024
Content may be subject to copyright.
Methodology development for estimating the impact of restriction factors
to promote national steel recycling
Han Gao
a
, Junxi Liu
b
, Ichiro Daigo
a,b,c,*
a
Department of Advanced Interdisciplinary Studies, Graduate School of Engineering, The University of Tokyo 4-6-1 Komaba Meguro-ku, 153-8904, Tokyo, Japan
b
Research Center for Advanced Science and Technology, The University of Tokyo 4-6-1 Komaba Meguro-ku, 153-8904, Tokyo, Japan
c
UTokyo LCA Center for Future Strategy, The University of Tokyo 4-6-1 Komaba Meguro-ku, 153-8904, Tokyo, Japan
ARTICLE INFO
Keywords:
Sustainable production and consumption
Circular economy
Steel scrap
Recycling
Steel quality
ABSTRACT
To promote sustainable iron and steel production and consumption, maximizing scrap utilization is crucial.
However, the restrictions and their impacts on promoting steel recycling are unclear. This paper aims to develop
a methodology to identify main factors hindering steel recycling promotion in various countries. We developed a
methodology based on material ow analysis and multiregional input-output analysis to quantify the impacts of
restrictions from scrap unavailability and contamination, portfolio of steelmaking facilities, and demand for
high-quality steel on promoting steel recycling. Results revealed signicant variation in steel recycling perfor-
mance across 25 countries, with recycling input rates ranging from 8 % to 99 %, primarily restricted by the
demand for high-grade steel in 20 countries. Our ndings highlight the diverse challenges and opportunities for
promoting steel recycling across countries, through promoting scrap recovery, constructing new infrastructure
for Electric Arc Furnaces (EAFs), and upgrading the EAFs to produce at steel products exclusively with scrap.
1. Introduction
To limit human-induced global warming to a specic level, it is
necessary to limit cumulative carbon dioxide (CO
2
) emissions
(Intergovernmental Panel on Climate Change, 2023). This requires
accelerating efforts to achieve net zero CO
2
emissions and implementing
signicant reductions in other greenhouse gas emissions
(Intergovernmental Panel on Climate Change, 2023). The iron and steel
industry is responsible for approximately 7 % of global CO
2
emissions
from energy systems (International Energy Agency, 2023), and is
considered a hard-to-abate sector (Wang et al., 2021; Watari et al.,
2023). Under the pressure of reducing CO
2
emissions, this sector needs
to shift towards long-term low-carbon and sustainable production
practices. Steel recycling can help the iron and steel industry to reduce
CO
2
emissions due to a lower life-cycle emission intensity of steel pro-
duced from scrap (Wang et al., 2021). Some other technological ad-
vancements in the production of virgin iron, such as renewable
hydrogen to produce DRI (Boretti, 2023), also lead to a reduction of CO
2
emissions, but natural resource use is not reduced. Ferrous metal, the
most used metal and most recyclable metal with a long lifetime world-
wide demonstrates the importance of recycling (Charpentier Poncelet
et al., 2022). The recycling strategy emphasizes the repeated use of
materials to minimize the extraction of new natural resources on the
earth (Graedel et al., 2011a). The recycling input rate (RIR) is a critical
measure of metal recycling rates, indicating the fraction of secondary
metal in the total metal input for metal production (Graedel et al.,
2011a). Especially for iron and steel production, the RIR reects the
ratio of steel scrap to the total raw material input to the steelmaking
process.
A previous study (Pauliuk et al., 2013) highlighted that the global
supply of scrap is insufcient to meet the steel demand. The unavail-
ability of scrap forces steelmaking facilities to rely on virgin iron to
compensate for the shortfall, which directly hinders steelmaking facil-
ities from consuming more scrap. Furthermore, the mixing and accu-
mulation of impurities in scrap necessitate the use of virgin iron to dilute
the impurities contained in scrap. The impurity in scrap also prevents
steelmaking facilities from increasing scrap consumption (Daehn et al.,
2017; Nakamura et al., 2012). Consequently, either scrap unavailability
or contamination of scrap will increase virgin iron input for steelmaking.
Ultimately, scrap unavailability and contamination are regarded as one
pivotal restriction factor that directly hinders efforts to promote steel
recycling (Pauliuk et al., 2013).
Two prevalent steelmaking facilities are the basic oxygen furnace
(BOF) and electrical arc furnace (EAF), representing ore-based and
* Corresponding author.
E-mail address: daigo@material.t.u-tokyo.ac.jp (I. Daigo).
Contents lists available at ScienceDirect
Resources, Conservation & Recycling
journal homepage: www.sciencedirect.com/journal/resources-conservation-and-recycling
https://doi.org/10.1016/j.resconrec.2024.108052
Received 15 May 2024; Received in revised form 5 November 2024; Accepted 29 November 2024
Resources, Conservation & Recycling 215 (2025) 108052
Available online 5 December 2024
0921-3449/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (
http://creativecommons.org/licenses/by/4.0/ ).
scrap-based steelmaking, respectively (Oda et al., 2013; Pauliuk et al.,
2013). Technically, different types of steelmaking facilities have varying
maximum scrap input ratios (Xi et al., 2019). BOFs are constrained to a
maximum of 30 % scrap consumption (Xi et al., 2019), due to limitations
imposed by the heat balance of the Basic Oxygen Process model
(Madhavan et al., 2021) and contamination in the scrap (Larsson and
Dahl, 2003). In contrast, EAFs can utilize up to 100 % scrap, provided
the scrap is sufcient with an acceptable level of impurities (Xi et al.,
2019). The type and capacity of prevailing steelmaking facilities deter-
mine the potential scrap consumption. Therefore, a portfolio of steel-
making facilities, encompassing their types and capacities, becomes a
crucial determinant for promoting steel recycling.
The quality of steel is dened or classied differently depending on
the analytical perspectives in previous studies (Daigo et al., 2017;
Dworak et al., 2022; Nakamura et al., 2014; Pauliuk et al., 2017). The
quality of steel could be classied based on steelmaking furnace types,
and the end-use applications of the nished steel. For instance, the
cold-rolled plate, which is ore-based steel and used in automobiles, is
regarded as high-quality steel. In contrast, the concrete reinforcing bars,
commonly used in construction, are regarded as low-quality steel
(Nakamura et al., 2014; Pauliuk et al., 2017). Additionally, the quality of
steel could be classied based on the tolerable content of tramp elements
in nished steel. The nished steel which tolerates a lower maximum
content of tramp elements is grouped into a higher class of quality. The
high-quality steel includes typically most at products such as cold
rolled coils (Dworak and Fellner, 2021). Flat steel products are usually
produced in BOF from virgin iron, and rarely produced in EAF (Xylia
et al., 2018). Accordingly, the distinction between high-quality and
normal steel lies in their raw material inputs: high-quality steel can
consume a small portion of scrap, while normal steel can consume up to
100 % scrap. In this sense, high-quality steel is assumed to be at steel
products that can consume no more than 30 % scrap. Therefore, the
demand for high-quality steel indirectly promotes the consumption of
virgin iron, which in turn impedes the promotion of steel recycling.
The RIR varies across countries due to disparities in scrap unavail-
ability and contamination, a portfolio of steelmaking facilities, and de-
mand for high-quality steel. However, the specic impacts of these
restriction factors on changes in the recycling input rate of iron and steel
remain unclear. This knowledge gap impedes a comprehensive under-
standing of the potential promotion of steel recycling in diverse econ-
omies and globally. Therefore, this study aims to develop a methodology
to identify the restriction factors affecting the promotion of steel recy-
cling across different countries.
2. Materials and methods
2.1. Hypothetical scenarios
This study investigates the three factors that can restrict steel recy-
cling: scrap unavailability and contamination, the portfolio of steel-
making facilities, and the demand for high-quality steel, as shown in
Fig. 1. Scrap unavailability and contamination directly restrict recy-
cling, while the inuences of the other factors are indirect, often
obscured by issues related to scrap availability and quality. Moreover,
the impact of the demand for high-quality steel is further obscured by
the portfolio of steelmaking facilities. To assess the impact of each re-
striction on potential promotions of steel recycling, the steel recycling
promotion strategies under hypothetical scenarios are devised to
remove the restrictions without considering the feasibility of strategies
that can hypothetically remove restrictions. The strategies under hy-
pothetical scenarios are not designed to foresee future scenarios, but to
set ideal conditions to remove each restriction factor in order to quantify
the impact of the corresponding restriction factor. The individual impact
of each restriction factor can be quantied by the change of RIR between
before and after hypothetical removal.
The basic assumption is that the demand for nal products and the
design of nal products in three hypothetical scenarios remain the same
as in current conditions. Three steel recycling promotion strategies are
designed to hypothetically remove the corresponding restriction and
quantify the impact of the corresponding restriction: maximizing scrap
input to steelmaking facilities, the installation of new EAFs to maximize
the capacity of EAFs to produce all long steel products alongside
maximum scrap input, and upgrading EAFs to produce at steel prod-
ucts exclusively from scrap. The actual steel recycling performance is
represented by RIR
0
at the baseline scenario. As shown in Fig. 1, hy-
pothetical scenarios 1, 2, and 3 assume that no constraint on scrap
supply and quality, thus the scrap input to BOFs and EAFs in any port-
folio of steelmaking facilities can be allowed to increase in three hypo-
thetical scenarios. Hypothetical scenario 1 assumes that steelmaking
facilities utilize maximum scrap input without supply and quality con-
straints, the corresponding potential steel recycling performance is
represented as RIR
1
. Hypothetical scenario 2 assumes that all long steel
products can be produced by EAFs. By maximizing the capacity of EAFs
to produce all long steel products alongside maximizing scrap input, the
corresponding potential steel recycling performance is represented as
RIR
2
. Hypothetical scenario 3 assumes that all at steel products can be
produced by EAFs exclusively with scrap. In the repeated recycling of
Fig. 1. Schematic diagram of restriction factors impacting RIR promotion: Analyzing the effects of scrap unavailability and contamination, portfolio of steelmaking
facilities, and demand for high-quality steel.
H. Gao et al.
Resources, Conservation & Recycling 215 (2025) 108052
2
steel, there are some losses in steelmaking processes, which should be
fullled by the input of primary resources, if we assume to keep the
demand constant. In our hypotheses, for the sake of simplication, we
do not consider repeated recycling and assume that all steel products can
be produced only from scrap without any restrictions. The correspond-
ing potential steel recycling performance is represented as RIR
3
. The
differences between two adjacent scenarios indicated as ΔRIR, reect
the impact of corresponding factors. The factor with the greatest impact
on improving the RIR is identied as the dominant restriction factor.
The assumptions and parameters for calculating the potential RIR for
each scenario are detailed in Table 1. In hypothetical scenario 1, scrap is
sufciently provided with an acceptable level of impurities, and the
scrap input to BOFs and EAFs can reach their maximum capacities of 30
% and 100 % respectively (Xi et al., 2019). RIR
1
can be calculated by 30
% of the total material input to BOF and the total material input to EAF.
In hypothetical scenario 2, the long steel products produced by BOFs can
be substituted with an equivalent amount by EAFs, and the scrap for
steelmaking is sufciently provided with an acceptable level of impu-
rities. RIR
2
can be calculated under the maximum capacity of EAFs to
produce all long steel products alongside the maximum scrap input. In
hypothetical scenario 3, EAFs produce all steel using sufcient scrap
without any quality restriction. Consequently, the RIR
3
is 100 %.
2.2. The construction of multiregional material ow table
Understanding the material ows of iron and steel is essential for
estimating steel recycling performance. To evaluate the status of metal
recycling, the recycled content and end-of-life recycling rate (EoL-RR)
are regarded as key metrics (Reck and Graedel, 2012). EoL-RR has been
used to evaluate the material recoverability in Japan (Daigo et al.,
2015), EU-27 (Panasiuk, 2019), and globally (Charpentier Poncelet
et al., 2022). Recycled content describes the portion of scrap, which
shares the same calculation with RIR (Graedel et al., 2011a). The
country-level RIR was assumed to be a certain percentage of scrap in
total raw materials input (Pauliuk et al., 2013; Wieland et al., 2022). The
STAF model provided sufcient steel ow data which can be used to
calculate country-level RIR, but the datasets need to be updated (Wang
et al., 2007). Therefore, the construction of material ows for
country-level RIR calculation with updated data is necessary.
A valid RIR result could be calculated based on a clear image of
material transformation. The material transformations in MFA are usu-
ally represented by the inputs and outputs of each process. To illustrate
the transformations of iron and steel during production processes, ma-
terial ows of each form of iron and steel are constructed. Form(s) with
the same degree of fabrication are grouped into the same layer, as
demonstrated by the material transformations between layers 1 and 6 in
Fig. 2(a). The ows of raw materials, liquid steel, and crude steel are
critical for calculating RIR
0
, RIR
1
, and RIR
2
, respectively, as depicted by
the material transformations between layer 2 and layer 3, layer 3 to
layer 4, and layer 4 to layer 5, respectively. However, the conditions for
raw material inputs to steelmaking are complicated. Therefore, this
study explicitly shows the material ows of each type of steelmaking
process with state transition, as detailed with corresponding processes in
Fig. 2(b).
The forms of iron and steel and their geographic locations are
together used to dene the state of iron and steel. The transformation to
the next form of iron and steel and the change in geographic location are
regarded as a state transition. The material ows model illustrated in
Fig. 2(a) simulates the state transitions of iron and steel through pro-
duction processes, international trade, and the use phase. Forty-seven
states of iron and steel, organized into seven layers, in a single econ-
omy include one state in natural resource, three in raw materials, three
in liquid steel, four in crude steel, sixteen in nished steel, ten in nal
products, and ten in nal uses of nal products. Iron ore and scrap are
initial states, of which source(s) are regarded as the item(s) in boundary
input. During the production processes, the iron ores are converted into
pig iron, direct reduced iron (DRI), and iron loss in slag from the blast
furnace through the ironmaking process. The pig iron, DRI, and scrap
are melted and converted into liquid steel in steelmaking furnaces. BOF
liquid steel is processed into various types of crude steel, which are then
transformed into at and long steel products. On the other hand, EAF
liquid steel is processed into billets and blooms which are limited to be
transformed into long steel products. These nished steel products are
then manufactured into various nal products, such as automobiles and
machinery. Iron and steel in a certain form included in layers 1 to 5
transits into the next form with no change in geographic location. In
international trade, iron and steel in a certain form included in layers 2,
4, 5, and 6 derive from or transit to other countries without material
transition. In the use phase, each type of nal product transits into the
nal use without any further state transition. Ten states in the nal uses
of nal products are regarded as the states in layer 7. The states without
any outows, including nal uses, iron losses in slag, forming scrap
generation, and fabrication scrap generation, are regarded as states in
boundary output.
To identify the state transitions of iron and steel for multiple econ-
omies, a multiregional material ow table (MR-MF table) of iron and
steel with a physical unit (tonnes) is created. The MR-MF table com-
prises state transition matrices, boundary input matrices, and boundary
output matrices. The state transition matrices record the ows transiting
among 37 states of each economy as a 37 ×37 block matrix on the di-
agonal, and record the bilateral trade ows between states of different
economies as a 37 ×37 block matrix on the off-diagonal. The categories
of states in the state transition table and the items in boundary input and
output initially referred to the categories of process- and product-
specic supply and use tables and boundary input and output tables
from the physical supply-use table (PSUT) of iron and steel (Wieland
et al., 2022).
2.3. Data preparation for recycling input rate
1) Correspondence of steel scrap input in the state transition table
The RIR is used to describe the condition of scrap used in metal
production (Graedel et al., 2011a). Especially for steel production, the
RIR is calculated as the fraction of steel scrap in the total material input
to steelmaking, which includes pig iron, DRI, and scrap, as described in
Table 1
Assumptions and parameterization for hypothetical scenarios.
Scenarios Scrap unavailability
and contamination
Portfolio of
steelmaking
facilities
Demand for
high-quality steel
Baseline
scenario
Constant Constant Constant
Hypothetical
scenario 1
Sufcient scrap with
fewer enough
contamination 1.
Increase scrap input
to 30 % for BOF 2.
Increase scrap input
to 100 % for EAF
Constant Constant
Hypothetical
scenario 2
Sufcient scrap with
fewer enough
contamination 2.
Increase scrap input
to 30 % for BOF 3.
Increase scrap input
to 100 % for EAF
Sufcient EAF
steelmaking
facilities 1. The
amount of long
steel products
produced by BOFs
is substituted with
an equivalent
amount from EAFs
Constant
Hypothetical
scenario 3
Sufcient scrap with
fewer enough
contamination
Sufcient EAF
steelmaking
facilities
Upgraded EAF
steelmaking
facilities produce
at steel
products
H. Gao et al.
Resources, Conservation & Recycling 215 (2025) 108052
3
Eq. (1).
RIR =Steel scrap input
Pig iron input +DRI input +Steel scrap input
=
i=4,jϵ{5,6,7}
xij
iϵ{2,3,4},jϵ{5,6,7}
xij (1)
Where xij represents the amount of iron and steel in form i that
transits into form j in the state transition matrices, where i and j denote
the form of iron and steel as indicated in Fig. 2(a), i,jϵ{1,2,3⋯,37}. Pig
Fig. 2. Material ow model of iron and steel, shown by material transformations (a) and by process and state (b). In Fig. 2 (a), the grey area represents the cor-
responding states at each layer in the state transition table. The green area shows the items of boundary input. The blue area shows the items of boundary output.
H. Gao et al.
Resources, Conservation & Recycling 215 (2025) 108052
4
iron, DRI, and scrap are represented by states 2, 3, and 4, respectively.
BOF liquid steel, open-hearth furnace (OHF) liquid steel, and EAF liquid
steel are represented by states 5, 6, and 7, respectively. i=4,jϵ{5,6,7}xij
represents the total of steel scrap that transits into three types of liquid
steel. iϵ{2,3,4},jϵ{5,6,7}xij represents the total of raw materials that transits
into three types of liquid steel.
The material ows through the processes with the state transition is
described in the state transition table. Therefore, the material ows
transited to liquid steel in the state transition table do not align with raw
material inputs in RIR calculation that are the inputs to the steelmaking
process. For calculating the RIR with the numbers described in the state
transition table, the denitions of states related to the raw material in-
puts are elaborated with technical appropriateness. BOF, OHF, and EAF
liquid steel are dened as molten steel just by melting the inputs before
rening them in the corresponding furnace. Hence, any material ow
transited to each type of liquid steel can correspond directly with this
raw material input to the corresponding steelmaking furnace. Especially
for calculating pig iron input, iron losses during pretreatment of pig iron,
which occurs before pig iron is charged into the steelmaking furnaces,
should be considered, as demonstrated in Fig. 2(b). Technical parame-
ters are incorporated to calculate pig iron input to steelmaking by sub-
tracting iron losses during pig iron pretreatment from total pig iron
consumption. The iron losses during pig iron pretreatment are quanti-
ed using the average yield loss from the desulfurization (de[S]) process
(Schrama, 2021), the total iron contents in slag generated during the
dephosphorization (de[P]) process (Diao et al., 2012), and slag gener-
ation per tonne hot metal in the de[P] process (Matsubae-Yokoyama
et al., 2010). Consequently, the pretreated pig iron input to steelmaking
can be directly obtained from the ow of pig iron to liquid steel in the
state transition table.
The scrap input, liquid steel production, and portion of long steel
production are needed and important for RIR
0
, RIR
1
, and RIR
2
, respec-
tively. The trade ows are indispensable for the mass balance of the state
which involves the trade ows. The following Section (2) introduced the
calculation of steel scrap input and identication of liquid steel pro-
duction in statistical data, Section (3) introduced data preparation of
trade ows, and Section (4) introduced keeping the portion of long steel
production in mass balance.
2) Steel scrap input
Former global steel ows (Wieland et al., 2022) assumed consistent
scrap input ratios of BOFs and EAFs across all countries. Scrap input to
steelmaking is needed to calculate RIR
0
, corresponding to the ow of
scrap, in layer 2, to furnaces, in layer 3. This study considered the
technical information on steelmaking processes to estimate the scrap
input to steelmaking in each country.
Firstly, the total raw material input to steelmaking is estimated using
data for liquid steel production by BOFs, EAFs, and OHFs (World Steel
Association, 2018) and the yield ratio of the corresponding steelmaking
process. The RIR
1
is inuenced by the data for liquid steel production by
each type of furnace. Liquid steel and solid forms are not distinguished
in statistical representations (World Steel Association, 2018). We used
statistics on outputs from BOFs, EAFs, and OHFs (World Steel Associa-
tion, 2018) as data for liquid steel production in the corresponding
furnace, and claried the denitions of liquid steel and crude steel in this
study. This study follows Worldsteel’s denition of crude steel – steel in
its rst solid (or usable form) form, which includes ingots, billets, bloom,
slabs, and liquid steel for castings (World Steel Association, 2018).
Liquid steel is considered as the molten steel in steelmaking furnaces as
dened above, which is quantied as the sum of liquid steel production
and iron loss in slag during steelmaking. The yield ratios of BOFs and
EAFs are determined based on the representative parameters for total
iron (T-Fe) content in slag (Nippon Slag Association, 2024a) and slag
generation per tonne liquid steel production (World Steel Association,
2021). The available local parameters for several countries are assumed
to be the corresponding national average levels utilized for a more
precise estimation (Chen et al., 2023; Nippon Slag Association, 2024a,
2024b; Saikia and de Brito, 2009). More details can be found in section
S1 of supplementary information (SI).
Secondly, the total scrap input to steelmaking is determined through
a mass balance of the steelmaking process. The total raw material input
to steelmaking equals the amount of liquid steel in steelmaking furnaces.
The inputs of pig iron and DRI are calculated based on statistics of
production, imports, and exports (UN Comtrade, 2018; World Steel
Association, 2018). The total scrap input equals the difference between
the total raw material input to steelmaking and the total input of pre-
treated pig iron and DRI. To address instances when lacking DRI pro-
duction data might result in an unreasonable mass balance with a
negative number on DRI input to steelmaking, the DRI production is
assumed to be the number that yields DRI input to be zero.
The allocation of each type of raw material input to the BOFs and
EAFs, corresponding to ows between layer 2 and layer 3, follows three
procedures. First, the forming scrap is completely consumed by the
steelmaking furnace that generates it. Second, both fabrication scrap
and End-of-Life (EoL) scrap are consumed by BOFs and EAFs. The
representative RIR of BOFs is 12 %, as derived from the statistics-based
results of China (China Iron and Steel Association, 2018) and Japan (The
Japan Iron and Steel Federation, 2018). BOFs consume fabrication and
EoL scraps up to a 12 % RIR. EAFs consume fabrication and EoL scraps
up to their input capacity. If the RIR of EAFs is lower than that of BOFs,
then BOFs do not consume any fabrication and EoL scraps. If any
fabrication and EoL scraps remain after EAFs consumption, then BOFs
consume the leftovers unless the RIR of BOFs exceeds 30 %. Third,
pretreated pig iron is primarily consumed by BOFs. DRI and any
remaining pretreated pig iron are subsequently consumed by EAFs. The
estimated scrap input for each type of steelmaking furnace is shown in
Table S11 in the SI. Forming scrap and fabrication scrap generations are
assumed to be completely consumed by BOFs and EAFs. EoL scrap
consumption is regarded as the difference between the total scrap input
for steelmaking and the total amount of forming and fabrication scrap
generations.
Forming scrap generations in steelmaking and the continuous casting
(CC) processes are determined by mass balances of liquid steel at layer 3,
and crude steel at layer 4, respectively. More details are shown in section
S3 of the SI. Additionally, fabrication scrap generation is estimated
based on the consumption of each type of nished steel and the yield
ratio of each end-use sector. More details are shown in Table S9 of the SI.
The estimated forming and fabrication scrap generations are presented
in Tables S8 and S10, respectively.
3) Trade ows
Data for trade ows derived from or transited to each state in layers
2, 4, 5, and 6 are necessary for a mass balance of each state. UN Com-
trade statistics for imports of raw materials, crude steel, and nished
steel represent the trade ows derived from or transited to each state in
layers 2, 4, and 5, aligned with concordance between HS code and
Worldsteel classication (Wieland et al., 2022). Additionally, indirect
imports of steel are estimated using the net weight of imports of nal
products from UN Comtrade, along with steel coefcients (International
Iron and Steel Institute, 1996) following Worldsteel’s methodology
(World Steel Association, 2015). Each indirect import by source country
and type of nal product, corresponding to each trade inow of each
state in layer 6, is calculated based on the statistical total indirect import
of steel for each country (World Steel Association, 2018) and the cor-
responding proportions derived from our estimates. The total export is
determined by aggregating all other countries’ relevant imports (UN
Comtrade, 2018).
4) Proportions of long steel products
H. Gao et al.
Resources, Conservation & Recycling 215 (2025) 108052
5
Hypothetical scenario 2 assumes that all long steel products can be
produced by EAFs. RIR
2
is thus calculated by maximizing the capacity of
EAFs to produce all long steel products alongside maximizing scrap
input. The proportion of total long steel production in total nished steel
production can successively inuence the RIR
2
by inuencing the pro-
duction of billets and blooms. Long steel products are only derived from
billets and blooms. However, the production of billets and blooms is not
available in statistics. Consequently, the production of slabs and pro-
duction of billets and blooms can be estimated with statistical data for
at and long steel production (World Steel Association, 2018), respec-
tively. But the estimated production of slabs and the estimated pro-
duction of billets and blooms are not consistent with the production of
liquid steel, resulting in a mass imbalance at layer 4. To realize mass
balances at layer 4, the proportion of billets and blooms production and
Fig. 3. The steel ows of China (a) and India (b) in the year 2017. Presented in million tonnes.
H. Gao et al.
Resources, Conservation & Recycling 215 (2025) 108052
6
the proportion of total long steel production are kept. Ultimately,
keeping the proportion of total long steel production can result in a
reasonable RIR
2
. Further details can be found in section S3 in SI.
Once the proportion of total long steel production is kept, the RIR
2
is
unaffected by disaggregated production of each type of at and long
steel product, corresponding to ows between layers 4 and 5. The hy-
pothetical scenarios assume that the design of nal products and de-
mand for nal products remain. Hence, the RIR
2
is unaffected by the
design of nal products, represented by various types of nished steel
transformed into each nal product corresponding to ows between
layers 5 and 6. The RIR
2
is also unaffected by demand for nal products,
represented by nal products transited to nal uses corresponding to
ows between layers 6 and 7. Layer 3 is deemed as the reliable and
initial layer for keeping mass balance in this study. Therefore, the ma-
terial ows between layers 1 and 2 are estimated based on the mass
balance of layer 2 which is closer to layer 3. The material ows from
layer 4 to layer 7 are estimated based on the mass balance of the pre-
vious one between adjacent layers. In addition, the ows of nished
steel transformed into nal products are constructed using local statis-
tics of China and Japan, which can provide reliable data for researchers
studying nished steel consumption by end-use sector. These procedures
are outlined in detail in section S3 of the SI.
3. Results & discussions
3.1. Data quality assessment
To demonstrate the mass balance of steel ows and present the
available data in the MRMF table for RIR calculation, the consistent
material ows that achieved the mass balance of each state in the layers
between layers 1 and 7 are presented in Fig. 3. The innovation of the
MRMF of iron and steel compared with PSUT of iron and steel (Wieland
et al., 2022) lies in a mass-balanced material ows with considering the
country-specic raw material inputs to steelmaking. Compared with
datasets for environmentally-extended multi-regional input-output
(MRIO) analysis, such as EXIOBASE (Merciai and Schmidt, 2018) and
Global MRIO (Lenzen et al., 2021), the MRMF table of iron and steel
shows a higher resolution by distinguishing the types of crude steel and
nished steel.
The material ow model was designed with a layer-wise structure to
simplify the complexity of the relationship between various states. This
design enables each state to transit to a limited number of subsequent
states, which are organized within a clearly dened layered structure.
Consequently, the material transformations to the subsequent state can
be explicitly tracked to clarify the material transformation in the
sequential steel production process.
A rigorous approach is demonstrated to understanding the com-
plexities of steelmaking processes by considering the technical knowl-
edge. It ensures that the MFA of iron and steel is grounded in realistic
parameters and reects practical considerations, which enhances the
reliability of the ows of raw materials transformed into liquid steel as
prepared data for the calculation of RIRs.
The lack of data on raw material inputs complicates RIR
0
calcula-
tions. Employing MFA based on the mass-balance principle and sup-
ported by reliable statistics addresses this. Flows of raw materials
transformed into liquid steel are prepared for RIR calculations. Due to
the general unavailability or low quality of recycling statistics (Graedel
et al., 2011b), production, import, and export data of pig iron and DRI
are deemed more reliable. The mass balances of pig iron and DRI can
obtain reliable inputs of pig iron and DRI. Finally, the mass balance of
steelmaking can result in a reasonable scrap input to steelmaking.
To verify the feasibility of the methodology for estimating the scrap
input, the permissible range of scrap input in BOFs and EAFs can be
corroborated by technological knowledge. It is important to note that
the RIR results for BOFs and EAFs do not inuence their weighted
average, which is presented as RIR in the baseline scenario in Section
3.3. Due to the limited data availability of raw material inputs for BOFs
and EAFs, calculating their RIRs is challenging. To overcome data
scarcity, our general methodology provides reliable RIR estimates for
each furnace type, regardless of data availability. The RIRs for BOFs and
EAFs across 25 countries are depicted in Fig. S12 in the SI.
Given our assumption for constant nal demand and the same design
of nal products across hypothetical scenarios, the outows of nished
steel and inows and outows of the nal product remain unchanged
and do not impact the calculation of RIRs. The integration of local sta-
tistical data for nished steel consumption by the end-use sector into the
MRMF structure is demonstrated in China’s case, as shown in Fig. 3(a).
The estimated scrap input for steelmaking in China is 196.0 million
tonnes, which is higher than the reported 147.9 million tonnes. This
mass imbalance arises because different statistics are used to calculate
each input and output. This leads to considering the priority of statistical
data quality. Liquid steel production is considered the most reliable, and
the recycling statistic is regarded as less reliable (Graedel et al., 2011a).
Therefore, the estimated scrap input for steelmaking in China is adopted
despite the recycling statistic. Due to the availability of data on nished
steel consumption by end-use sectors in Japan and data on apparent
consumption of nished steel and market shares in China, the data for
various nished steel transformed into various nal products, corre-
sponding to ows between layer 5 and layer 6, are compiled using local
statistics for these two countries, respectively. These local statistics are
valuable for researchers studying nished steel consumption by end-use
sector in these countries.
For the other twenty-three countries, country-specic market shares
of nished steel and global average yield ratios for end-use sector
fabrication are derived from PSUT (Wieland et al., 2022). Fig. 3(b)
presented consistent steel ows using India as an example due to the
emerging market’s lack of comprehensive data. The data for nished
steel consumption by end-use sector for other countries can be compiled
with their local statistics to meet the requirements of future works.
Overall, the multiregional material ows achieved the mass balance
at each state. The newly compiled MRMF table specied the country-
specic raw material inputs of steelmaking. Moreover, our methodol-
ogy constructed a layer-wise material ow model to simplify the re-
lationships between states. The calculation of raw material inputs also
took a rigorous consideration with technical knowledge of the steel-
making process. The reliable raw material inputs can be ensured by
prioritizing the use of more reliable statistical data.
3.2. Country-specic main restriction factor on promoting steel recycling
The RIRs under current conditions and hypothetical scenarios across
25 countries are shown in Fig. 4. The estimated scrap input closely
matches the reported scrap use for steelmaking for six countries (US,
Japan, China, Canada, South Korea, and Russia)(Bureau of International
Recycling, 2022). More details are shown in Table S11 in the SI.
Table S13 in the SI provides the data for scrap and virgin iron con-
sumption in each economy under baseline and hypothetical scenarios.
The twenty-ve countries are categorized into three clusters based on
their primary restriction factor. For India, Mexico, Italy, and Greece, the
main restriction on promoting steel recycling is the shortage of scrap or
contamination issues. The estimated amounts for long steel production
by EAFs in each country determine the improvement from RIR
1
to RIR
2
.
The liquid steel demand for at steel production is estimated to be close
to BOF liquid steel production in India, Mexico, and Italy. The negligible
estimates of their long steel production from BOF result in no increment
from RIR
1
to RIR
2
, as indicated in Table S7 of the SI. In contrast, several
countries in Table S7 show that billets and blooms are produced via BOF.
Consequently, these countries have a big potential to replace the pro-
duction of billets and blooms by BOF with equivalent production by
EAF. In Greece, after ceasing cast iron production in 1981, the steel
industry completely transitioned to scrap smelting through open electric
arc furnaces (Zevgolis et al., 1999). Since then, the Greek steel industry
H. Gao et al.
Resources, Conservation & Recycling 215 (2025) 108052
7
has primarily produced long steel products, with large imports of at
steel products (Zevgolis et al., 1999), resulting in the highest RIR among
countries. These four countries could promote recycling by maximizing
the scrap input by furnace type, such as sourcing cleaner scrap (Daehn
et al., 2017) and increasing the production of recycled steel through
domestic recovery or scrap imports. In the Czech Republic, the pre-
dominant restriction is the portfolio of steelmaking facilities. BOF
steelmaking dominates 93 % of long steel production, as shown in
Table S7 of the SI. The Czech Republic could promote recycling by
maximizing the capacity of EAFs to produce all long steel products under
no supply and quality constraints of scrap, such as constructing new
EAFs.
Demand for high-quality steel emerges as the primary restriction to
promoting steel recycling in 20 countries, as shown by light blue bars in
Fig. 4. Analysis of the RIRs in hypothetical scenarios 1 and 2 shows that,
the global average RIR of 25 countries and the rest of the world could
increase from approximately 30 % to 64 % by shifting BOF-based long
steel products to EAF-based ones exclusively using scrap. This indicates
that the ongoing demand for high-quality steel makes it difcult to
replace over a third of the raw material input with scrap. While indi-
vidual countries could potentially reduce virgin iron consumption by
importing crude steel, nished steel, and nal products, the global
consumption of virgin iron for high-quality steel production remains
inevitable unless the steel industry undergoes technological advance-
ments that allow the exclusive production of high-quality steel only
using scrap, or unless there is a reduction in the demand for high-quality
steel in the design of nal products. Ultimately, addressing the challenge
posed by high-quality steel demand is essential for achieving signicant
promotions in global steel recycling. This will require a combination of
technological innovation and changes in the design and demand dy-
namics within the steel industry.
3.3. Limitations
A limitation of the RIR is its inability to accurately reect the sec-
ondary ratio within intermediate steel products, nal products, and their
nal uses. Given the global trade in intermediate and nal steel prod-
ucts, a high RIR may not necessarily represent a comprehensive measure
of steel recycling performance. To assess steel recycling performance
across various stages, it is crucial to analyze the secondary material
ratios in intermediate steel products, nal products, and nal uses.
Evaluating how scrap ratios within steel ows are affected by the in-
ternational trade of intermediate and nal steel products, and deter-
mining the extent of scrap consumption in the nal uses of each
economy, requires a detailed assessment of the secondary ratios at these
stages. This approach provides a consumption-based evaluation of steel
recycling that accounts for the actual use of recycled material. Conse-
quently, future research is expected to focus on estimating consumption-
based steel recycling performance at the level of individual economy.
This will address the current limitations and contribute to a more
comprehensive understanding of the dynamics of steel recycling,
particularly how it is inuenced by international trade and the con-
sumption of steel products.
This study has several limitations. First, it operates under the
assumption of xed demand at the economy level and does not consider
changes in the design of the nal product structure. Consequently, the
study cannot provide insights into potential promotions in steel recy-
cling that could be achieved through upgrades in the design of nal
products. Second, EAFs are assumed to only produce long steel products
in the material ow model of iron and steel, which may differ from the
fact in North America as reported in previous studies (Cullen et al.,
2012; Daehn et al., 2017; Xylia et al., 2018) and in India as reported by
the local statistical information (Ministry of Steel, 2017). Third,
high-quality steel is synonymous with BOF-based at steel products in
this study, since at steel products typically demand higher inputs of
Fig. 4. The main restriction for steel recycling in domestic production systems for 25 countries grouped into three clusters based on their common main restriction
factor. The countries in each group are ordered by the RIR of the baseline scenario. The black points represent the RIR of steelmaking, indicating the weighted
average of RIRs of BOFs and EAFs. The blue and grey bars display the potential promotion in steel recycling by maximizing scrap input by the furnace type (ΔRIR
1
)
and maximizing the capacity of EAFs alongside maximizing scrap input (ΔRIR
2
), respectively. The gap between the red point and one hundred percent illustrates the
potential promotion by upgrading EAFs to produce at steel products exclusively with scrap (ΔRIR
3
). The primary restriction factor for each country is identied by
determining the largest value among ΔRIR
1
, ΔRIR
2
, and ΔRIR
3
.
H. Gao et al.
Resources, Conservation & Recycling 215 (2025) 108052
8
primary resources. But high-quality steel could incorporate more types
of steel if considering the properties of steel. Fourth, the specic coun-
termeasures to increase scrap supply about where to get scrap (domestic
scrap recovery or imports of scrap) cannot be provided by this study.
Fifth, the feasibilities of each steel recycling promotion strategy are not
considered, for instance, practical implementation costs and energy
demand of replacing BOFs with EAFs are not considered, and the tech-
nology for upgrading EAFs to produce at steel products with more
scrap is explored in the laboratory scale experiment (Huellen et al.,
2006), and may continue developing in mini-mill steelworks (Umezawa
et al., 2019), and company’s research and development (Nucor Corpo-
ration, 2020). The consequence of each steel recycling promotion
strategy is not considered, such as the change of slag supply and demand
or the life-cycle environmental impact from the additional and saving
demand under the new portfolio of steelmaking facilities. Sixth, the
scrap input to blast furnaces for ironmaking is not considered in this
study, as the RIR indicates the scrap input ratio for steelmaking.
4. Conclusion
This study identied restriction factors on promoting steel recycling,
including scrap unavailability and contamination, the portfolio of
steelmaking facilities, and the demand for high-quality steel. We
developed steel ow models and designed hypothetical scenarios to
quantify the impact of various restriction factors on promoting national
steel recycling. It should be noted that the hypothetical scenarios are
designed to hypothetically remove the impact of the corresponding re-
striction factor that hinders more scrap consumption, which are not
estimations for the future and do not consider the feasibility and
consequence of strategies. A mass-balanced multiregional material ow
table for iron and steel for the year 2017 was meticulously constructed.
Our ndings revealed considerable variations in the recycling input
rates of steelmaking across 25 countries, ranging from 8 % to 99 %. The
demand for high-quality steel is the main restriction factor on promoting
steel recycling across 20 countries. The signicant disparities in poten-
tial promotions highlight the diverse opportunities for promoting steel
recycling through promoting scrap recovery and constructing new EAFs
across these countries. Upgrading the EAFs to produce at steel products
with more scrap consumption may be an urgent and upcoming challenge
for countries that may have sufcient scrap availability in the future.
CRediT authorship contribution statement
Han Gao: Writing – original draft, Visualization, Validation, Meth-
odology, Funding acquisition, Formal analysis, Data curation, Concep-
tualization. Junxi Liu: Writing – review & editing, Validation,
Methodology, Formal analysis. Ichiro Daigo: Writing – review & edit-
ing, Validation, Supervision, Methodology, Funding acquisition, Formal
analysis, Data curation, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Acknowledgment
This work was supported by JST SPRING (Grant Number
JPMJSP2108), the Environment Research and Technology Development
Fund (JPMEERF20S11816) of the Environmental Restoration and Con-
servation Agency of Japan, and the New Energy and Industrial Tech-
nology Development Organization (NEDO) (JPNP21019).
Supplementary materials
Supplementary material associated with this article can be found, in
the online version, at doi:10.1016/j.resconrec.2024.108052.
Data availability
Data will be made available on request.
References
Boretti, A., 2023. The perspective of hydrogen direct reduction of iron. J. Clean. Prod.
429, 139585. https://doi.org/10.1016/j.jclepro.2023.139585.
Bureau of International Recycling, 2022. World Steel Recycling in Figures 2017-2021.
Steel scrap- a Raw Material for Green Steelmaking. Bureau of International
Recycling.
Charpentier Poncelet, A., Helbig, C., Loubet, P., Beylot, A., Muller, S., Villeneuve, J.,
Laratte, B., Thorenz, A., Tuma, A., Sonnemann, G., 2022. Losses and lifetimes of
metals in the economy. Nat. Sustain. 5, 717–726. https://doi.org/10.1038/s41893-
022-00895-8.
Chen, H.-J., Lin, Y.-R., Tang, C.-W., Hung, Y.-C., 2023. Application of biomineralization
technology in the stabilization of electric arc furnace reducing slag. Appl. Sci. 13,
10435. https://doi.org/10.3390/app131810435.
China Iron and Steel Association, 2018. China Steel Yearbook 2018. The editorial board
of China Steel Yearbook, Beijing.
Cullen, J.M., Allwood, J.M., Bambach, M.D., 2012. Mapping the global ow of steel:
from steelmaking to end-use goods. Environ. Sci. Technol. 46, 13048–13055.
https://doi.org/10.1021/es302433p.
Daehn, K.E., Cabrera Serrenho, A., Allwood, J.M., 2017. How will copper contamination
constrain future global steel recycling? Environ. Sci. Technol. 51, 6599–6606.
https://doi.org/10.1021/acs.est.7b00997.
Daigo, I., Fujimura, L., Hayashi, H., Yamasue, E., Ohta, S., Huy, T.D., Goto, Y., 2017.
Quantifying the total amounts of tramp elements associated with carbon steel
production in Japan. ISIJ Int. 57, 388–393. https://doi.org/10.2355/
isijinternational.ISIJINT-2016-500.
Daigo, I., Iwata, K., Ohkata, I., Goto, Y., 2015. Macroscopic evidence for the hibernating
behavior of materials stock. Environ. Sci. Technol. 49, 8691–8696. https://doi.org/
10.1021/acs.est.5b01164.
Diao, J., Xie, B., Wang, Y., Guo, X., 2012. Recovery of phosphorus from
dephosphorization slag produced by duplex high phosphorus hot metal rening. ISIJ
Int 52, 955–959. https://doi.org/10.2355/isijinternational.52.955.
Dworak, S., Fellner, J., 2021. Steel scrap generation in the EU-28 since 1946 – sources
and composition. Resour. Conserv. Recycl. 173, 105692. https://doi.org/10.1016/j.
resconrec.2021.105692.
Dworak, S., Rechberger, H., Fellner, J., 2022. How will tramp elements affect future steel
recycling in Europe? – A dynamic material ow model for steel in the EU-28 for the
period 1910 to 2050. Resour. Conserv. Recycl. 179. https://doi.org/10.1016/j.
resconrec.2021.106072.
Graedel, T.E., Allwood, J., Birat, J.-P., Buchert, M., Hagelüken, C., Reck, B.K., Sibley, S.
F., Sonnemann, G., 2011a. Recycling Rates of Metals: a Status Report. United
Nations Environment Programme, Nairobi, Kenya.
Graedel, T.E., Allwood, J., Birat, J.-P., Buchert, M., Hagelüken, C., Reck, B.K., Sibley, S.
F., Sonnemann, G., 2011b. What do we know about metal recycling rates? J. Ind.
Ecol. 15, 355–366. https://doi.org/10.1111/j.1530-9290.2011.00342.x.
Huellen, M., Schrade, C., Wilhelm, U., Zulhan, Z., Fuchs, S.V., 2006. EAF-based at-steel
production applying secondary metallurgical processes.
Intergovernmental Panel on Climate Change, 2023. Climate Change 2021 – The Physical
Science Basis: Working Group I Contribution to the Sixth Assessment Report of the
Intergovernmental Panel On Climate Change. Cambridge University Press. https://
doi.org/10.1017/9781009157896, 1st ed.
International Energy Agency, 2023. CO2 emissions of Industry [WWW Document].
IEA50. URL https://www.iea.org/energy-system/industry#tracking.
International Iron and Steel Institute, 1996. Indirect Trade in Steel 1989-1993 (IISI).
International Iron and Steel Institute (IISI), Brussels.
Larsson, M., Dahl, J., 2003. Reduction of the specic energy use in an integrated steel
plant-the effect of an optimisation model. ISIJ Int. 43, 1664–1673. https://doi.org/
10.2355/isijinternational.43.1664.
Lenzen, M., Geschke, A., West, J., Fry, J., Malik, A., Giljum, S., Mil`
a I Canals, L.,
Pi˜
nero, P., Lutter, S., Wiedmann, T., Li, M., Sevenster, M., Potoˇ
cnik, J., Teixeira, I.,
Van Voore, M., Nansai, K., Schandl, H., 2021. Implementing the material footprint to
measure progress towards sustainable development goals 8 and 12. Nat. Sustain. 5,
157–166. https://doi.org/10.1038/s41893-021-00811-6.
Madhavan, N., Brooks, G.A., Rhamdhani, M.A., Rout, B.K., Overbosch, A., 2021.
Application of mass and energy balance in oxygen steelmaking. Ironmak. Steelmak.
48, 995–1000. https://doi.org/10.1080/03019233.2020.1850170.
Matsubae-Yokoyama, K., Kubo, H., Nagasaka, T., 2010. Recycling effects of residual slag
after magnetic separation for phosphorus recovery from hot metal
dephosphorization slag. ISIJ Int. 50, 65–70. https://doi.org/10.2355/
isijinternational.50.65.
Merciai, S., Schmidt, J., 2018. Methodology for the construction of global multi-regional
hybrid supply and use tables for the EXIOBASE v3 database. J. Ind. Ecol. 22,
516–531. https://doi.org/10.1111/jiec.12713.
H. Gao et al.
Resources, Conservation & Recycling 215 (2025) 108052
9
Ministry of Steel, 2017. National conference of secondary steel producers on the theme
‘Make in Steel for Make in India’ by ministry of steel in New Delhi on 5 April 2017
[WWW Document]. Press Inf. Bur. Gov. India. URL https://pib.gov.in/PressNot
eDetails.aspx?NoteId=149963®=3&lang=1(accessed 10.15.24).
Nakamura, S., Kondo, Y., Kagawa, S., Matsubae, K., Nakajima, K., Nagasaka, T., 2014.
MaTrace: tracing the fate of materials over time and across products in open-loop
recycling. Environ. Sci. Technol. 48, 7207–7214. https://doi.org/10.1021/
es500820h.
Nakamura, S., Kondo, Y., Matsubae, K., Nakajima, K., Tasaki, T., Nagasaka, T., 2012.
Quality- and dilution losses in the recycling of ferrous materials from end-of-life
passenger cars: input-output analysis under explicit consideration of scrap quality.
Environ. Sci. Technol. 46, 9266–9273. https://doi.org/10.1021/es3013529.
Nippon Slag Association, 2024a. Chemical characteristics of iron and steel slag [WWW
Document]. Nippon Slag Assoc. URL https://www.slg.jp/e/slag/character.html.
Nippon Slag Association, 2024b. Types of iron and steel slag [WWW Document]. Nippon
Slag Assoc. URL https://www.slg.jp/e/slag/kind.html.
Nucor Corporation, 2020. Nucor expands steelmaking capacity with new EAF. Met. Cent.
News.
Oda, J., Akimoto, K., Tomoda, T., 2013. Long-term global availability of steel scrap.
Resour. Conserv. Recycl. 81, 81–91. https://doi.org/10.1016/j.
resconrec.2013.10.002.
Panasiuk, D., 2019. Dynamic material ow analysis for estimation of iron ows, stocks
and recycling indicators in EU-27.
Pauliuk, S., Kondo, Y., Nakamura, S., Nakajima, K., 2017. Regional distribution and
losses of end-of-life steel throughout multiple product life cycles—insights from the
global multiregional MaTrace model. Resour. Conserv. Recycl. 116, 84–93. https://
doi.org/10.1016/j.resconrec.2016.09.029.
Pauliuk, S., Milford, R.L., Müller, D.B., Allwood, J.M., 2013. The steel scrap age. Environ.
Sci. Technol. 47, 3448–3454. https://doi.org/10.1021/es303149z.
Reck, B.K., Graedel, T.E., 2012. Challenges in metal recycling. Science (1979) 337,
690–695. https://doi.org/10.1126/science.1217501.
Saikia, N., de Brito, J., 2009. Use of industrial waste and municipality solid waste as
aggregate, ller or ber in cement mortar and concrete. Advances in Materials
Science Research. Nova Science Publishers, Inc., pp. 27–30
Schrama, F.N.H., 2021. Desulphurisation in 21st Century Iron- and Steelmaking. Delft
University of Technology. https://doi.org/10.4233/UUID:4F1E8FF6-E910-4048-
8A17-2A959C74F508.
The Japan Iron and Steel Federation, 2018. JFE Techno-Research Corporation, National
Steel Sales Federation, Zenkoku Atsuita Shearing Kohgyokumiai, National Coil
Center Industrial Union, Special Steel Association of Japan, Steel Castings and
Forgings Association of Japan, Japan Ferroalloy Association, Japan Metal Daily,
2018. Steel Yearbook. Japan Metal Daily.
Umezawa, O., Shinohara, Y., Halada, K., 2019. Growth of Ecomaterials and Eco-
efciency in Major Metallic Structural Materials, Handbook of Ecomaterials.
Springer Nature.
UN Comtrade, 2018. United nations commodity trade statistics database.
Wang, P., Ryberg, M., Yang, Y., Feng, K., Kara, S., Hauschild, M., Chen, W.Q., 2021.
Efciency stagnation in global steel production urges joint supply- and demand-side
mitigation efforts. Nat. Commun. 12, 1–11. https://doi.org/10.1038/s41467-021-
22245-6.
Wang, T., Müller, D.B., Graedel, T.E., 2007. Forging the anthropogenic iron cycle.
Environ. Sci. Technol. 41, 5120–5129. https://doi.org/10.1021/es062761t.
Watari, T., Hata, S., Nakajima, K., Nansai, K., 2023. Limited quantity and quality of steel
supply in a zero-emission future. Nat. Sustain. 6, 336–343. https://doi.org/10.1038/
s41893-022-01025-0.
Wieland, H., Lenzen, M., Geschke, A., Fry, J., Wiedenhofer, D., Eisenmenger, N.,
Schenk, J., Giljum, S., 2022. The PIOLab: building global physical input–output
tables in a virtual laboratory. J. Ind. Ecol. 26, 683–703. https://doi.org/10.1111/
jiec.13215.
World Steel Association, 2015. Indirect Trade in Steel. World Steel Association.
World Steel Association, 2018. Steel statistical yearbook 2018.
World Steel Association, 2021. Steel industry co-products (Fact sheet).
Xi, L., Qianguo, L., Hasan, M., Ming, L., Qiang, L., Jia, L., Alisa, W., Muxin, L.,
Francisco, A., 2019. Assessing the economics of CO2 capture in China’s iron/steel
sector: a case study. Energy Procedia 158, 3715–3722. https://doi.org/10.1016/j.
egypro.2019.01.886.
Xylia, M., Silveira, S., Duerinck, J., Meinke-Hubeny, F., 2018. Weighing regional scrap
availability in global pathways for steel production processes. Energy Efc. 11,
1135–1159. https://doi.org/10.1007/s12053-017-9583-7.
Zevgolis, M.N., Georgakellos, N., Spyropoulos, P., 1999. Ferrous metals and ferroalloys in
Greece (Present state and perspective).
H. Gao et al.
Resources, Conservation & Recycling 215 (2025) 108052
10
