-
ifeu - Institut für Energie- und Umweltforschung Heidelberg
GmbH
CONVERSION OF EUROPEAN PRODUCT FLOWS INTO RAW MATERIAL
EQUIVALENTS
Final report of the project: Assistance in the development and
maintenance of Raw Material Equivalents conversion factors and
calculation of RMC time se-ries
Karl Schoer, Jürgen Giegrich, Jan Kovanda, Christoph
Lauwigi,
Axel Liebich, Sarka Buyny, Josefine Matthias
commissioned by Statistical Office of the European Communities –
Eurostat; Directorate E – Agriculture and Environmental
Statis-tics; Statistical Cooperation Unit E3: Environment
statistics Contract no. 50902.2010.001.-2010.612
ifeu - Institut für Energie- und Umweltforschung,
Heidelberg
in cooperation with:
Sustainable Solutions Germany – Consultants,
Wiesbaden (SSG)
Charles University in Prague, Environment Centre
CUEC
Heidelberg, May 2012
-
Assistance in the development and maintenance of Raw Material
Equivalents
conversion factors and calculation of RMC time series
Content
Summary
_______________________________________________________________
1
1. Introduction
_________________________________________________________ 4
1.1. Background
___________________________________________________________ 4
1.2. The concept of raw material equivalents
____________________________________ 6
2. The RME calculation model
____________________________________________ 10
2.1. General considerations
_________________________________________________ 10
2.2. Overview calculation model
_____________________________________________ 12
2.3. Disaggregated monetary IOT
____________________________________________ 15 2.3.1. The concept
and rationale for disaggregating the IOT
_____________________________ 15 2.3.2. Calculation method and data
sources __________________________________________ 17
2.3.2.1. Structural information based on German expanded IOT
_________________________ 18 2.3.2.2. EU level balance column and
line total _______________________________________ 20 2.3.2.3.
Iterative adjustment approach
_____________________________________________ 21 2.3.2.4. Data
issues
_____________________________________________________________
25
2.4. Disaggregated hybrid IOT
_______________________________________________ 27 2.4.1. The
concept and rationale for a hybrid IOT
______________________________________ 27 2.4.2. Calculation method
and data sources __________________________________________ 30
2.4.2.1. Price adjustment of monetary use
structure___________________________________ 30 2.4.2.2. Physical
use structure for biomass and other minerals
__________________________ 31 2.4.2.3. Physical use structure for
metals ____________________________________________ 32 2.4.2.4.
Physical use structures for energy carriers
____________________________________ 34
2.5. Annual environmental extensions
________________________________________ 35 2.5.1. Domestic
extraction used
____________________________________________________ 35 2.5.2.
External approach for estimating RME of imported LCA products
____________________ 35
2.6. Metal model for estimating RME coefficients
_______________________________ 38 2.6.1. Introduction to the metal
model ______________________________________________ 38
2.6.1.1. Representative RME coefficients
____________________________________________ 39 2.6.1.2.
Methodology for polymetallic ores
__________________________________________ 43 2.6.1.3. Auxiliary
materials _______________________________________________________
47 2.6.1.4. Metal content of alloys
___________________________________________________ 48
2.6.2. Metal content of imported metal concentrates
__________________________________ 49 2.6.3. Treatment of secondary
metals _______________________________________________ 50 2.6.4.
RME coefficients for imported metal products
___________________________________ 53
2.7. A mixed Leontief / LCA approach
_________________________________________ 54
2.8. Data quality considerations
_____________________________________________ 56 2.8.1. Comparison
to standard approach _____________________________________________
56 2.8.2. Comparison to other calculation models
________________________________________ 58
2.8.2.1. Czech approach
_________________________________________________________ 59
2.8.2.2. German approach
________________________________________________________ 60
2.8.2.3. EXIOPOL approach
_______________________________________________________ 61
3. Analytical properties of RME accounting and selected
results_________________ 66
-
Assistance in the development and maintenance of Raw Material
Equivalents
conversion factors and calculation of RMC time series
3.1. Economy wide indicators
_______________________________________________ 67
3.2. Ecological link
________________________________________________________ 68 3.2.1.
Disaggregation by raw material categories
______________________________________ 68 3.2.2. Pressure profiles
of raw materials _____________________________________________
71
3.3. Economic link
_________________________________________________________ 75 3.3.1.
RME by products and categories of final uses
____________________________________ 75 3.3.2. Structural
decomposition of RME _____________________________________________
78
4. Outlook RME model
__________________________________________________ 81
4.1. Migration from CPA 2002 to CPA 2008
_____________________________________ 81
4.2. Options for regionalization of RME calculation
______________________________ 82
4.3. Internalized capital formation
___________________________________________ 89 4.3.1. Concept
__________________________________________________________________
89 4.3.2. Calculation approach
_______________________________________________________ 91 4.3.3.
Results
___________________________________________________________________
92
5. Acknowledgements
__________________________________________________ 95
List of Abbreviations
_____________________________________________________ 96
Bibliography
____________________________________________________________ 97
ANNEX 1 Product classification for expHIOT
_________________________________ 101
ANNEX 2 Data sources
___________________________________________________ 106
ANNEX 3 Conversion factors imported metal to RME
__________________________ 107
ANNEX 4 Data documentation metal model
_________________________________ 108
-
Assistance in the development and maintenance of Raw Material
Equivalents page 1
conversion factors and calculation of RMC time series
Summary
English:
This report presents an approach for converting product flows
into raw material equiva-
lents (RME). The RME concept takes the perspective of raw
materials embodied in prod-
ucts. The RME of a product indicates how much extraction of
material was necessary over
the whole production chain for manufacturing a specific product,
irrespective whether
those raw materials where extracted from the domestic or the
rest of the world environ-
ment. The weight of the consumed raw materials is measured at
the point of extraction
from the environment.
It was the objective of the project to improve the current
European material flow indicator
“Domestic Material Consumption” (DMC) by overcoming the
asymmetry of measuring at
the one hand the mass of domestic extraction of raw materials
from the environment at
the point of extraction, but on the other to measure the weight
of material which is import-
ed or exported at the time of crossing the border. By expressing
all flows in RME it be-
comes possible to make up a mass balance of domestic uses,
imports and exports by a
coherent unit of measurement. It is intended to use the
indicator “Domestic Raw Material
Consumption” (RMC), which is derived from the RME calculation
system, as a central in-
dicator of the European Resource Strategy.
The estimation of RME is based on the Leontief approach, which
is a well-established
method for environmental economic analysis. That approach
applies Input-Output analy-
sis for assigning direct environmental pressures – measured in
physical units – by the
individual production activities to the products of final use
and of imports.
As far as the application of the “standard version” of the
Leontief approach – which is
based on the standard European monetary IOT of the size 60x60 -
for the estimating RME
is concerned, the following major deficits where identified:
a) The resolution of the European monetary standard Input-Output
table (IOT) with
the format 60x60 is not sufficient for depicting the flow of raw
materials through the
economy with an acceptable degree of accuracy. Therefore an
approach was de-
veloped for estimating an expanded IOT of the format 166x166 at
an annual basis.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 2
conversion factors and calculation of RMC time series
b) IOT use structures in physical units are able to provide much
more accurate re-
sults for calculation of RME than monetary use structures, as
far as raw products
and some primary processed raw products are concerned. Therefore
a hybrid IOT
was established, i.e. monetary structures of the monetary IOT
where replaced by
use structures in physical units for selected product groups
c) The assumption that imported products are produced with
domestic production re-
lationships, which is inherent to the standard Leontief
approach, is especially not
maintainable in case of calculation of RME for metals. Therefore
the RME of se-
lected imported products where estimated by an external
approach. That external
approach is based on annual preparation of business reports of
mining companies
for 160 mines worldwide and of results from Life Cycle
Assessment.
A comparison of the outcome of the project approach and of the
“standard approach”
demonstrates that the new method yields calculation results
which are significantly more
accurate.
An automated calculation procedure furnishes annual results for
Raw Material Equivalents
(RME) in a detailed disaggregation by categories of final use
and imports, product groups
and type of raw material. From that data the central economy
wide material flow indicator
of the European Sustainable Development Strategy “Domestic Raw
material Consump-
tion” (RMC) is derived by aggregation. At the same time that
data set provides the basis
for analyzing the RMC in an environmental economic context by
linking it to the underlying
economic driving forces (economic link) and to the environmental
impacts (ecological link)
which are intended to be addressed by that general
indicator.
Deutsch:
Dieser Bericht stellt ein Verfahren zur Konvertierung von
Güterströmen in Rohstoffäquiva-
lente vor (Raw Material Equivalents, RME). RME beschreiben die
Menge an Rohstoffen in
Gewichtseinheiten, die über die gesamte Produktionskette hinweg
für die Herstellung ei-
nes Gutes im In- oder Ausland aufgewendet wurden (embodied raw
materials). Das Ge-
wicht der eingesetzten Rohstoffe wird einheitlich zum Zeitpunkt
der Entnahme aus der
Natur gemessen.
Zielsetzung des Projektes war es, den bisherigen Europäischen
Rohstoffindikator
"Domestic Material Consumption" (DMC) zu verbessern durch
Überwindung der Asym-
-
Assistance in the development and maintenance of Raw Material
Equivalents page 3
conversion factors and calculation of RMC time series
metrie von Messung der inländischen Rohstoffentnahme als
Rohstoffgewicht bei der Ent-
nahme aus der Natur und der Import- und Exportströme mit dem
Gewicht beim Grenz-
übergang. Durch die Darstellung der entsprechenden Ströme in RME
wird es möglich,
inländische Verwendung sowie die Importe und Exporte von Gütern
in Gewichtseinheiten
nach einem einheitlichen Maßstab zu bilanzieren. Es ist
vorgesehen, den aus dem Be-
rechnungssystem abgeleiteten Indikator „Domestic Raw Material
Consumption“ (RMC) als
zentralen Rohstoffindikator für die europäische
Ressourcenstrategie zu verwenden.
Die Berechnung von RME beruht auf dem in der umweltökonomischen
Analyse etablier-
ten Verfahren des Leontief Ansatzes. Dabei handelt es um ein
Verfahren der Input-Output
Analyse, das direkte Umweltbelastungen - gemessen in physischen
Einheiten - durch die
einzelnen Produktionsaktivitäten den Gütern der letzten
Verwendung und dem Import
zuordnet.
Bezüglich der Verwendung dieses Verfahrens in der
"Standardvariante" für die Schätzung
von RME wurden allerdings folgende grundlegende Defizite
identifiziert:
a) Der Detaillierungsgrad der Europäischen monetären Standard
Input-Output Tabel-
le (IOT) mit dem Format 60x60 reicht nicht aus, um den Fluss von
Rohstoffen
durch die Wirtschaft mit hinreichender Genauigkeit abzubilden.
Deshalb wurde ein
Verfahren zur jährlichen Schätzung einer erweiterten IOT im
Format 166x166 ent-
wickelt.
b) Für Rohstoffgütergruppen und einige rohstoffnahe Gütergruppen
führen Verwen-
dungsstrukturen in physischen Einheiten zu deutlich genaueren
Ergebnissen bei
der RME Berechnung, als monetäre Strukturen. Deshalb wurden
Verfahrenswei-
sen zur Schätzung einer hybriden IOT entwickelt, bei der
monetäre Verwendungs-
strukturen selektiv durch physische ersetzt wurden.
c) Die dem Leontief Verfahren inhärente Annahme, dass
importierte Güter mit inlän-
discher Produktionstechnologie produziert wurden, ist
insbesondere mit Hinblick
auf die Berechnungen für Metalle nicht haltbar. Aus diesem Grund
wurden die
RME ausgewählter Importgüter mit Hilfe eines externen Ansatzes
geschätzt, der
sich vor allem auf die Auswertung der Geschäftsberichte für
weltweit etwa 160 Mi-
nen und auf Ergebnisse von sogenannten Lebenszyklusanalysen
(Life Cycle
Assessment) stützt.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 4
conversion factors and calculation of RMC time series
Ein Vergleich der Projektergebnisse mit den Ergebnissen nach dem
Standardverfah-
ren verdeutlicht, dass das im Projekt entwickelte Verfahren zu
wesentlich genaueren
Ergebnissen führt.
In einem automatisierten Berechnungsverfahren werden jährliche
Ergebnisse in Roh-
stoffäquivalenten (RME) in tiefer Untergliederung nach
Kategorien der letzten Ver-
wendung und Importen, Gütergruppen und Rohstoffarten
bereitgestellt. Aus diesen
Daten wird der zentrale gesamtwirtschaftliche Rohstoffindikator
der Europäischen
Ressourcenstrategie "Domestic Raw Material Consumption" (RMC)
durch Aggregation
abgeleitet. Der Datensatz bildet aber zugleich die Grundlage, um
diesen Indikator in
einem umweltökonomischen Kontext zu analysieren und dabei den
Indikator sowohl
mit den dahinter stehenden ökonomischen Triebkräften (economic
link) als auch mit
den durch den Indikator indizierten Umweltauswirkungen
(ecological link) zu verknüp-
fen.
1. Introduction
1.1. Background1
Based on the Thematic Strategy on the Sustainable Use of Natural
Resources [EC 2005]
and the 6th Environmental Action Programme [EC 2002] the EU
Commission decided to
provide policy makers and other stakeholders with a framework of
information about the
use of resources and products. Therefore, Environmental Data
Centres had been estab-
lished to fulfil this purpose. Within an agreement between other
Directorates of the EU the
responsibility for the Environmental Data Centres for natural
resources, products, and
waste had been assigned to Eurostat.
The main task of the Environmental Data Centres on resources and
products is to im-
prove knowledge about the relationship between resource use,
economic growth and en-
vironmental impacts. The combination of economic information and
environmental input
and output information – both handled at Eurostat – is an
important basis for fulfilling the
needs of such an Environmental Data Centre. Resource
productivity is considered to be
an important indicator of Eurostat’s resource strategy.
1 First results of this project are reported in the Journal
Environmental Science and Technology.
See: Schoer, et al. (submitted)
-
Assistance in the development and maintenance of Raw Material
Equivalents page 5
conversion factors and calculation of RMC time series
For the measurement of the resource productivity the domestic
material consumption
(DMC) is related to the gross domestic product (GDP). Therefore
DMC is a key indicator
for measuring the use of natural resources in an economy. This
indicator is derived from
economy-wide material flow accounts (EW-MFA) which covers all
material inputs into na-
tional economies, the changes of stocks and their respective
outputs.
Domestic material consumption (DMC) is defined as the total
amount of material directly
used in a given economy. DMC is calculated by subtracting the
exports from the direct
material input (DMI) of an economy.
Now a couple of research projects and expert groups pointed out
that the value of DMI as
a basis for the DMC depends strongly from where the input is
coming from. If e.g. metal
ore is extracted domestically the total amount of ore is
accounted but if metals are im-
ported only their imported mass is used. This asymmetry led to
the proposal to express all
imported goods (and also exported goods) in terms of raw
material. Consequently all im-
ported semi-finished and finished goods have to be be expressed
in raw material equiva-
lents (RME).
Eurostat aims to convert the current asymmetric DMC indicator
into an indicator which is
expressed in raw materials equivalents (RMC). Conversion factors
are needed to translate
the masses of any imported good and exported good into their
mass expressed in RME. A
methodology had been developed in a past project2 how to
calculate these RME with a
reasonable effort. The methodology is based on an integration of
IOT data and life cycle
based data. That former project was conducted by the same
consortium as this project.
In the former project a first set of RME conversion factors were
developed and applied for
selected situations of the EU-27 and Germany as a pilot country.
The set of RME factors
where regarded as a first suggestion. It was the aim of this
project to consolidate the
method and the results and to develop an approach for
calculating time series of RME by
an automated procedure. A major issue for methodological
consolidation was to develop
an empirically broader founded "metal model" for estimating raw
materials coefficients for
imported metals.
2 Conceptual framework for measuring the environmental impact of
the use of natural resources
and products; Eurostat Contract 50304.2008.008-2008.715
-
Assistance in the development and maintenance of Raw Material
Equivalents page 6
conversion factors and calculation of RMC time series
The approach for calculating RME presented in this report is
fully integrated into the sys-
tem of Integrated Environmental and Economic Accounting (SEEA)
[UN (2003)] which is
designed for depicting the interaction between the economy and
the environment in a sys-
tematic and coherent manner. The economic system at European
level is described by
the European economic accounting system of ESA [ESTAT (1996)].
Input output tables
(IOT) are an integral part of the ESA. Applying an IOT based
approach for estimating
RME supports the objective of embedding those indicators into
the accounting system.
Behind that background a specific IOT based calculation model
was developed which is
called mixed Leontief / Life Cycle Assessment (LCA) approach.
This type of approach is
sometimes referred to as a hybrid LCA in the literature [Joshi
(2000)].
This report presents an annual automated calculation model for
converting product
flows into raw material equivalents (RME). The calculation
approach provides detailed
annual results on product flows in RME in a breakdown by the
following dimensions:
1. Categories of final uses and imports
2. 166 product groups
3. 52 raw material categories
1.2. The concept of raw material equivalents
The main purpose of establishing estimates in RME is to improve
the existing environ-
mental mass indicator DMC regarding meaningfulness and
interpretability. The improve-
ments refer to three main aspects:
a) Improvement of the mass flow indicator DMC by converting
product flows into
flows of raw material equivalents.
b) Establishing an ecological link by detailed disaggregation of
RME by 52 mate-
rial flow categories. A detailed disaggregation by material flow
categories takes
account of the fact that the different raw material categories
can have extremely
different pressure characteristics (environmental impact). That
disaggregation is
therefore a precondition of a meaningful ecological analysis and
interpretation of
material flows.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 7
conversion factors and calculation of RMC time series
c) Establishing an economic link by relating the direct and
indirect demand for
raw materials by 52 material flow categories to the final use of
products by 166
product groups. Further, the underlying IOT establishes a link
between the final
use of products and the direct economic production and
consumption activities.
That link is essential for embedding the material flow indicator
into an environ-
mental-economic context which is able to relate the demand for
raw materials to
the economic driving forces.
The term raw materials equivalent (RME) corresponds conceptually
to the term domes-
tic extraction used (DEU) of the EW-MFA system. The RME is
supposed to "convert"
trade flows expressed in simple product weight into their
equivalent of domestic extraction
used (DEU). The RME concept takes the perspective of raw
materials embodied in prod-
ucts. The RME of a product indicates how much extraction of
material was necessary over
the whole production chain for manufacturing that specific
product, irrespective whether
those raw materials were extracted from the domestic or the rest
of the world environ-
ment.
One important aim of physical flow accounting of the EW-MFA
system is to derive indica-
tors – one of the most important indicators being DMC - which
are used for the purpose of
the European Resource Strategy.
However, the DMC indicator inherently bears two major
asymmetries:
a) The imports and export are measured in a kind of different
"unit" than the
domestic extraction used (DEU). The latter is measured in virgin
raw mate-
rial extraction (e.g. tonnes of gross iron ore which is much
heavier than the
resulting steel). Trade is measured only in simple product
weight (e.g. ton-
nes of steel).
b) The composition of imports and exports can differ leading to
a further
asymmetry. For the EU one knows that the imports are dominated
by prod-
ucts with a rather low degree of processing, like metal ores or
primary en-
ergy carriers. These raw or semi-manufactured products are
usually much
heavier than more finished products and have a rather low RME
per tonne
product. EU exports are rather dominated by finished products,
like cars,
machinery etc. The weight of the finished products usually
represents only
-
Assistance in the development and maintenance of Raw Material
Equivalents page 8
conversion factors and calculation of RMC time series
a fraction of the weight of raw materials which were originally
extracted for
their manufacturing. This means that 1 million tonnes of
imported products
– measured in simple product weight – have a significant lower
RME than
1 million tonnes exported products.
The RME concept overcomes these asymmetries in considering the
imports and exports
measured in raw material equivalents which have the same "unit"
as the domestic extrac-
tion used (DEU).
Figure 1 illustrates the differences between original EW-MFA
figures and the correspond-
ing results which are converted into RME
Figure 1: Comparison EW-MFA and RME accounts
The figure shows that imports expressed in RME are double as
high as the figures from
EW-MFA accounts. The exports in RME are even more than three
times higher than the
original EW-MFA results.
Eurostat aims to establish economy wide indicators where the
trade components are
"converted" from product weight into RME. The following
accounting rule shows how the
indicators – following the RME concept – are derived:
6,711
1,681
513
7,932
6,711
3,401
1,676
8,435
DEU Imports Exports DMC / RMC
2005Mill tonnes
EW-MFA RME accounts
-
Assistance in the development and maintenance of Raw Material
Equivalents page 9
conversion factors and calculation of RMC time series
+ Domestic extraction used (DEU)
+ Imports in raw material equivalents (IMPRME)
= Raw material input (RMI)
– Exports in raw material equivalents (EXPRME)
= Raw material consumption (RMC)
It has to be pointed out that it is an important property of the
RME accounting system that
the above accounting identity is not only valid for total
materials, but also at the level of
individual raw material categories. Compared to that, the DMC
can only be disaggregated
to the level of aggregated raw material categories (biomass,
metal ores, non-metallic min-
erals and fossil energy resources), as semi-finished and
finished imported and exported
products can only be lumped together into rather broad material
categories like “products
mainly from biomass” or “products mainly from metals”.
Compared to the RME concept the TMR concept includes in addition
also unused extrac-
tion (e.g. mining overburden). Ideally TMR should be designed as
a methodical extension
of the RME accounts, i.e. RME supplemented by unused flows.
Technically that would
mean, that the RME including unused extraction are derived from
RME by multiplying the
RME of the individual raw material categories by a factor which
expressed the average
relationship between total (used plus unused) and used
extraction.
However in practice it is rather questionable whether it would
be worth to estimate total
extraction for data and methodical reasons.
As described below in section 2.6, the statistical coverage of
unused extraction is rather
poor for metals. Unused extraction of metal mining is a major
source of unused materials.
Therefore, including unused extraction is likely to increase
inaccuracy of the accounting
system significantly, as it is necessarily burdened by applying
quite rough estimates. Sec-
ondly, in a general manner it is rather disputable that the
specific element of unused ex-
traction correlates with environmental pressures or impacts.
However establishing mass
flow indicators like DMC, RMC or TMR is not an end in itself in
the sense that all materials
which are activated by economic activities should be counted in
the most comprehensive
manner, but the ultimate purpose is to indicate the development
of environmental pres-
sures or impacts in an aggregated manner.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 10
conversion factors and calculation of RMC time series
The structure of the report can be described as follows:
Chapter 2 of the report presents the RME calculation model. It
starts with some general
considerations and a short overview over the steps and the data
sources of the calculation
model. In the following subsections the principal steps of the
calculation model are de-
scribed in a more detailed manner. One central issue is the
“metal model” which provides
coefficients for converting metal flows into RME. In a final
section of that chapter it is tried
to assess the quality of the final results.
In chapter 3 the analytical properties of the RME accounting
system are discussed and
illustrated by analysing selected empirical results of the
project.
In chapter 4 issues for further methodological development of
the RME accounting system
are considered, as migration of the calculation system to new
classification, regionaliza-
tion of calculation of RME and the alternative approach of using
an IOT matrix with “inter-
nalized capital formation”.
2. The RME calculation model
2.1. General considerations
This project has developed an annual automated calculation model
converting product
flows into raw material equivalents (RME). The RME are estimated
by a specific Leontief
type calculation procedure.
Generally, the Leontief method is an approach of IOT analysis
which is based on an IOT
matrix and an environmental extension which carries the
information on generation of
pressures (direct pressures) by economic activities (homogenous
branches and final use
categories). By that approach, which is based on the so called
Leontief inverses, the di-
rect pressures are assigned to the products of final uses, as
exports, consumption expen-
diture and capital formation (embodied pressures). The embodied
pressures of imports
are generally estimated under the assumption that the imported
products are manufac-
tured under domestic production conditions ("domestic technology
assumption"). That
general Leontief type approach can also be applied to raw
material consumption by treat-
ing direct raw material consumption by production activities as
type of environmental ex-
tension.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 11
conversion factors and calculation of RMC time series
The Leontief approach for estimating embodied pressure is a
well-established approach in
environmental-economic analysis. For example, the
environmental-economic accounting
units of the Federal Statistical Office Germany and of EUROSTAT
are regularly calculat-
ing and publishing results on embodied air emissions
[Statistisches Bundesamt (2011)]
[Schoer et al (2007)]3.
The "standard Leontief approach" is based on the monetary IOT as
it is published by the
National Accounts and the pressure flow tables (environmental
extension) which are pro-
vided by the Environmental Accounts. The standard monetary input
output table (MIOT) at
European level has the format 60 product groups by 60
homogeneous branches
(MIOT60).
That standard Leontief approach was taken as a conceptual
starting point for designing a
method for converting product flows into RME at European level.
For that purpose annual
data from Eurostat are available for MIOT60 and data of the
material flow accounts (EW-
MFA) on domestic extraction used (DEU) of materials from the
environment by 42 mate-
rial categories.
But the standard approach had to be considerably modified in
order to cope with a num-
ber of deficiencies which had been identified with respect to
applying it for calculation of
RME.
Figure 2 shows the major deficits of the standard Leontief
approach and the solutions
which were adopted for establishing a specific approach for the
purpose of calculation of
RME.
Further the following issue had to be regarded, which is not a
specific feature of the stan-
dard Leontief approach. The degree of detail of the breakdown of
the domestic extraction
used (DEU) of the Economy Wide Material Flow Accounts by 42 raw
material categories
was considered not to be adequate for metals. Therefore the
standard breakdown was
disaggregated for metals (expanded material classification of
EW-MFA).
3 See Eurostat's reference database:
http://epp.eurostat.ec.europa.eu/portal/page/portal/environment/data/database
http://epp.eurostat.ec.europa.eu/portal/page/portal/environment/data/database
-
Assistance in the development and maintenance of Raw Material
Equivalents page 12
conversion factors and calculation of RMC time series
Figure 2: Leontief standard approach versus project approach for
RME calculation
The description of the calculation model below follows those
principal steps
2.2. Overview calculation model
The core of the calculation model is a specific annual expanded
hybrid input output
table of the format 166 product groups by 166 homogeneous
branches (expHIOT166).
That table is applied together with physical information of raw
material inputs from domes-
tic extraction and embodied RME of selected imported products
into the economy in a
type of Leontief approach (mixed Leontief / LCA approach).
Figure 3 illustrates the steps
and the data sources of the RME calculation.
Problem Solution
The degree of resolution of the
standard 60x60 IOT matrix is not
sufficient for tracking the flows of
individual raw materials.
Disaggregated monetary IOT of to the
size of 166x166 (expMIOT166)
Monetary use structures are in some
cases not appropriate for depicting
flows of raw material through the
economy
Establishing a hybrid IOT
(expHIOT166), i.e. monetary
structures of the expMIOT166 are
replaced by use structures in physical
units for selected product groups
The "domestic technology
assumption" is not suitable for
representing the raw material content
of a number of imported products in a
sufficient manner
External approach for estimating RME
of selected imported products (mixed
Leontief / LCA approach) was
introduced to the model.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 13
conversion factors and calculation of RMC time series
Figure 3: Calculation steps and data sources for RME
calculation
(3) DE 2000 expMIOT166Expanded monetary IOT for Germany
166x166
(Structural information)
(7) EU expMIOT166Annual expanded monetary IOT for EU-27
166x166(iterative adjustment approach for integrating German
structural information into existing
EU-level framework)
(11) EU expHIOT166Annual expanded hybrid IOT for EU-27
166x166(replacement of selected monetary use structures by physical
use structures
(5) EU MIOT60 Annual monetary IOT for EU-27 60x60
(6) Balance column for total domestic uses and total line for
inputs by 166 product groups / omogeneous branches for EU-27
(9) Physical flow tablesbv 166 branches- Biotic raw materials-
Energy carriers- Metal ores and basic
metals- Other minerals
(2) DE 2000 MIOT72Monetary IOT for Germany 72x72
(1) Auxiliary information forDE 2000
- Supply and use table by about3000 product groups and 120
branches (unpublished.- MIOT 120x120 (unpublished)- Agricultural
IOT by 46
agricultural product groups andbranches
- External trade statistics,structural business statistics,
and other sources
(4) Annual auxiliary monetary information for EU-27- COMEXT-
Balance of payment
statistics- Structural business
statistics- Agriculural accounts- other sources
(10) EW-MFAAnnual domestic extraction used in tons by material
categoriesl
(8) Annual auxiliary physical information for EU-27- COMEXT
(tons)- Energy balance
statistics- Metal flow information from USGS and BGS (mine
production of ores, prices, recycling, typical uses)- Agricultural
statistics- other sources
(13) Metal modelAnnual conversion coefficients (tonnes traded
weigh to tonnes RME, total metal to primary metal) for imported
metals by 60 metal categories
(12) Auxiliary information for metal model for EU-27- annual
business reports for
about 160 mines - Information from life
cycle assessment (LCA)
- other sources
(14) Auxiliary information for environmental extension for
EU-27- COMEXT: imports in tons by 60
metal categories)
(15) EU ENVEXT166Annual environmental extension (52 raw material
categories) by 166 branches for domestic extraction and imports
(selected product groups)
(16) EU Raw Material Equivalents (RME)Annual output, imports
exports and 6 final domestic use categories in RME by 52 raw
material categories and 166 product groups(calculation by mixed
Leontief / LCA approach)
Annual calculation of Raw Material Equivalents (RME)
Calculation for Germany 2000
Annual calculation for EU-27
-
Assistance in the development and maintenance of Raw Material
Equivalents page 14
conversion factors and calculation of RMC time series
The EU level calculation procedure can be subdivided into the
following principal steps:
a) Annual expansion of the monetary EU-level standard IOT 60x60
to the size of
166x166 (expMIOT166).
b) Converting the expMIOT166 into an annual hybrid IOT
(expHIOT166) by insert-
ing physical use structures for selected product groups
c) Estimation of annual environmental extensions (ENVEXT166),
i.e. input of raw
materials into the economy
d) Converting product flows into RME by a specific Leontief
approach
Ad (a): The expMIOT166 can be regarded as a necessary interim
step for estimating
expHIOT166. However, data of that IOT are also required for
appoaches of analysing the
the "economic link".
The expMIOT166 at EU-level is estimated by fitting structural
information from the
German expMIOT166 into the EU-level data framework of MIOT60 and
monetary vectors
by 166 product groups / homogeneous branches for outputs,
imports, exports and inputs.
Major data sources for estimating the German expMIOT166 are the
national German
MIOT72 and detailed information from internal supply and use
tables of the format about
3000 product groups x 120 homogeneous branches and agricultural
IOTs by 46
agricultural production processes.
Ad (b): The expHIOT166 is obtained by replacing selected
monetary use structurs of the
expMIOT166 by physical use structurs. The physical use
structures are obtained from flow
tables in different physical units for biomass products
(excluding animal production), metal
ores and basic metals, non-metallic minerals and energy
carriers. It was considered that
for those selected products the underlying raw material flow are
more closely represented
by physical than by monetary information.
Ad (c): Eenvironmental extensions in a breakdown by 166
homogenous branches
(ENVEXT166) are required as an input to the Leontief calculation
together with the ex-
pHIOT. In the context of calculation of RME the ENVEXT denotes
the input of raw mate-
rial into the economy in physical units. Following the specific
approach of Leontief calcula-
tion of this project two types of ENVEXT166 are required for
each raw material category:
1. domestic extraction used of raw materials,
2. raw materials embodied in selected imported products (called
"LCA products").
-
Assistance in the development and maintenance of Raw Material
Equivalents page 15
conversion factors and calculation of RMC time series
The RME of LCA products are estimated by an external approach
which is based on the
"metal model". The metal model provides RME coefficients (tonne
RME per tonne traded
weight) which are combined with data from external trade
statistics (COMEXT) on imports
in tonnes traded weight.
Ad (d): In the final step RME of product flows are estimated by
a "mixed Leontief / LCA
approach". Whereas the standard Leontief approach estimates RME
by assuming that
the imported products are produced under the same conditions
than the domestic prod-
ucts, the mixed Leontief / LCA approach avoids the domestic
technology assumption for
"LCA products" by using external information which enters the
calculation system as an
additional environmental extension for imported products.
The calculation approach provides detailed results on product
flows in RME in a break-
down by the following dimensions:
1. Categories of final uses (consumption expenditure, capital
formation, exports)
and imports
2. 166 product groups
3. 52 raw material categories
2.3. Disaggregated monetary IOT
2.3.1. The concept and rationale for disaggregating the IOT
The standard 60x60 IOT is primarily designed for analysing
economic relationships. With
respect to the relationship between the environment and economy
it has to be regarded
that – depending on the environmental problem under
consideration – the focus may
change. Some product flows which are not important in terms of
income generation or
other economic aspects can be highly relevant under an
environmental perspective. In
those cases the standard IOT may not able to provide the
necessary degree of detail for
tracking those flows.
A high level of aggregation may seriously impair the accuracy of
RME estimation for a
number of reasons:
-
Assistance in the development and maintenance of Raw Material
Equivalents page 16
conversion factors and calculation of RMC time series
a) Flows of a raw material are usually related to rather
specific production activi-
ties, as far as the first steps in the production chain are
concerned
b) The degree of inhomogeneity of a product group usually
increases with the level
of aggregation
Ad (a): With respect to estimating embodied raw material an
important feature of raw ma-
terial flows compared to many other pressure flows has to be
taken into consideration.
Other pressure flows, as for example CO2 emissions, usually
originate from a wide range
of production and consumption activities. Compared to that, each
raw material enters the
economy by a specific production activity (extraction). And also
the following step of trans-
formation is usually confined to one or a rather limited number
of specific production ac-
tivities (primary processing). Therefore it is very crucial to
depict those first steps in a suf-
ficient degree of detail for tracking the flow of a raw material
through to economy. How-
ever, in the standard IOT those flows are presented only in a
rather aggregated manner.
The production of all agricultural crops is grouped together
with animal production in one
product group / homogenous branch "agriculture". A substantial
part of primary processing
of agricultural crops as input to animal production takes place
within agriculture. And the
food production which is the major primary processing branch for
agricultural crops and
animal products outside agriculture is again grouped to only one
branch.
A similar situation can be stated for metal ores and other
mining and quarrying products
with one branch each for extraction and one (in case of metals)
or two (in case of other
mining) primary processing branches.
Ad (b): It has to be pointed out that the degree of
inhomogeneity – in terms of monetary
value per unit of raw material equivalent or some other weight
unit – is likely to increase
with the level of aggregation. That is the composition of
imports, exports or intermediate
consumption of such a rather aggregated product group may
differ. For example, the ex-
ports and the imports of the product group "other mining and
quarrying" (CPA 14) are in
monetary terms highly dominated by "Other mining and quarrying
products n.e.c" (CPA
14.5) whereas the domestic use is prevailed by stone, sand and
clay which have a much
lower price than other mining products. For reference see annex
1 Product classification
for expHIOT. RME calculation which is based on monetary
relationships assumes that
the average price for exported products of a product groups is
identical with the prices for
domestic uses of that product groups. That is not the case, if
at the same time the struc-
-
Assistance in the development and maintenance of Raw Material
Equivalents page 17
conversion factors and calculation of RMC time series
ture of the exports within a product group differs from final
domestic uses and the prices of
the components are not nearly at the same level. In that case an
estimation of RME which
is based on monetary relationships for a product group is not
able to provide a realistic
result.
As a solution for overcoming those deficits of the standard IOT
an expanded IOT matrix of
the size 166 product groups by 166 homogeneous branches
(expMIOT166) was devel-
oped as a first step.
Raw materials are reported by the calculation system in a
breakdown by 52 raw material
categories. In order to meet the requirements for calculation of
RME for each of those raw
material categories a corresponding product group was
established for the purpose of
expMIOT166. Furthermore, also the branches of primary processing
of raw materials, like
agricultural animal production, food production, basic metal
production, other non-metallic
mineral products and energy transformation were disaggregated.
In addition some further
branches which are considered to be important under the
perspective of raw material con-
sumption (e.g. chemical industry, metalworking industry) are
also presented in a disag-
gregated manner. See annex 1.
Beyond improving the accuracy of the calculation results for
RME, the disaggregated
IOT has a further important advantage. The breakdown of the
calculation results by type
of product follows the classification of the underlying IOT.
Therefore also the analytical
relevance of the RME indicators is strengthened by establishing
an “economic link” at a
much more detailed level.
2.3.2. Calculation method and data sources
The expansion of the MIOT60 to the format of MIOT166 means that
individual cells of
the MIOT60 have to be disaggregated. That disaggregation
requires two principal types of
auxiliary information:
a) Structural information for those arrays which have to be
disaggregated and
b) Information for subdividing the IOT balance column for total
domestic use and the
total line for inputs to the size of 166 product groups /
homogeneous branches.
A systematic overview of all important data sources is presented
in annex 2 data sources.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 18
conversion factors and calculation of RMC time series
2.3.2.1. Structural information based on German expanded IOT
At European level only few auxiliary data are available
establishing the required struc-
tural information. Exceptions are energy (energy balance data in
physical terms and
some information from SBS on purchases of energy carriers) and
agriculture (data on use
of fodder crops from the agricultural statistics).
Compared to that the data situation for Germany is much more
favourable. The German
economy accounts for roughly 20 per cent of total EU economy.
Therefore, in a first step
an expanded German IOT for the year 2000 was established with
the objective to utilize
the structural information of that IOT matrix for disaggregating
the European IOT.
The standard IOT of the German Statistical Office has the format
72 x 72. The expansion
of that IOT was predominantly based on detailed supply and use
tables of the size 3000
product groups by 120 homogeneous branches. Those detailed
tables are not published.
They are designed as an internal tool of the German national
accounting department for
tuning different parts of the national accounting system and for
generating the published
IOT. Those parts of that table which carry significant
information for disaggregating the
standard IOT to the size of 166x166 where made available to this
project by the German
Federal Statistical Office.
The breakdown by products of the detailed supply and use table
widely covers the re-
quirement for estimating the breakdown of the expMIOT by 166
product groups, with the
exception of disaggregation for other metals (CPA 2745,
excluding nickel). For further
subdividing that item the output, the imports and the exports
information from COMEXT
and a special preparation of the German production statistics
for metal was used. Such
original disaggregated information was available for total
supply (imports plus output) and
for exports. As far as the domestic use is concerned, it was
generally assumed that the
disaggregated rows have the same use structure as the aggregated
row. However some
information from life cycle inventories on typical composition
of steel alloys (e.g. nickel,
manganese, chromium. tungsten, titanium, molybdenum) was applied
for modifying the
use structures of those metals.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 19
conversion factors and calculation of RMC time series
For the disaggregation by homogeneous branches further
information was required for
disaggregating agriculture (18 branches), metal ores (17
branches) and non-ferrous basic
metals (16 branches).
For subdividing agriculture a detailed agricultural IOT by 46
agricultural production
processes for the year 1999 could be used. The agricultural IOT
was estimated in a joint
project of the Johann Heinrich von Thünen Institut and the
German Federal Statistical
Office [Schmidt, Osterburg (2009)]. The agricultural IOT
provides detailed information on
intra agricultural flows. But other inputs to the agricultural
production processes are only
reported in a summary manner. However some additional auxiliary
information from the
German Ministry of Agriculture on the allocation of important
non-agricultural inputs (e.g.
animal fodder from food industry, fertilizers, pesticides,
energy and medical services)
could be used for filling the most important gaps.
Regarding disaggregation of agriculture it has to be paid some
attention to the issue of so
called ancillary activities. It is a special feature of
agricultural production that quite fre-
quently different crops and animal products are produced in the
same holding. Above all
crop and animal production are strongly intermingled. Animals
consume fodder crops,
crop residues like straw and gazed biomass on the one hand and
on the other hand farm
manure which can be regarded to be a by-product of animal
production is used as input
into crop production. Also within animal production there are
interrelationships, like milk
consumption by animals.
Generally, the agricultural accounts treat the outcomes of
activities which are usually mar-
keted as products. That is, they are regarded also as output and
as input, if consumed in
the same holding. But the results of other activities which are
usually not marketed, like
grass which is directly taken up by the animals (grazed biomass)
or farm manure, are re-
garded as ancillary activities which are not treated as output
but only as an interim stage
within a production process. For example milk is the output of a
specific production proc-
ess. The milk producing animals may consume grazed biomass.
Input which are neces-
sary for producing that grazed biomass are assigned to
production of milk, but the grazed
biomass does not appear as a separate output. And the other way
round also the produc-
tion and use of farm manure is only regarded as an internal
flow.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 20
conversion factors and calculation of RMC time series
However if total agriculture is subdivided into a branches
production of grazed biomass
and of milk the flows between both activities (e.g. grazed
biomass and farm manure) have
to be made explicit by treating those flows as outputs and
inputs. Doing so would mean
that the total outputs from and the total inputs to agriculture
are higher for the disaggre-
gated IOT than for the standard IOT. That effect has also to be
regarded accordingly for
disaggregating the EU level IOT.
Regarding the disaggregation of metal branches it has to be
noted that iron and casting
services are shown separately in the German 72x72 IOT. The ore
input into basic non-
ferrous metal production was disaggregated by the general
assumption that the ore of a
metal almost fully enters basic metal production of the
corresponding metal. Some rather
insignificant corrections where made for alloys. The other
inputs into metal production
were widely allocated according to output relationship. Energy
inputs into extraction and
primary processing of metals were cross checked and adjusted by
some information from
life cycle inventories on typical energy consumption.
2.3.2.2. EU level balance column and line total
The second element for expanding the MIOT60 is monetary balance
columns and line
totals by 166 product groups / homogeneous branches for total
domestic use and for in-
puts.
The total domestic use is obtained by the following
calculation:
Outputs + imports = total supply = total use - exports = total
domestic use
Annual European information is widely available from Eurostat’s
online reference data
base4 for disaggregating the above items to the size of 166
product groups or homogene-
ous branches.
Monetary data from the European external trade statistics
(COMEXT) are applied for dis-
aggregating the import and export vector for goods. The product
groups for services are
4 Eurostat. Eurostat’s reference database:
http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/themes
http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/themes
-
Assistance in the development and maintenance of Raw Material
Equivalents page 21
conversion factors and calculation of RMC time series
disaggregated by information from Balance of Payment Statistic
(BoP). Main source for
the subdivision of the output and the input vector is the
Structural Business Statistic
(SBS). The SBS data set had to be gap filled in order to cope
for missing values, mainly
due to suppressing of data for confidentiality reasons. For
agriculture information from
agricultural accounts on output structure can be utilized. For
full disaggregation of metal
mining and basic metal production it has also to be referred to
data from the Production
Statistics (PRODCOM) and information from USGS and BGS on
physical production and
on metal prices.
2.3.2.3. Iterative adjustment approach
The EU level expMIOT166 is estimated by fitting in the
structural information from German
expMIOT166 to the European data framework of MIOT60 and the
disaggregated balance
column and line total by a RAS-type approach5. The
disaggregation of individual cells of
the European MIOT60 is conducted by three principal steps:
a) Estimation of raw values by disaggregating individual cells
of the European IOT
by structural information from the German IOT,
b) Adjustment of the raw values of step a) to the vector of
total domestic uses and
to the corresponding cells of the MIOT60 in an iterative
adjustment approach,
c) Adjustment of the results of step b) to the vector of total
inputs and the corre-
sponding cells of the MIOT60 in an iterative adjustment
approach.
The principal steps of the calculation approach are
schematically illustrated by Figure 4 to
Figure 6.
The green cells represent the European data framework and the
red cells denote the dis-
aggregated values. In the first calculation step raw values are
estimated by subdividing
individual cells (indicated by the arrows) of the EuropeaMIOT60
IOT (light green) by verti-
cal, horizontal or matrix type disaggregation. Those raw values
sum up to the disaggre-
gated cell but they do not match to the vectors for total inputs
and total final domestic
uses.
5 The general RAS method is an approach to reconcile
inconsistencies in data that should match or
sum up to the same amount. “It is a bi-proportional adjustment
algorithm that balances matri-
ces in a mechanical way. See: [Bouwmeester (2007)]
-
Assistance in the development and maintenance of Raw Material
Equivalents page 22
conversion factors and calculation of RMC time series
Figure 4: Scheme for the disaggregation of 60X60 IOT to the
format of 166X166 (Step 1)
Disaggregation of 60x60 IOT to the format of 166x166
Step (1) raw values
Schematic description
(1) Adjustment of German structural information to the cells of
the EU-level MIOT60
To
tal
do
mesti
c u
ses b
y 1
66 p
rod
uct
gro
up
s
Total inputs / final uses by 166 homogeneous branches / final
use categories
EU-level framew ork Totals by 166 product groups /homogeneous
branches
Cells of 60x60 IOT not disaggregated
Cells of 60x60 IOT disaggregated
German structural information f it
into the EU-level framew ork by an
iterative adjustment approach
Cells of 166x166 IOT estimated by horizontal disaggregation
Cells of 166x166 IOT estimated by vertical disaggregation
Cells of 166x166 IOT estimated by matrix shape
disaggregation
-
Assistance in the development and maintenance of Raw Material
Equivalents page 23
conversion factors and calculation of RMC time series
Figure 5: Scheme for the disaggregation of 60X60 IOT to the
format of 166X166 (Step 2)
Step two shows the adjustment of the raw values of step one to
the total domestic use
vector in an iterative approach by observing the coherence to
the cells of the MIOT60.
The arrows indicate the adjustment steps.
Disaggregation of 60x60 IOT to the format of 166x166
Step (2)
Schematic description
(2) Adjustment of (1) to total domestic uses and MIOT60 by an
iterative approach
To
tal
do
mesti
c u
ses b
y 1
66 p
rod
uct
gro
up
s
Total inputs / final uses by 166 homogeneous branches / final
use categories
EU-level framew ork Totals by 166 product groups /homogeneous
branches
Cells of 60x60 IOT not disaggregated
Cells of 60x60 IOT disaggregated
German structural information f it
into the EU-level framew ork by an
iterative adjustment approach
Cells of 166x166 IOT estimated by horizontal disaggregation
Cells of 166x166 IOT estimated by vertical disaggregation
Cells of 166x166 IOT estimated by matrix shape
disaggregation
-
Assistance in the development and maintenance of Raw Material
Equivalents page 24
conversion factors and calculation of RMC time series
Figure 6: Scheme for the disaggregation of 60X60 IOT to the
format of 166X166 (Step 3)
The result of step two is adjusted in a final iterative approach
to the vector of total inputs
into intermediate consumption and of total final uses for the
individual final use categories
by observing the coherence to the cells of the MIOT60.
Disaggregation of 60x60 IOT to the format of 166x166
Step (3)
Schematic description
(3) Adjustment of (2) to total inputs / final uses and MIOT60 by
an iterative approach
To
tal
do
mesti
c u
ses b
y 1
66 p
rod
uct
gro
up
s
Total inputs / final uses by 166 homogeneous branches / final
use categories
EU-level framew ork
Cells of 60x60 IOT not disaggregated
Cells of 60x60 IOT disaggregated
Totals by 166 product groups /homogeneous branches
German structural information f it
into the EU-level framew ork by an
iterative adjustment approach Cells of 166x166 IOT estimated by
vertical disaggregation
Cells of 166x166 IOT estimated by matrix shape
disaggregation
Cells of 166x166 IOT estimated by horizontal disaggregation
-
Assistance in the development and maintenance of Raw Material
Equivalents page 25
conversion factors and calculation of RMC time series
As a result a fully coherent monetary IOT matrix of the format
166x166 is obtained. It has
to be pointed out that that IOT to a large extent is an
approximation to real European rela-
tionship and not the German conditions, as the German structural
information was not
simply adjusted to the European balance column and total line.
The disaggregated IOT
fully carries the information on European average production
technology as it is repre-
sented in the European MIOT60. Only below that level German
technological relations
have been used. But also for that level it has to be considered
that the structural informa-
tion from the German IOT was further adjusted to European demand
and input totals.
2.3.2.4. Data issues
Structural Business Statistics
Regarding Structural Business Statistics (SBS) data two major
data issues have to be
mentioned, filling of data gaps and transition to new
classification beginning in 2008.
Gap filling: At the EU-27 level there are many data gaps at the
required level of detail,
mainly due to confidentiality cases or non-reporting of
individual countries for some years.
Those gaps were filled by the following general approach: It was
referred to country level
data. At that level most gaps could be filled by referring to
corresponding results of previ-
ous or following years. In some cases further auxiliary
information had to be consulted,
like information from COMEXT, BGS, USG and even annual business
reports.
Change of classification: Data of the SBS are reported in NACE
Rev. 1.1 until the year
2008. For 2008 data are reported also according to the revised
classification NACE Rev. 2
and after 2008 data are only available in NACE Rev. 2. The
calculation model is based on
CPA 2002, which fully corresponds to NACE Rev. 1.1, except that
the branches according
to CPA are reported as homogeneous branches whereas branches
according to NACE
are demarcated as local kind of activity unit (establishments).
From 2009 on SBS data
have to be recoded from new to old classification. For that
purpose a conversion key was
developed which is based on two elements:
a) A correspondence table which conceptually relates the
individual items of both
versions of the classification to each other at the most
detailed level.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 26
conversion factors and calculation of RMC time series
b) An empirical conversion key which indicates what proportion
of an item of the
one classification has to be assigned to what item of the other
classification.
The conceptual correspondence tables are provided by Eurostat6.
The empirical conver-
sion tables are based on the conceptual correspondence tables
and on double prepara-
tions according to both versions of the classification for one
year. The empirical conver-
sion matrix for SBS shows what proportions of the items of the
NACE Rev. 2 are assigned
to the items of NACE Rev. 1.1. The disaggregation of NACE Rev. 2
follows the 4 digit
level. NACE Rev. 1.1 is allocated according to the “166 level
breakdown” of the expHIOT,
which is a mix of 2 digit, 3 digit, 4 digit and some special
disaggregation of 4 digit level
items (metals). That matrix was estimated for production
values.
Beginning in 2009 the original data for production values and
purchases of goods and
services of SBS are converted by that matrix into the project
specific 166 level breakdown
of NACE Rev. 1.1.
COMEXT
In some cases original COMEXT data had to be corrected. The
corrections refer to both,
results in EUR as well as in metric tonnes. The corrections go
back to two major reasons,
apparent assignment errors and non-plausibility.
A number of apparent assignment errors were especially detected
for gold and some oth-
er high priced metals.
Cross checking with external trade data from BGS lead to
corrections of import and export
figures in case of some high priced metals.
In case of tin applying the original COMEXT data lead to
inconsistent IOT relationships
(negative domestic uses) for some years.
Therefore it has to be noted that especially the calculation of
separate results for gold and
tin should be used with care.
6 See: RAMON Eurostat’s metadata server:
http://ec.europa.eu/eurostat/ramon/index.cfm?TargetUrl=DSP_PUB_WELC
http://ec.europa.eu/eurostat/ramon/index.cfm?TargetUrl=DSP_PUB_WELC
-
Assistance in the development and maintenance of Raw Material
Equivalents page 27
conversion factors and calculation of RMC time series
2.4. Disaggregated hybrid IOT
2.4.1. The concept and rationale for a hybrid IOT
For the specific purpose of calculation of RME the environmental
IOT of the type of
expMIOT166 was further developed by converting it into a hybrid
IOT (expHIOT166)
where monetary use structures were replaced by use structures in
physical units for some
selected product groups7.
The aim of that modification is to arrive at an improved
presentation of flows of raw mate-
rials through the economy by the use structures of the IOT.
For analysing raw material flows it would be crucial that the
use structures for the ex-
tracted raw materials (raw material products) represent the
physical flows as much as
possible. This may hold to a certain extent also for products of
primary processing of raw
materials, like basic metals or refinery products which still
largely contain the original raw
materials8.
In principal the aim of depicting physical flows of raw material
products and of primary
processed raw materials could also be achieved by applying
monetary use structures.
However monetary units can only work well if there are no
significant price effects, i.e. if
the unit value is the same for all uses. However, it can be
shown that this is sometimes
not the case in practice.
Physical use structures for raw materials (raw products) and
primary processed raw mate-
rials are required for the following reasons:
a) Price differentiation: different users pay different prices
for the same product
b) Structural effects: the composition within disaggregated
product groups differs by
users
7 An early approach for a hybrid IOT was developed by Beutel and
Stahmer. See: [Beutel (1982)]
8 There were some attempts in the past by statistical offices
also to estimate full physical input out-
put tables (PIOT) where all use structures are expressed in mass
units. One example is the PIOT of the German Statistical Office.
However to establish such a table is extremely data demanding. In
case of Germany the calculation were predominantly based on
detailed mone-tary supply and use tables by 3000 product groups.
For German PIOT see: [Waldmüller. (2001)]
-
Assistance in the development and maintenance of Raw Material
Equivalents page 28
conversion factors and calculation of RMC time series
Case a) is especially relevant for energy carriers. Examples for
case b) are: Significant
structural differences for exports and domestic uses for other
minerals n.e.c; differences in
the share of secondary metal for exports and domestic uses;
prepared animal feed (CPA
156.7) for pets and farm animals has considerably different
prices.
Table 1 compares the prices as reported by COMEXT of imports
with the prices of exports
for raw materials and some products which represent the stage of
primary processing of
raw materials.
That comparison reveals that the prices for imports and exports
differ for almost all prod-
uct groups9 which are presented in the table. Moreover the
differences change over time
in many cases. In some cases the differences may simply reflect
statistical inconsisten-
cies. But more generally it can be assumed that different prices
rather reflect structural
effects.
It has to be regarded that especially the quantitative
relationship between the imports and
exports in RME is most crucial for estimating the central
indicator RMC. Therefore, it can
be concluded that applying physical instead of monetary use
structures for raw materials
and some products of primary processing is likely to improve the
accuracy of the calcula-
tion results for RME.
It has to be pointed out that for none of the products under
consideration full physical in-
formation is available for converting monetary into physical use
structures. Therefore, the
estimated physical use structures have to be always based on a
mixture of physical and
monetary information. In most cases only imports, exports and
output are available in
physical terms and monetary relationships have to be applied for
disaggregating of do-
mestic uses in physical terms. Only for energy carriers the
energy balance provides rather
detailed physical information on the consumption of energies by
economic activities.
9 As a matter of fact unit prices for imports and exports prices
may not only differ for raw materials,
but also for semi-finished and finished products. However, for
those more complex products physical use structures are not likely
to be advantageous compared to monetary structures for estimating
the raw material content.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 29
conversion factors and calculation of RMC time series
Table 1: Relationship unit value of imports to unit value of
exports
Selected product groups
Excluding scrap
Classification for expHIOT 2000 2001 2002 2003 2004 2005 2006
2007 2008 2009
01.11.1 Cereals 1.4 1.2 1.1 1.2 1.2 1.3 1.2 1.0 1.0 1.3
01.11.21 Potatoes 1.4 1.3 1.1 1.3 1.1 1.2 0.9 0.9 1.1 1.1
01.11.22 Dried leguminous vegetables, shelled 0.9 1.3 1.8 2.0
1.3 1.1 1.4 1.6 1.7 2.1
01.11.23
Edible roots and tubers with high starch or
inulin content 0.2 0.3 0.1 0.1 0.1 0.2 0.3 0.2 0.4 0.9
01.11.3 Oil seeds and oleaginous fruits 1.1 0.9 0.8 0.8 0.9 1.0
0.9 0.7 0.6 0.9
01.11.4 Unmanufactured tobacco 1.7 1.7 1.8 1.4 1.3 1.2 1.3 1.2
1.1 1.0
01.11.5 Plants used for sugar manufacturing 2.7 3.0 24.7 47.3
23.6 92.9 21.8 41.7 29.2 23.2
01.11.7 Raw vegetable materials used in textiles 1.2 1.4 1.2 1.0
1.2 1.1 1.2 1.1 1.1 1.2
01.12
Vegetables, horticultural specialities and
nursery products 1.0 0.9 0.8 0.8 0.8 0.9 0.8 0.8 0.9 0.9
01.13 Fruit, nuts, beverage and spice crops 1.5 1.4 1.3 1.2 1.3
1.4 1.4 1.3 1.2 1.5
01.11.8, 01.11.9, 01.19Other crop products 0.7 0.8 0.8 1.0 0.9
1.0 1.3 1.0 1.0 1.1
02
PRODUCTS OF FORESTRY, LOGGING
AND RELATED SERVICES 0.7 0.7 0.8 0.8 0.8 0.8 0.7 0.8 1.1 0.8
05
FISH AND OTHER FISHING PRODUCTS;
SERVICES INCIDENTAL TO FISHING
0.8 0.8 0.9 0.7 0.7 0.7 0.8 0.8 0.7 0.9
10.1, excl 10.10.12 Coal, not agglomerated 1.0 1.1 0.9 1.0 0.8
0.9 0.9 0.9 0.9 0.8
10.2.a Lignite, not agglomerated 0.8 1.1 0.3 0.8 0.4 0.6 0.5 0.5
0.7 0.8
10.3 Peat 0.7 0.7 0.6 0.6 0.5 0.5 0.6 0.6 0.5 0.5
11.10.1
Petroleum oils and oils obtained from
bituminous minerals, crude 1.0 1.0 1.0 1.0 1.0 1.0 0.9 1.0 1.0
1.0
11.10.2
Natural gas, liquefied or in gaseous state
0.8 0.9 0.9 0.9 0.8 0.8 0.8 0.9 0.8 0.7
13.1 Iron ores 0.9 0.8 0.8 0.8 0.9 0.7 0.6 0.8 0.7 1.0
13.20.11 Copper ores and concentrates 1.3 1.8 1.1 1.3 1.2 1.4
1.4 1.3 1.1 1.2
13.20.12 Nickel ores and concentrates 1.4 1.0 1.0 0.7 0.7 1.9
2.9 2.1 3.6 3.1
13.20.13 Aluminium ores and concentrates 1.3 1.2 1.0 1.0 0.9 1.2
0.9 0.7 0.9 0.5
13.20.14.a Gold 601.9 43.2 48.1 59.2 80.3 679.7 24.5 48.7 28.4
236.2
13.20.14.b Silver 3.6 39.9 11.5 2.3 8.7 49.9 1.1 81.5 18.7
0.3
13.20.15.a Lead 2.9 2.9 3.2 1.8 1.7 1.8 2.5 1.8 2.2 2.3
13.20.15.b Zinc 1.0 1.1 0.7 0.7 0.9 0.9 0.9 0.9 1.0 1.2
13.20.15.c Tin 1.2 0.6 0.0 0.3 0.4 1.1 1.5 0.1
13.20.16.a Tungsten ores and concentrates 0.6 0.9 0.4 12.6 0.6
0.8 0.6 0.7 0.8 2.0
13.20.16.d
Titanium ores (Ilmenite) and concentrates
0.2 0.3 0.4 0.4 0.6 0.1 0.1 0.3 0.4 1.4
13.20.16.e Manganese ores and concentrates 0.8 0.9 1.2 0.8 0.9
1.1 1.1 1.0 1.0 1.0
13.20.16.f Chromium ores and concentrates 0.5 0.5 0.3 0.3 0.4
0.3 0.4 0.4 0.6 1.1
13.20.16.g Other ores and concentrates 0.7 1.0 0.8 0.8 1.3 1.0
1.0 1.7 2.2 1.6
14.1 Stone 1.6 1.3 1.3 1.2 1.2 1.3 1.7 1.6 1.4 1.4
14.2 Sand and clay 1.3 1.3 1.1 1.0 1.0 1.0 0.9 0.9 1.0 0.9
14.3 Chemical and fertilizer minerals 0.6 0.6 0.6 0.5 0.5 0.5
0.5 0.5 0.7 0.9
14.4 Salt 0.6 0.5 0.5 0.5 0.5 0.5 0.4 0.5 0.6 0.4
14.5
Other mining and quarrying products
n.e.c. 0.3 0.2 0.2 0.2 0.3 0.3 0.3 0.2 0.2 0.3
23.1 Coke oven products 0.7 0.8 0.7 0.8 0.9 0.9 0.9 1.0 1.1
1.5
23.2 Refined petroleum products 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9
1.0 1.0
23.3 Nuclear fuel 1.2 1.0 0.8 0.7 0.5 0.7 0.7 1.2 1.0 0.5
27.1-3
Basic iron and steel and ferro-alloys,
tubes and other first processed iron and
steel 0.7 0.7 0.7 0.7 0.8 0.7 0.6 0.6 0.7 0.7
27.41.1, 27.41.5, 27.41.62.a Silver and silver products 1.8 1.6
1.9 1.2 1.1 1.3 0.8 1.1 1.4 1.3
27.41.2, 27.41.4Gold and gold products 1.2 1.5 3.0 4.4 1.7 1.1
1.1 1.2 1.4 1.2
27.41.3, 27.41.62.b Platinum and platinum products 1.2 1.3 1.0
1.2 1.2 1.1 1.0 1.0 1.0 1.0
27.42 Aluminium and aluminium products 0.6 0.5 0.5 0.5 0.5 0.5
0.6 0.6 0.5 0.4
27.43.11, 27.43.21-3 Lead and lead products 1.0 1.0 0.8 1.6 2.1
1.2 1.7 1.9 1.4 1.4
27.43.12, 27.43.24-6 Zinc and zinc products 0.8 0.7 0.7 0.8 0.9
0.6 0.8 0.7 0.7 0.8
27.43.13, 27.43.27-9 Tin and tin products 1.0 1.1 0.9 1.0 1.0
0.7 0.5 0.6 1.2 1.1
27.44 Copper products 0.7 0.6 0.6 0.6 0.7 0.7 0.7 0.7 0.7
0.7
27.45.1, 27.45.2, 27.45.42Nickel products 0.7 0.6 0.5 0.7 0.7
0.7 0.8 0.8 0.7 0.7
27.45.a Tungsten products 0.9 0.7 1.1 0.9 0.8 0.8 0.9 0.6 0.4
0.6
27.45.b Tantalum products 2.1 1.6 1.9 1.2 1.6 1.8 1.3 1.6 1.5
1.4
27.45.c Magnesium products 0.5 0.4 0.6 0.5 0.6 0.4 0.4 0.5 0.6
0.5
27.45.d Titanium products 1.0 0.9 0.9 0.8 0.6 0.8 0.7 0.8 0.6
0.6
27.45.e Manganese products 0.5 0.5 0.5 0.6 0.7 0.7 0.6 0.8 0.8
0.7
27.45.f Chromium products 0.9 0.8 0.7 0.7 0.9 0.9 0.8 1.1 1.0
0.9
27.45.g Other non-ferrous metal products 0.5 0.4 0.4 0.4 0.7 0.3
0.2 0.3 0.4 0.3
Source: COMEXT
-
Assistance in the development and maintenance of Raw Material
Equivalents page 30
conversion factors and calculation of RMC time series
But notwithstanding, introducing physical information for
imports, exports and outputs
alone is a crucial step for improving the quality of the
calculation result for RME.
In a first step the monetary use structures for some product
groups were corrected in or-
der to cope for some identifiable price effects. In a second
step, by integrating elements of
that corrected monetary use structures, a set of different
physical flow table in a break-
down by 166 homogeneous branches was developed for the following
product categories:
a) biotic raw materials (agricultural crop production, wood,
fish)
b) energy carriers
c) metals (ores and basic metals)
d) other minerals
For establishing those physical flow tables different methodical
approaches are applied
and different physical units are used. Those flow tables are
inserted to the expMIOT166 in
order to convert it to an expHIOT166.
2.4.2. Calculation method and data sources
2.4.2.1. Price adjustment of monetary use structure
In a first step the monetary use structures for selected product
groups were corrected in
order to cope for some identifiable price effects. Those
corrections are based on plausibil-
ity checks which compare the inputs and the outputs of
individual product groups in physi-
cal units.
For example, for checking the plausibility of the input-output
relationship of swine pro-
duction (CPA 01.23) the physical output was compared to the
amount of relevant physi-
cal inputs. For that comparison the output of the product under
review and the relevant
inputs have to be converted into tonnes. With regard to swine
production cereals (CPA
01.11.1) and prepared animal feed (CPA 15.7) were identified as
the relevant inputs to be
considered. The monetary use structures of both input product
groups were converted into
mass units by using information in tonnes for imports, exports
and output and the relation-
-
Assistance in the development and maintenance of Raw Material
Equivalents page 31
conversion factors and calculation of RMC time series
ships of the monetary use structure for domestic uses. As a
result the inputs of cereals
and prepared animal feed into swine production were obtained in
mass units.
It turned out that the calculated relationship fodder input to
meat output was far below the
usual ratio of pig fattening as it is known from literature. In
particular the assumption of
equal prices for cereals as well as for prepared fodder for all
users, which was inherent to
the IOT based calculation approach, was identified as a reason
for that discrepancy. That
assumption neglects that the cereals which are used for animal
fodder and human con-
sumption represent different qualities which have different
prices. The same holds for pre-
pared animal feeds. The price for pet feed is much higher than
for farm animal feed.
Therefore, under the "equal price assumption" the amount of
cereals and of prepared
animal feed in tonnes which was allocated to swine production
was too low. In order to
cope for that effect of "price differentiation" the implicit
calculative prices were adjusted in
a manner that the physical input-output relationship for swine
production arrived at a plau-
sible order of magnitude.
Following that consideration the original monetary use structure
for cereals and prepared
animal feed were corrected accordingly in order to represent
rather the physical and not
the monetary use structures.
Similar considerations and corrections were made for the use
structures of fish, stone,
sand and clay and salt.
2.4.2.2. Physical use structure for biomass and other
minerals
For agricultural crop products, products from forestry and
fishery and for other mining and
quarrying products physical use structures were estimated in the
unit tonnes traded
weight. Physical information on imports and exports are taken
from COMEXT10 and the
information on output is obtained from EW-MFA (domestic
extraction)11. The physical use
structures for domestic uses were estimated by applying the
monetary relationship after
correction for price effects (see section 2.3.2.1).
10
It can be assumed that the RME of imported and exported raw
products of biomass and other
minerals are reflected quite well by the traded weight. However,
diamonds are an exception.
The special case of diamonds has not been explicitly regarded so
far in the calculation model.
But it is planned to take account of that issue during the next
round of updating. 11
The MFA compilation manual of Eurostat recommends estimating the
RME of domestic extrac-tion of those materials by referring to the
traded weight. See [ESTAT (2011)]
-
Assistance in the development and maintenance of Raw Material
Equivalents page 32
conversion factors and calculation of RMC time series
Those physical use structures take into account effects of
differences in the structural
composition of imports, exports and domestic uses. Moreover,
also within domestic use
activities also some important price or structural effects are
regarded.
2.4.2.3. Physical use structure for metals
Similar to the approach for biomass and other minerals the
benchmark values for outputs,
imports and exports of metals are estimated in physical units
and for the disaggregation of
the domestic uses the monetary relationships are applied.
However, compared to those materials there are some
peculiarities for metals which re-
quire a more complex estimation approach for the benchmark
values. One rather out-
standing feature is that the differences between tonnes traded
weight and tonnes RME
are extremely high for most metals. Also the trade flows of ores
are usually not reported in
gross ore but for concentrates. Another issue is that the
external trade figures for metals
have to be adjusted for scrap and secondary metal (which is made
of recycled metal).
Finally the lack of reliable direct information on output of
basic metals has to be regarded.
In a first step of the calculation approach the benchmark values
are expressed in metal
content. That step is necessary for adjusting exports for
secondary metal content. In a
final step the benchmark values are converted into the unit
tonne RME.
The model is based on a number of sources as:
a) Imports and exports of ores and basic metals (excluding
scrap) in tonne traded
weight (COMEXT)
b) Mine production of metal ores in tonne metal content
(BGS)
c) DEU of metal ores in tonne RME (EW-MFA)
d) Conversion coefficients tonne ore concentrate to tonne metal
content
(metal model)
e) Conversion coefficients tonne metal content to tonne RME for
ores and basic
metal (metal model)
f) Conversion coefficients tonne total metal to tonne primary
metal for imports
(metal model)
-
Assistance in the development and maintenance of Raw Material
Equivalents page 33
conversion factors and calculation of RMC time series
The RME concept refers to consumption of material that was
extracted from the environ-
ment, i.e. in the case of metal the amount of embodied ore.
Metal scrap and secondary
metal is not generated from new ore but goes back to ore which
was already counted
earlier. Therefore, it has to be excluded for the purpose of RME
calculation12.
The monetary IOT includes scrap and secondary metal which is a
further reason that
monetary relationships are not a suitable basis for estimating
RME for metals.
In a first step the benchmark values for imports, exports and
outputs are estimated in
tonne of metal content.
The imports and exports of ores in tonnes traded weight are
converted by coefficients
from the metal model into metal content. The outputs of ores
(mine production in tonne
metal content) are reported by BGS.
For converting the flows of basic metal into metal content of
primary metal, scrap and
secondary metal has to be excluded. As far as the data source
COMEXT for imports and
exports of basic metals is concerned, scrap can be excluded
easily, as it is reported sepa-
rately. But that is not the case for primary and secondary
metal, except for aluminium.
Therefore, average recycling coefficients for imported metal
from the "metal model" (see
section 2.6) are applied for estimating imports of basic metals
in metal content. The coef-
ficients represent the world average.
As no direct reliable and complete information on domestic
production of basis metals
is available in tonne of metal content that item was estimated
by assuming that the output
of basic metal is equal to the input of ore, which of course is
a simplification, as some of
the metal which is contained in the ore maybe gets lost. The
input of ore into the same
metal – which is usually quite near 100% – was estimated by the
relationships of the
monetary use structure.
The secondary metal content for exports cannot be simply
estimated by applying an ex-
ternal coefficient as it was used for imports. Exports of
secondary metals can be ap-
12
In principal this issue is also of relevance for other
materials, like wood as also a rather high share of paper, which is
the main use category for wood, originates from recycling. However
due to the lesser quantitative importance for the overall results
that issue was neglected so far for non-metal materials.
-
Assistance in the development and maintenance of Raw Material
Equivalents page 34
conversion factors and calculation of RMC ti