PDF format (503 kB) - ECN

Workshop Factor 2/Factor 10 Proceedings, Utrecht, 2 April 1998
Introduction
Concrete and binding targets for reduction of greenhouse gas (GHG) emissions for Annex-1 countries are now
agreed to in the Kyoto protocol. Emissions in the European Union will be reduced by 8% in the period 2008-
1012, compared to the emissions in the reference year (1990/95). In subsequent periods after 2010, further reductions
will be required in order to arrive at stabilisation of global GHG concentrations at acceptable levels as
called for by the UN Climate Convention. In this respect the Factor 4 and Factor 10 concepts, emerging from
the eco-efficiency discussion, come into play. Doubling of economic outputs, coupled to a simultaneous 50%
reduction of GHG emissions, requires a Factor 4 reduction of the specific emissions per unit of economic activity.
If this were a global goal, Western economies will in all likelihood face even much further reaching goals:
efficiency improvements as high as a Factor 10 might eventually become the longer term requirement.
In order to achieve such challenging goals, all available options and strategies must be utilised. One group of
strategies that has received relatively little attention for GHG policies focuses on bulk material flows in the
economy. Under current conditions, approximately one quarter of the GHG emissions can be attributed to the
production of materials, and these emissions can be reduced through changes in the materials use. However,
such strategies compete with strategies that reduce the emissions in the production of materials itself. An integrated
life cycle approach is required for proper assessment of the various materials related options in close
conjunction with the energy sector. The MATTER project (MATerials Technologies for greenhouse gas Emission
Reduction) is a joint project of five Dutch institutes within the framework of the National Research Programme
on Global Air Pollution and Climate Change that focuses on the potential to reduce GHG emission in
the life cycle of materials.
The goal of this "Factor 2/Factor 10" MATTER workshop is to indicate the potential of materials strategies for
GHG emission reduction. Five papers were presented, one focusing on integrated systems calculations with the
MARKAL model, three on product groups (passenger cars, packagings, and building and infrastructure) and
one on a materials group (metals). Together these case studies cover a significant part of the total materials
production and materials use.
The title of the workshop “Factor 2/Factor 10” refers to the widely accepted notion that a factor 2 emission reduction
is possible with technological improvements, while a factor 10 cannot be achieved through technological
improvements alone. The case studies intend to put this hypothesis to the test.
The workshop was attended by representatives from the industry, research and policy communities in the Netherlands.
The organizers hereby want to express their gratitude for the active participation of all, and recognizes
in particular the invaluable input provided by discussants and panelists. The National Research Programme
(NOP-MLK) kindly provided financial and logistic support; and last but not least the workshop chairman, professor
Han Brezet of Delft Technical University, is acknowledged for creating the lively and stimulating atmosphere
under which the event took place.

Programma
tijd onderwerp
10.00 Opening door de dagvoorzitter
Prof.dr.ir. J. Brezet, TUD
10.05 Presentatie NOP
drs. M. Kok, NOP-MLK
10.15 Overzicht MATTER studie
ir. T. Kram, ECN
10.30 Waarom materialenbeleid ?
drs.ir. D.J. Gielen, ECN
11.00 Verpakkingen
drs. M. Hekkert, NW&S
11.45 Transportmiddelen
drs. M. Bouwman, IVEM
12.30 Lunch
13.30 Gebouwen en infrastructuur
drs.ir. D. Gielen, ECN
14.30 Metalen
drs. B. Daniëls, IVEM
15.30 Pauze
16.00 Discussie beleidsrelevantie
17.00 Borrel

The Impact of GHG Emission Reduction on the Western European Materials System
D.J. Gielen
ECN-Policy Studies, PO Box 1, 1755 ZG Petten, The Netherlands
tel. +31-224-564460 E-mail: Gielen@ecn.nl
Preliminary analysis for the MATTER workshop Factor 2/Factor 10, Utrecht, 2 April 1998
Abstract
This paper discusses preliminary results for the analysis of Western European GHG emission reduction in the
materials system, based on MARKAL model calculations. The results show that the materials system contributes
up to 50% to the total emission reduction at moderate emission reduction goals. Major changes due to
GHG emission penalties occur in the materials production and in waste handling, while materials use is only
significantly affected at penalty levels abov 100 ECU/t CO2.
Introduction
Greenhouse gas (GHG) emission reduction is a key issue for environmental policies in the first half of the 21st
century. The countries of the European Union, the United States and Japan agreed at the UNFCCC conference
in Kyoto in December 1997 to reduce their emissions by 8, 7, and 6% in the period 2008-2012, respectively,
compared to their emissions for a reference year 1990 or 1995. Further emission reductions can be expected
beyond this period.
The Kyoto agreement covers six categories of greenhouse gases: carbon dioxide (CO2), methane (CH4), nitrous
oxide (N2O), perfluorocarbons (PFCs), hydrofluorocarbons (HFCs) and sulphurhexafluoride (SF6). These emissions
are aggregated on the basis of their global warming potential (GWP) for a time horizon of 100 years.
A significant part of the GHG emissions can be attributed to the life cycle of materials. As a consequence, the
production and the use of materials will be affected by GHG emission reduction policies. However, GHG emission
reduction from a materials life cycle perspective has received little attention yet, and the impacts on the
materials life cycle are still unclear. This paper discusses strategies to reduce emissions that are related to the
life cycle of materials, and compares these strategies to other emission reduction strategies (see Figure 1). Special
attention is paid to the interaction of emission reduction strategies. “Materials” includes all substances that
are not produced for energy purposes. Food products are excluded from the category materials.
MINING/
CONVERSION
ENERGY SYSTEM
ENERGY
PRODUCTION
ENERGY
PRODUCTION
MATERIALS SYSTEM
MATERIALS
INDUSTRY
MATERIALS
PRODUCTION
CONSUMER
USE
WASTE
MANAGEMENT
ENERGY
RECOVERY
Figure 1 Definition of the energy system and the materials system

This analysis focuses on Western Europe (the 15 countries of the European Union, Iceland, Norway, and Switzerland).
This region has been selected because it poses a relatively closed area from the point of view of material
flows with high related GHG emissions [1] Large transboundary material flows complicate the analysis and
complicate policy making.
Figure 2 shows the contribution of individual categories of greenhouse gases in Western Europe. The areas in
the figure are proportional to the emissions. The total emission is approximately 4259 Mt CO2 equivalents per
year (1990/1995 reference year figures). The emissions that are not covered in this study are also indicated.
This includes the agricultural emissions of CH4 and N2O, SF6 and HFCs. Figure 2 shows that CO2 constitutes
the bulk of the GHG emissions. The CO2 emissions can be split into two sources: emissions that originate from
the combustion of fossil fuels and emissions that originate from the decomposition of limestone. Cement clinker
production and quicklime production are the main sources of the latter type of emission.
The methane emissions can be split into three categories of a similar order of magnitude: deep coal mining,
landfill sites, and agricultural emissions (from manure and from ruminants, not included in the analysis). N2O
emissions can be split into industrial emissions (mainly production of adipic acid and production of nitric acid)
and agricultural emissions (use of nitrogen fertilizers, not included in the analysis). Primary aluminium smelters
are the main source of PFCs.
The figure does not show important net biomass carbon storage in the increasing forest stock, in products, and
in landfill sites. The total storage effect is approximately 300 Mt CO2 per year. This storage is not accounted
for within the framework of the Kyoto agreement.
Figure 2 does not show the relevance of foreign emissions for Western European consumption. For example
methane leakages from Russian gas pipelines, emissions in the production of metals ores and emissions for
tropical timber production are not accounted. The total net GHG emission from these sources for Western
European consumption is approximately 100-150 Mt CO2 equivalents per year. On the other hand, Western
Europe is an important exporter of materials and finished products. For example steel, machinery and scrap is
exported in significant quantities. As a consequence, an emission of 50 Mt CO2 equivalents can be attributed to
foreign consumers.
The following study is based on an “end use” system boundary. All emissions related to the use of products in
Western Europe are included, whether they arise within Western Europe or abroad. All emissions related to the
production of materials in Western Europe that are net exported are excluded from the analysis (net export =
export - import). Net exports of materials within products and net exports of waste materials are not valued in
GHG-emission terms. One one hand, data are scarce, on the other hand the available data suggest that these
flows are less relevant from a GHG emission point of view.

N2O INDUSTRY
OTHER GHG
CH4 LANDFILLS/MINING CH4 AGRICULT.
CO2 CARBONATES
CO2 FOSSIL FUELS
N2O AGRICULT.
= NOT COVERED IN THIS STUDY
Figure 2 Western European GHG emissions (area is proportional to the emission in 1995).
Emissions that are not covered in this study are indicated
Western European bulk material flows are quantified in [2. The related CO2 emissions have been derived from
data for energy balances and process emissions [1] The materials system represents a CO2 emission of approximately
744 Mt CO2 per year, and additionally an emission of non-CO2 GHGs of 254 Mt CO2 equivalents
per year (excluding carbon storage for biogenous materials). The production of less than 20 materials, characterised
by their uniform production process, represents more than 75% of the greenhouse gas emissions for
materials that are consumed within Western Europe. This paper focuses on the reduction of GHG emissions
through changes in production, use, and waste handling of these materials. The following questions will be adressed:
• What are the material flows in Western European where policies can have a significant greenhouse gas
emission reduction impact ?
• Which emission reduction options exist ?
• Is it necessary to consider interactions of improvement options in the analysis ?
• What are the dynamics of the materials system; is long-term planning possible and sensible ?
• Must technological progress be considered in the analysis ?
• To what extent can materials options reduce greenhouse gas emissions ?
• How compare the costs of these materials related emission reduction options to the costs of energy related
emission reduction options ?
• Which materials strategies should be further developed (and which strategies not) ?
Characteristics of the system from a GHG emission point of view
A limited number of materials constitutes the bulk of the GHG emissions. Table 1 provides an overview of the
most relevant categories.

Table 1 The relevance of materials for GHG emissions, end use based system boundaries, based on actual
emissions according to IPCC accounting guidelines
CO2 Non-CO2 GHG Total Fraction
[Mt CO2 equiv.pa] [Mt CO2 equiv.pa] [Mt CO2 equiv.pa] [%]
Metals 244 11 255 26
Synthetic organic
materials
167 53 220 22
Natural organic materials 93 130 223 22
Inorganic materials 49 60 109 11
Ceramic materials 191 - 191 19
Total 744 254 998 100
The flows of energy and materials through the economy can be analysed in a systems approach. “A system is a
structured assemblage of elements and subsystems, which interact through interfaces. The interaction occurs
between system elements and between the system and its environment. The element and their interactions constitute
a total system, which satisfies functional, operational and physical characteristics, as defined by the user
and customer needs and requirements, over a defined total system life cycle of the system in existence, including
the life cycle of bringing the system into being” [3]
The system elements in the economy are called processes. They are defined by their energy and material inputs
and outputs, by their emissions and by their costs. Examples are a blast furnace (input coal and coke, output
liquid iron), a car manufacturing plant (input steel sheet, other materials and energy, output cars), landfill sites
(input waste materials). The interfaces between the processes are called flows. They include energy flows and
material flows. The function of the system is to satisfy a demand from the final users (e.g. a demand for housing,
transportation, lighting and packaging). Constraints are for example added for the availability of resources,
for the social acceptance of processes (like nuclear power plants or public transportation) and for the environmental
impact of the system (like GHG emissions).
How long is the life cycle of the system ? The definition encompasses all steps from the early system conception
to its final abolishment. For the economy as a whole, it is difficult to set these boundaries. From a systems engineering
point of view for GHG emission reduction, it is more relevant to analyze the system dynamics: how
much time is required to achieve significant changes in the system configuration in order to change its environmental
impact.
The dynamics of the energy and materials system are defined by the rate of technological change, the turn-over
rate of production processes, and the product life. The rate of technological change is often overestimated.
Technological breakthroughs require decades, even if the development is successful. Many key processes in the
energy and materials system show little technological change over the last century. Examples are blast furnaces,
Hall-Heroult smelters for aluminium production, pulp mills, ignition engines for e.g. passenger cars.
Gradual efficiency improvements have occured, but the technical basic principle has not changed. Development
of alternatives was not succesful because of technological problems or because of worse process economics.
Even if technology development is successful, the actual introduction is hampered by the slow diffusion processes
for new technology. Literature data suggest the the time required for a new process to unfold from 10% to
90% of the market share ranges from one decade to several centuries, depending on the pervasiveness of the
technology and the interaction with existing practices and technologies [4]. Diffusion of new technologies is
also hampered by the life span of existing capital equipment. The life span of industrial equipment ranges from
1 to 5 decades. The life span of infrastructure ranges from 5 to 10 decades.
Data for materials production suggest similar slow system dynamics. Figure 3 shows the materials production
trends for a number of “old” materials (steel, cement) and “new” materials (polyethylene, aluminium). It is interesting
to see the strong growth of paper production, despite the fact that this is the material with the longest

history of use. The figure indicates that the introduction of new materials can also be considered as a slow diffusion
process.
INDEX [-]
120
100
80
60
40
20
0
1960 1965 1970 1975 1980 1985 1990 1995
[YEAR]
STEEL
CEMENT
ALUMINIUM
PAPER/BOARD
Figure 3 Western European materials production, period 1960-1995 (1990=100)
An additional barrier for materials substitution is the relatively closed life cycle of many materials (contrary to
the energy carrier life cycle). Recycling of waste materials beyond the product life limits the rate of change in
the materials system. In conclusion, the slow dynamics of the materials system makes long term planning
feasible and sensible.
General model description
The MARKAL linear programming model was developed 20 years ago within the international IEA/ETSAP
framework (International Energy Agency/Energy Technology Systems Analysis Programme). More than 50
institutes in 27 countries use nowadays MARKAL [5,6] MARKAL is an acronym for MARKet ALlocation.
The model was originally developed for energy systems analysis. In recent years, the model has been extended
for materials system analysis [7] The model covers now the whole materials life cycle ‘from cradle to grave’.
Figure 4 shows the materials system model structure.
1 PRIMARY
PRODUCTION
MATERIAL
3 PRODUCT
ASSEMBLY
PRODUCT
4 PRODUCT
USE
WASTE PRODUCT
5 REMOVAL
&SEPARATION
2 RECYCLING
6 ENERGY
RECOVERY
Figure 4 Materials system model structure
WASTE
MATERIAL
7 DISPOSAL
PE

MARKAL represents all activities that are necessary to provide a fixed set of products and services. Products
and services can be generated through a number of alternative (sets of) processes. The model contains a database
of several hundred processes, covering the whole life cycle for both energy and materials. The model calculates
the least-cost system configuration.
The database of processes and the constraints for individual processes and for the whole region are defined by
the model user. Constraints are determined by the demand for products and services, the maximum introduction
rate of new processes, the availability of resources, environmental policy goals for energy use and for emissions
etcetera. All environmental impacts are endogenised into the process costs and into the costs of energy
and matter flows between processes.
MARKAL is a dynamic model. The time span to be modeled is divided into seven periods of equal length,
covering the period 1990-2050. Changing technology is modeled through changing parameters in time for
individual processes. The future availability of new alternative processes is modeled separately. The model is
used to calculate the least-cost system configuration for the whole time period, meeting exogenously defined
product and service demands and meeting emission reduction targets. This optimization is based on a so-called
‘perfect foresight’ approach, where all time periods are simultaneously optimized. Future constraints are taken
into account in current investment decisions.
The model contains approximately 50 energy carriers, 125 materials, 100 product service categories and 32
types of waste materials. The whole system consists of several hundred processes. A detailed description of the
model input parameters for processes, materials and products can be found in separate reports.
[8,9,10,11,12,13,14,15,16,17]
Different GHG emission penalties have been analyzed. These penalties are shown in Figure 5. The base case is
run without penalties. In the emission reduction cases, the penalties increase from zero in the year 2000 to their
maximum level in 2020 and stabilize afterwards.
[ECU/T CO2]
200
150
100
50
"200 ECU/T CO2"
"100 ECU/T CO2"
"50 ECU/T CO2"
"20 ECU/T CO2"
0
1990 2000 2010 2020
[YEAR]
2030 2040 2050
Figure 5 GHG emission penalty scenarios

Emission mitigation options
The following improvement options are considered within the materials life cycle:
1. Increased energy efficiency: new production technology
2. Increased materials efficiency: increased materials quality
3. New recycling technologies
4. Cascading: waste separation and product re-use
5. New energy recovery technologies for waste materials
6. Substitution of energy carriers for materials production
7. Substitution of natural resources/feedstocks
8. Substitution of materials/product re-design
9. End-of-pipe technology for catalytic and thermal conversion of CH4 (from landfill sites)and N2O (from
nitric acid and adipic acid production) and removal and underground storage for CO2 (for iron, cement
clinker and ammonia production)
The list covers all stages of the materials life cycle “from cradle to grave”. The list covers strategies that are
currently not applied because of technological problems or because they are not cost-effective. R&D can overcome
the technological barriers; the cost-effectiveness of alternatives may change in the future if GHG emissions
are penalized, or if environmentally friendly alternatives are subsidized.
Improvement strategies that affect the consumer lifestyle or that affect the product performance have not been
considered. The problems regarding their implementation is beyond the scope of techno-economic optimization
and cannot be analyzed fruitfully with a techno-economic optimization model. In order to show the impact of
lifestyle and economic growth, a number of demand scenarios have been analyzed.
The reference scenario in this paper is based on a 5% discount rate, moderate economic growth, and moderate
estimates for demand growth. Key scenario parameters for demand growth are shown in Table 2.
Table 2 Demand scenario parameters
1990 2010 2040
Passenger cars 100 128 176
Trucks 100 124 164
Single family residences 100 115 130
Residential other electricity 100 233 378
Commercial electricity 100 133 178
Results
In order to analyze the contribution of the materials system for the total GHG emissions, two base case runs are
presented. One includes the demand for materials (E+M), the other excluded the materials demand categories
(E). The difference represents the contribution of the materials system to the GHG emissions (somewhat underestimated,
because biogenous carbon storage in products and disposal sites is subtracted from the materials
system emissions). This is shown in Figure 6. The figure shows that the calculated emissions are in line with
the bottom-up estimates regarding the relevance of the materials system. Figure 6 shows that both emissions in
the materials system and emissions in the energy system increase in time.

[MT CO2 EQUIV./YEAR]
5,000
4,000
3,000
2,000
1,000
NOT COVERED
0
1990 2000 2010 2020
[YEAR]
2030 2040 2050
Figure 6 Base case GHG emissions for the energy system and the integrated energy and materials system
(E+M)
The emission reduction for the integrated energy and materials system (E+M) is shown in Figure 7. The figure
shows that a reduction of the total GHG emissions from 4500 Mt to less than 1000 Mt in 2030 is achieved if a
(high) penalty of 500 ECU/t CO2 equivalent is introduced. This is a reduction of more than 75%. This result
shows that a factor 4 emission reduction is possible, even in a situation with growing demand. The emission
penalty may seem unlikely high, but the factor 4 emission reduction is an ultimate goal for industrialized countries
in order to minimize the risk for major climate change.
Note the high contribution of CO2 to the total GHG emissions in the base case. Autonomous developments reduce
emissions of other GHGs (e.g. the closure of the German and UK coal mines and the reduction of waste
disposal).
[MT CO2 EQUIV./YEAR]
5,000
4,000
3,000
2,000
1,000
0
BC
50 ECU/T CO2 200 ECU/T CO2
20 ECU/T CO2 100 ECU/T CO2 500 ECU/T CO2
Figure 7 Changing GHG emissions with increasing emission penalties, 2030
IMPORTS
PFC
N2O
CH4
CO2
E+M
E

The contribution of the materials system to the emission reduction is shown in Figure 8. This result is based on
a comparison of a set of calculations without materials (E) and a set of calculations with materials demand
(E+M). The figure shows a very significant contribution for the materials system, up to 50% of the total emission
reduction at lower penalty levels.
EMISSION REDUCTION [MT CO2 EQUIV.]
3500
3000
2500
2000
1500
1000
500
E+M
E
0
0 50 100
PENALTY [ECU/T CO2]
150 200
Figure 8 Aggregated emission reduction for the energy system (E) and the integrated energy and materials
system (E+M)
The contribution of individual materials strategies is elaborated in Figure 9. Because of the definition of the
materials system (see Figure 1), some emission reduction can be attributed to emission reduction in electricity
production, fuel switches and increased industrial energy efficiency (e.g. for electricity that is used for materials
production). End-of-pipe technology (for CO2 removal, industrial N2O conversion, CH4 capture from landfill
sites and reduced landfilling) proves to be significant. The contribution of biomass feedstocks for the petrochemical
industry is also significant. Charcoal is introduced for injection in blast furnaces. Some materials substitution
occurs. On the waste management side, plastics incineration is replaced by hydrogenation and natural
organic materials are used for energy recovery. Improved materials quality is modeled for concrete (high
strength concrete) and for steel. Both options are introduced.

[MT CO2 EQUIV./YEAR]
1400
1200
1000
800
600
400
200
0
BC
50 ECU/T 200 ECU/T
20 ECU/T 100 ECU/T
OTHER
WASTE
MANAGEMENT
MATERIALS
EFFICIENCY
MATERIALS
SUBSTITUTION
OTHER RESOURCES
SUBSTITUTION
PETROCHEMICAL FEEDSTOCK
SUBSTITUTION
END-OF-PIPE
Figure 9 Contribution of individual materials strategies, 2030, increasing emission penalties
The impact of GHG emission reduction on individual materials use and production
The changing materials consumption due to GHG emission reduction is shown in Figure 10. The figure shows
the impact of a penalty of 200 ECU/t CO2. The consumption of cement decreases, while the consumption of
wood and of aluminium increases. This change can be attributed to materials substitution in the transportation
sector and materials substitution in the building sector.
SHIFT [%]
60
40
20
0
-20
-40
2000 2010 2020 2030 2040 2050
[YEAR]
CEMENT STEEL PE PAPER SAWN WOOD ALUMINIUM
Figure 10 Shifts in materials consumption due to GHG emission reduction, 200 ECU/t
CO2 reduction case
The impact of GHG emission penalties on the materials production of aluminium, cement, and ethylene is
shown in Figures 11-13. The figures show a significant switch of production technology and of production location
(for aluminium, based on hydroelectricity based production in Iceland).

[MT AL/YEAR]
25
20
15
10
5
0
BC
50 ECU/T CO2 200 ECU/T CO2
20 ECU/T CO2 100 ECU/T CO2 500 ECU/T CO2
RECYCLING
Figure 11 Changing aluminium production with increasing emission penalties
[MT CEMENT/YEAR]
200
150
100
50
0
BC
50 ECU/T CO2 200 ECU/T CO2
20 ECU/T CO2 100 ECU/T CO2 500 ECU/T CO2
Figure 12 Changing cement production with increasing emission penalties
INERT ANODES
ICELAND
HALL-HEROULT
ICELAND
INERT ANODES
CONTINENTAL
HALL-HEROULT
CONTINENTAL
HIGH STRENGTH
POZZOLANS
GEOPOLYMERIC
BF SLAG
ACTIVATED FLY ASH
FLY ASH
PORTLAND

[MT ETHYLENE/YEAR]
25
20
15
10
5
0
BC
50 ECU/T CO2 200 ECU/T CO2
20 ECU/T CO2 100 ECU/T CO2 500 ECU/T CO2
Figure 13 Changing ethylene production with increasing emission penalties
PLASTIC WASTE
PYROLYSIS
WOOD
FLASH PYROLYSIS
METHANOL
MTO
NATURAL GAS
OX. COUPLING
LPG
CRACKING
ETHANE
CRACKING
GASOIL
CRACKING
NAPHTHA
CRACKING
The results of these calculations do not significantly differ from the results for earlier calculations for the Netherlands
[7]. The contribution of the materials system proves to be significant in both case studies. The results
indicate positive impacts on the use of wood and aluminium, and negative impacts for cement. Major impacts
can be expected for materials production and waste handling processes.
Conclusions
The conclusion will focus on the questions that were posed in the introduction.
What are the material flows in the Western European economy where policies can have a significant
greenhouse gas emission reduction impact ?
The production of iron and steel, cement and quicklime, petrochemical products, and a limited number of inorganic
materials like nitrogen fertilizers constitutes the major part of the GHG emissions that can be attributed
to the production of materials. Carbon storage in products proves to be important for both synthetic organic and
natural organic materials. The organic materials are also increasingly important from an energy recovery point
of view.
Which emission reduction options exist in the materials life cycle ?
Emission reduction can be achieved in all stages of the materials life cycle. They can be characterised as efficiency
options, recycling/reuse options (including energy recovery), and substitution options. Changing lifestyle
or tempered consumption growth are not considered in this analysis.
Is it necessary to consider interactions in the materials system in the analysis ?
The results show that it is very important to consider interactions between improvement options in the optimization.
Especially the reduction of emissions in materials production affects the cost-effectiveness of further on
in the product life cycle. If such interactions are not considered, emission reduction potentials are overestimated.

What are the dynamics of the materials system; is long-term planning possible and sensible ?
The dynamics of the matarials system are determined by the pace of technological change and the life span of
capital equipment. Regarding technological change, development of new technology for bulk materials requires
a period of several decades from labscale to a significant section of the market. As a consequence of such slow
system dynamics, long term planning is possible.
Should technological development be considered in the analysis ?
Technological development is one of the key driving forces that determine the system configuration. Within a
time horizon of half a century, technology will change significantly. In the framework of the energy and materials
system, major changes can be expected in power generation (gasification technology, new technology for
renewables), in the transportation sector (fuel cells, electric vehicles, ethanol), for materials production
(smelting processes for iron production, near net shape casting, biomaterials). Regarding product use, technological
development seems to be a less important driving force. For recycling, cheap materials separation technology
and improved recycling technology for plastics pose a challenge.
How interact the improvement options in the materials system ?
Significant emission reduction in materials production and electricity production reduces the cost-effectiveness
of emission reduction in materials use.
To what extent can materials options reduce greenhouse gas emissions ?
The materials system can contribute up to 1200 Mt CO2-equivalents of emission reduction.
How compare the costs of these materials related emission reduction options to the costs of energy related
emission reduction options ?
Emission reduction in the materials system contributes up to 50% to the total emission reduction for a fixed
emission penalty. This figure shows that the costs are in line with the costs for emission reduction options in
the energy system.
Which strategies should be further developed (and which strategies not) ?
Greenhouse gas emission penalties will have much more impact on materials production and waste handling
than on materials consumption.
End-of-pipe technology does already receive a lot of attention. More attention should be paid to feedstock
substitution in the petrochemical industry, to improved materials quality, materials substitution/product redesign,
and to the GHG consequences of future waste management technologies. Carbon storage in products
seems a less relevant strategy at acceptable emission penalty levels. The potential for increased recycling seems
to be limited.
It is possible to achieve a factor four reduction in the greenhouse gas emissions in the materials system through
an improved materials efficiency of the economy without affecting the lifestyle. A factor 10 is not possible with
the improvement options and demand growth paths that have been considered in this study. Product re-design,
increased product life and improved materials quality are examples of promising options that are not yet fully
included in the calculations. More attention for this type of improvement strategies is warranted.
References

[1] Gielen: Western European Materials as Sources and Sinks of CO2: a Materials Flow Analysis Perspective.
Journal of Industrial Ecology, Forthcoming.
[2] R. van Duin: Production and Consumption of Materials in Western Europe in Year 2000. Bureau B&G,
Emst, the Netherlands, July 1997
[3] O.A. Asbjornsen: Systems Engineering Principles ans Practices. Skarpodd, Armold (MA) USA, 1992
[4] A. Grübler: Time for a Change: On the Paths of Diffusion and Innovation. In J. Ausubel, H.D. Langford:
Technological Trajectories and the Human Environment. p. 14-32. National Academy Press, Washington
DC, 1997
[5] T. Kram: National energy options for reducing CO2 emissions, Volume 2: Country Studies. ECN-C--94-
024. Petten, March 1994
[6] T. Kram, D. Hill: A multinational model for CO2 reduction. Defining boundaries of future CO2 emissions
in nine countries. Energy Policy vol. 24, no. 1, pp. 39-51, 1996
[7] D.J. Gielen: Toward integrated energy and materials policies? A case study on CO2 reduction in the
Netherlands. Energy Policy vol. 23, no.12, pp. 1049-1062, 1995
[8] J.R. Ybema et al: Scenarios for Western Europe on Long Term Abatement of CO2 emissions. ECN-C--97-
051. Petten, December 1997
[9] P. Lako, J.R. Ybema: CO2 abatement in Western European power generation.
ECN-C--97-053. Petten, May, 1997
[10] I. Kok, J.R. Ybema: CO2 abatement in the built environment of Western Europe. ECN, Petten,
forthcoming
[11] D.J. Gielen: Technology characterisation for ceramic and inorganic materials. Input data of Western
European MARKAL. ECN-C--97-065. Petten, September 1997
[12] D.J. Gielen: Building materials and CO2 Western European emssion reduction strategies. ECN-C--97-65.
Petten, October 1997
[13] M. Hekkert, E. Worrell: Technology characterization for Natural organic Materials. Input Data for
Western European MARKAL. NW&S, Utrecht, forthcoming.
[14] B.W. Daniels, H.C. Moll: The base metal industry: Technological description of processes and production
routes, status quo and prospects. IVEM, Groningen,
[15] M.E. Bouwman, H. Moll: Status quo and expectations concerning the material composition of road
vehicles and consequences for energy use. IVEM, Groningen, forthcoming
[16] M.P. Hekkert, L.A.J. Joosten, E. Worrell: Packaging Tomorrow. Modeling the material Input for
European Packaging in the 21 st Century. NW&S, Utrecht, forthcoming
[17] L.A.J. Joosten: MATTER datasheets synthetic organic materials. NW&S, Utrecht, forthcoming

Material efficiency improvement for European packaging in the period
2000 - 2020
Marko Hekkert, Louis Joosten and Ernst Worrell
Utrecht University, Department of Science, Technology and Society,
Padualaan 14, NL-3584 CH Utrecht, The Netherlands
Abstract
In this paper the current material consumption for packaging making in Europe is described. Per packaging
type (food bottles, non-food bottles, boxes for primary packaging, flexible packaging, carrier bags, industrial
boxes and pallets) options for improved material efficiency are described. The options are in the field of using
thinner materials, using less material by changing the shape of the package, using recycled material and using
refillable packages. This paper shows that many option are available to reduce the future material input for
packaging and that a reduction of CO2 emissions by this sector with a factor 2 is possible. A substantial share of
this reduction can be achieved without any changes in consumer behavior.
Introduction
Packaging is an important product category from a CO2 point of view. In Table 1.1 a first order estimate for the
energy and CO2 balance for packaging materials are stated. The total direct and indirect CO2 emission of 109
Mtonne for all packaging materials can be compared to the current Western European emission of approximately
3300 Mtonne [Gielen, 1997]. The packaging materials that are analyzed in this report represent 3.3 percent
of the total Western European CO2 emission or about 10% of European industrial CO2 emissions. Plastics,
paper and metal are the most important contributors to the CO2 emissions from packaging.
This paper describes possible material efficiency options by the packaging industry and the material processing
industry that may lead to reduced CO2 emissions in Europe. Material efficiency is defined in this paper as the
amount of primary material that is needed to fulfill a specific function. Material efficiency improvement allows
the same function to be fulfilled with less material. At several stages in the life cycle intervention is possible in
order to increase the material efficiency over life cycle. Figure 1.1 shows these efficiency improvement measures
and at which stages in the life cycle they intervene. The standard life-cycle is presented within the box.
The efficiency improvement measures that can be applied are depicted outside the box.
Even though in this paper we focus on efficiency improvement of packaging materials we need to take into account
that packaging is needed for the protection and marketing of the packaged products. In many cases this
paper states the technical possibilities but before actual implementation a good understanding of the purpose of
packaging material is needed. Savings on packaging material may in some cases lead to less efficient product
use which may even lead to greater consumption of energy per consumption unit than the original situation.
However, arguments like these should not prevent that resources are used in the most efficient way possible.
This paper investigates which packaging options are suitable to reach an efficiency improvement with a factor 2
in 2020 and also if a factor 10 is possible on the long run (2050).

Figure 1.1. The life-cycle of materials, adapted from Worrell et al., (1995).
Life cycle
Resources
Material
production
Manufacturing
Application/
filling
Consumption
Waste
processing
Waste
Efficiency Improvement Options
Thinner materials
Product
design
Product reuse
Material
substitution
Good housekeeping
Material
recycling
Cascading
In order to model possible changes in material consumption we discerned 7 packaging categories: food bottles,
non-food bottles, boxes for primary packaging, flexible packaging, carrier bags, industrial boxes and pallets.
For these categories we studied the present material input and the possible material options.
In this paper we will first describe the material efficiency improvement options for the different packaging
categories. Next we will describe how we calculated the costs and the effect on CO2 emissions for the material
efficiency options. Finally we will present the CO2 emissions and costs for all material efficiency improvement
options. These calculations are first order estimates. More precise calculations require detailed insight in the
material production processes of the packaging materials. This will be done by means of a MARKAL model
that focuses on CO2 emissions reductions that are related to the Western European materials system. The options
described in this paper will be use in this model as input data. The first order estimates are given in order
to give insight in the question: is it possible to reduce CO2 emissions related to the use of packaging by a factor
2 or a factor 10 by means of material efficiency improvement.
To give an idea about which packaging materials are used for which products, Table 1.2 states the material input
of packaging material per demand category.

Table 1.1 First order estimate of energy and CO2 balance of packaging in Europe based on van Heijningen
(1992), van Heijningen (1992a), APME (1996), de Beer et al., (1994) and van Duin (1997) 12
Consumption of Packaging materials Consumption CO2
Share in total
(Mtonne/yr) (Mtonne/yr) CO2 emission (%)
Paper 28 14 13 - 17
Glass 17 6 6 - 7
Plastics 12 23 - 61 28 - 56
Metal 6 15 14 - 18
Others (incl. Wood) 13 13 11 - 12
Total 75 84 - 109 100
Table 1.2 Material input per packaging service category
Demand Material Material Demand Material Material
category input type category input type
[ktonnes] [ktonnes]
beverages, 252 Steel non-food 798 Cardboard
carbonated 171 Aluminum 672 PVC
1131 PET 260 PE
15605 Glass 140 PP
199 PE 10 Paper
745 Board
447 PS carrier bags 430 PE
73 PP
industrial bags 600 PE
wet food 800 Glass 504 Paper
188 Steel 400 PP
non-food liquids 1050 PE transport 2610 PE
Packaging 259 Wood
dry food 2577 Cardboard 9400 Corrugated board
224 PVC
840 PE Pallets 4956 Wood
1530 PP 840 PE
47 Paper
110 PET
2. Material efficiency options for different packaging categories
2.1 Food bottles
The category food bottles is defined as all bottles, cans and jars that are used to pack food. A very large category
is liquid packaging. The main liquid types that are packed within this category are dairy products, soft
drinks, beer, wine and spirits. The total European consumption of these products amounted to 111 billion liters
[EC, 1997]. Also preserved fruit and vegetables are important in this category, 4 billion liters in 1994 [EC,
1997]. Current most used materials to pack the products in this category are glass, PET, liquid board, steel and
aluminum
Glass
Glass bottles and jars are used to pack all of the liquid and food categories that are defined earlier. The glass
bottles have had competition from other materials since a long time. Years ago, milk in The Netherlands was
sold in glass bottles but today the major packaging used are liquid cartons [van der Ent, 1995]. In the soft-drink
sector, a similar situation occurred. For a long time only glass was used while today PET bottles have taken a

very large market share in Europe. Still, glass has a large market share in the packaging used in Europe. About
23% of the packaging materials is glass which corresponds to 17250 ktonne in 1994 [APME, 1996]. In 1995
54% of glass packaging in Europe was collected for recycling [Anon. 1996].
For the MARKAL- model we will define 2 glass bottles (large and small) and 1 jar. The large bottle is defined
as a bottle with a volume of 1 liter while the small bottle has a volume of 0.3 liter. The jar has a volume of 0.5
liter (in between the often used standards of 72 and 37 centiliter). Based on SVM (1993a) we estimate the
weight of these bottles and jars at 500, 250 and 250 grams respectively.
The glass bottles can be improved in order to reduce the amount of packaging waste. In the Netherlands many
projects have taken place to reduce the weight of glass bottles. The weight of milk bottles was reduced with
33%, the weight of several liquor bottles was reduced with 20% and 22% [SVM, 1994]. It seems possible to
reduce the weight of large glass bottles in Europe with 25% in 2020. Projects have taken place to reduce the
weight of small glass bottles like beer bottles with 5.5% [SVM, 1994]. The glass industry in The Netherlands
expected in 1993 a weight reduction of 15% in 1995 compared to 1991. We will use this figure for the improvement
of the small glass bottle [SVM, 1993]. The weight of mushroom jars was reduced with 20% by a
company in 1994 and in 1992 vegetable jars were reduced in weight with 14%. Furthermore, some jam and
jelly bottles were reduced in weight by 10% in 1995 [SVM, 1992, 1994, 1995].
Besides weight reduction a lot of resources can be saved by glass recycling. Two types of recycling are possible:
product re-use and material recycling. Currently, the European recycling rate is already 50%. The Swiss recycling
rate is the highest in Europe (85%) and can be seen as the absolute maximum for Europe. However, due
to the large transportation distances in Europe in rural area’s this figure is not very likely to be reached. We
assume a maximum recycling rate in Europe of 70%. We furthermore assume that color separation is possible
for all the glass that is collected separately; in the U.K. already 95% of the recycled glass is sorted on color. In
The Netherlands beer bottles and some jar types are recycled with a deposit system (product recycling). We assume
that this system is also an option for Europe after 2000. . The success of such a system depends on the
willingness of the consumers to return the package (this can be influenced by the height of the deposit fee) and
the willingness of the producers to implement such a system. Standardization of packaging is a strong tool to
make product recycling work. In this way it doesn't matter if the package is returned to producer A or producer
B. Standardization for beer bottles is proven technology in The Netherlands. We will therefore only use this
option for beer bottles in Europe. We assume a trip number of 20 trips per bottle
PET bottles
PET (Poly Ethylene Terephthalate) bottles were introduced in the soft drink sector to replace the standard 1 liter
glass bottles. PET bottles are especially suited to pack carbonated soft drinks. PET bottles also replace PVC
bottles that are often used in South Europe for the packaging of mineral water [Ent, 1995]. 50% of the PET
packaging in Europe are used to pack soft drinks, 27% is used to pack mineral water, and 5% is used to pack
other drinking liquids. The rest (18%) is used for other purposes like food and non food packaging [Clausse
and Mitchell, 1996].
In 1993 about 700 ktonne PET for bottles was used in Europe and projections for 2000 and 2005 suggest a demand
for PET of 1.55 million tonne and 2.12 million tonnes respectively [Anon., 1995, Clausse and Mitchell,
1996]. These trends are based on expectations that PET bottles will replace all PVC and a lot of glass packaging.
Most PET bottles used in Europe are one way PET bottles. We will model these bottles as having a volume of
1.5 liter and a weight of 50 grams. In The Netherlands and germany many PET bottles used are refillable. This
development was possible because new PET types became available that could be cleaned at higher temperatures
(58°C). In 1994 Spadel introduced the Hotwash Pet bottle that can even be cleaned at temperatures up to
75°C [Hentzepeter, 1996]. The refillable PET bottles (REF-PET) are designed to make 25 trips during a lifetime
of 4 years [Kort, 1996]. Many bottles, however, make less trips because of the damage done to the bottles
during the refill process (scuffing). We will model the refillable PET bottle as having a volume of 1.5 liter, a
weight of 103 grams and a trip number of 20.

PET bottles normally are made out of virgin PET but the three layer PET bottle with a recycled PET inner layer
can be seen as an improvement option. We will use the bottle with 25% recycled PET as improvement option
for the virgin bottle.
Liquid board
Cartonboard has been used to pack liquids for a long time. The Tetra Classic was introduced as early as in 1952
[PPI, 1996]. The most important markets for liquid cartonboard are milk and juice packaging. Less important
are wine, water, and soup packaging [PPI, 1996].
In order to hold liquids liquid board is laminated with other materials like PE and aluminum. Tetra Briks for
juice packaging for example contain 75% cardboard, 20% PE and 5% aluminum and the total weight is 28
grams for a 1 liter package [Buelens, 1997]. Cardboard is used as middle layer with a PE and aluminum layer
on the inside and a PE outer layer. The liquid board package is not expected to undergo radical changes in future
years. To keep up the competition strength more plastics may be used for easier openings and better closures.
Material savings are reported by increasing the size of the 1 liter package to 1.5 liter. This saved 9%
packaging material per liter [SVM, 1994].
Metal packaging
In 1995 around 4 million tonnes of packaging steel was used in Europe. Furthermore another 3.1 million tonnes
of aluminum was used. For the beverage market 14.3 billion steel cans were used in 1995. The consumption
of aluminum cans in the beverage sector was even larger (17.5 billion cans) [Depijpere, 1996].
In Europe there is a strong competition between steel and aluminum for beverage cans. Almost all lids of European
beverage cans are made out of aluminum while 50% of the bodies of the cans are made out of steel and
another 50% out of aluminum [Depijpere, 1996]. In contrast in the U.S. almost all cans that are used in the
beverage industry are made out of aluminum (95%) [Meert, 1995]. For food cans the situation in Europe is entirely
different. Tin-plated steel commands 100 percent of the food can market [Abbott, 1995]. The size of the
“steel” food market is estimated at 2000 ktonnes [Meert, 1995, Depijpere, 1996].
Many developments have been going on in the last decades to reduce the weight of steel beverage cans in order
to save materials costs. In the last decade the weight of steel beverage cans have been reduced by 20%
[Depijpere, 1996]. The current body of a steel 33 ml can weighs about 27 grams. It is already possible to produce
a steel can that weighs 23 grams. Hoogovens is working on ultra thin steel that should make it possible to
produce cans that weigh 18 grams in 2000 [van der Ent, 1995, Depijpere, 1996, van Deijck, 1994].
A new development in the steel can business is the introduction of the all steel can. The can is developed by
Hoogovens, British Steel and Rasselstein. The difference with the normal steel beverage can is the steel ‘push
in’ lid. The advantage of the all steel can is that the can be recycled entirely. Aluminum that normally is part of
the can, can not be recycled because it is incinerated in the recycling process [van Deijck, 1994]. The lid of the
all steel can weighs about 8 grams. The total weight of the all steel can in 2000 will be around 26 grams.
The developments in aluminum beverage cans are very similar to the developments in steel cans. Producers
have also been working on reducing the weight of the cans. The current body of an aluminum can weighs
around 11.5 grams. The current lid weighs another 2.7 grams which leads to a total can weight of around 14
grams [Depijpere, 1996, van der Ent, 1995]. Alcan, a large aluminum producer, estimates that an aluminum
can in 2000 can weigh 13 grams (including lid) [Goddard, 1994].
Besides developments in the beverage sector also food cans are made lighter. A liter can weighs around 88
grams in stead of 115 grams in 1985 (savings of 35%) [van Stijn, 1996]. Continental Can is currently working
on a ‘honeycomb can’. This can has a honeycomb structure which makes the can stronger. This structure
makes it possible to produce a can that weighs 30% less [van Stijn, 1996]. We expect the market penetration to

e lower then 100% because labels can not be attached as easily and the printability is worse than for normal
cans.
PS and PP cups
Cups made from thermoformed plastics are used for packing of yogurt and butter. Popular materials are PS, PP
and PVC with market shares of respectively 52, 15 and 25% [APME, 1996]. We will model two cups. One is
made from PS and has a weight of 14 grams and capable of packing 500 gram product. The other cup is made
from PP and weighs 12 gram [Phylipsen, 1993].
PE pouch and PC bottle
Both Tetra Pak and Elopak have introduced flexible packaging (pouches) for milk and juice. The pouches are
made out of plastic. Tetrapak uses LLDPE and Elopak uses multiple layer PP laminates [Couwenhoven, 1996].
The advantage of using pouches for liquid packaging is that they are extremely light. An empty 1 liter pouch
from Elopack weighs 10 grams while an empty 1 liter pouch from Tetra Pak only weighs 4 grams. We will
model the Tetra Pack pouch as improvement option for current packaging options. It needs to be taken into account
that this option asks for a change in consumer behavior because a pouch does not have the same packaging
characteristics as non-flexible packaging.
In the Netherlands the PC bottle is introduced as an alternative for liquid board milk packaging. We will model
this bottle (trip number of 30 and a weight of 80 grams for a liter bottle) as alternative for current packaging
types for non carbonated drinks.
2.2 Non-food bottles
Non-food bottles are used to pack shampoos, detergents and other cleaning liquids, lubricants, and light cleaning
chemicals. Contrary to food bottles not as many different materials are used to pack the liquids. The food
sector puts high demands on the quality of the material because properties like CO2 and oxygen permeability
are very important for the durability and quality of the packed product. In the non-food sector these factors are
less important.
The main material used to produce non-food bottles is HDPE. We estimate the amount of non-food bottles on
the HDPE blow moulding data for Europe. In 1994 about 1125 ktonne HDPE bottles were used in Europe
[APME, 1996]. Besides non-food bottles also the U.K. milk bottles are made out of HDPE. This leaves 1050
ktonne HDPE for non-food bottles.
For the MARKAL model we define a non-food bottle with a volume of 0.5 liter and a weight of 50 grams. Several
improvement options can be modeled. Based on SVM (1992-1996) we estimate that it is possible to reduce
the weight with 25%. Furthermore bottles can be made with a recycled HDPE content of 50% [SVM, 1994].
Two other options are more difficult to implement: a refill package made out of cardboard (14 grams) and a
plastic pouch (5 grams).
2.3 Boxes for primary packaging
The category ‘boxes (primary packaging)’ consists of all the non-flexible packaging that is used as primary
packaging to pack food and non-food products. To estimate the amount of boxes used as primary packaging in
Europe we make use of statistics on the materials that are used to make the boxes. In PPI (1997) the total
amount of packaging board consumption in Western Europe is estimated at 4.8 Mtonnes. Subtracting the share
for liquid packaging (745 ktonne) leaves 4 Mtonnes of board.
Besides cardboard, also plastic is a popular material for the production of boxes. In general plastic boxes have
very thin walls and are not very durable. We estimate that the majority of these boxes are made out of thermoformed
sheets. In Europe the amount of thermoformed sheets that are used as packaging is 1220 ktonnes in
1994 [APME, 1996].

To model the category boxes we will define two basic types of boxes: the cardboard box and the plastic box. We
will also model a cardboard box with an inner bag. The standard cardboard box has a volume of 1 liter and
weighs 35 grams. The inner bag will add another 3.0 grams (PE) or 7 grams (paper). A market share of 20% of
the cardboard box market is assumed for the box with the inner bag. The plastic box will contain some cardboard
in order to simulate the share of blister packaging. We will define a plastic box with a volume of 0.5 liter
that contains 10 grams of plastics (HDPE) and 2 grams of cardboard.
Three improvement options are defined. The first deals with the use of smaller boxes, removal of trays, increasing
product quantity, refill packages and use of thinner material. This is modeled by lighter boxes (28
grams). The second option replaces the inner bag by sealing of the box. The third improvement replaces the
plastic blister with a cardboard blister (17.5 grams) as done intensively in the DIY sector [SVM 1992-1996].
2.4 Flexible packaging
From a materials point of view this group of packaging materials basically consists of three subgroups: plastic
films, paper wrappings and aluminum films. Plastic films can be further broken down into different types of
monolayer films and laminates. The total European market for flexible packaging amounts to some 5.2 million
tonnes, of which PE films are estimated at 4.2 million tonnes and PP film at 0.8 million tonnes [APME, 1996].
Possible material changes in flexible packaging depends heavily on the products that are packed. For many
food products barrier properties for moisture and gasses, especially oxygen and carbon dioxide, play a crucial
role. Other products do not need high barrier properties. For this reason we modeled high and low barrier films.
In case of low barrier food films we modeled 1 liter bags out of LDPE (3.7 gram) and PP (2.7 gram). Substitution
between the films can already improve material efficiency but specific improvement options are metallocene
films which are about 20% thinner (2.2 grams) due to improved polymerization control in the production
process [Anon. 1995, 1996, v. Stijn, 1996, 1997] and the use of paper wrappings (8 grams). In case of high barrier
films several laminates are modeled which consist of two or more layers of different materials. Laminates
either consist of two or more different plastics, or of a plastic with another material (e.g. aluminum or paper).
We modeled PP, PET and metallocene laminated with PVdC (a super thin coating with excellent barrier properties)
and the same plastics with a coating of aluminum.
Flexible packaging can also be used to pack non-food packaging and industrial packages. For non-food packaging
we used the same films as for low barrier food packaging. For industrial packaging typical films are
stretch films and shrink covers. These are used to bundle several packages together and often fix these bundles
to pallets. Shrink covers are assumed to be 140 m thick and a reduction to 100 m is assumed to be possible
[Zoethout, 1997]. Stretch films are assumed to be 30 m thick but multiple turns around a pallet are necessary.
Shrink films can in some cases also replace corrugated boxes. For shrink films we used an average thickness of
50 m and a possible reduction in thickness of 10% [SVM, 1994, Plasthill, 1997, Zoethout, 1997].
2.5 Carrier bags
Carrier bags are most often made out of plastics, more specifically PE. Both LDPE and HDPE are used for the
production. HDPE bags are normally lighter than LDPE bags [SVM, 1992]. Many different kinds of carrier
bags are used in Europe which depends on the specific function of the bag. The thickness varies between 10 to
200 m [Donker, 1993].
Several initiatives have been taking place in order to reduce the amount of plastics used for plastic bags. Many
projects focussed on prevention. In the Netherlands this resulted in agreements in 1991 between the government
and stores that plastic bags will not be handed out for free [CV, 1992]. These measures resulted in a reduction
of the amount of carrier bags because consumers started to reuse bags or make use of durable carrier
bags. Other initiatives focussed on alternative materials for carrier bags like paper.

We estimate the amount of carrier bags in Europe in 2000 at 430 ktonnes based on a European consumption in
1990 of 370 ktonnes which is linearly extrapolated with the estimated developments of PE in that period
[APME, 1992, APME, 1994, APME, 1996]. Furthermore we assume an average weight per bag of 20 grams
[Donker, 1993]. This results in a European demand for carrier bags of 21.5 billion bags per year.
Improvement options for the normal carrier bag are a bag that consists of recycled PE, a lighter PE bag (15
grams), a paper bag (56 grams) and a multiple use PE bag which weighs 240 grams but can be reused 100
times [SVM, 1993, Donker, 1993].
2.6 Industrial bags
Industrial sacks are used to pack plastics granulate, cement, animal feed, fertilizers, flour, soda, gypsum, compost
etc. In APME (1992) the total use of plastic industrial bags is estimated at 460 kton. We will use this figure
as an estimate for the amount of plastic industrial bags in 2000. The total plastic industrial bag consumption
is estimated at 4.4 billion bags considering an average weight of 105 grams per bag with a carrying capacity
of 25 kg [calculations based on: Bührmann-Vromen, 1997, Zoethout, 1997].
Besides plastic bags also paper bags are used. We will make an estimate for the amount of paper bags based on
the amount of cement that is consumed in Europe (10% is carried in bags). Cement bags are made out of multiple
layers of paper because of the bag needs to be very strong. The sacks have a standard carrying capacity of 50
kg which equals 25 liters in the case of cement. Today a clear trend is visible towards smaller bags because they
are easier to handle [Ayoup, 1997]. The amount of paper bags is calculated at 340 million bags (assuming bags
of 50 kg). The paper bag is modeled as a bag that weighs 262 of which 10 grams PE.
A possible improvement option for industrial bags is the Flexible Intermediate Bulk Container (FIBC). This is
very strong bags made out of woven PP straps. We modeled a multiple use and a one way FIBC with carrying
capacities of 1000 kg and weights of 2500 and 2000 grams respectively [van Well, 1997].
2.7 Industrial Boxes
Industrial boxes are used for several purposes. First of all they can be used for packing loose product like fruit,
vegetables and machine parts. Secondly, they can be used to pack several (cardboard) boxes or (plastic)
pouches. We will define the first category as crates and the second as transport boxes.
Three types of crates can be defined that are used in Europe to ship mostly loose products like fruit, vegetables,
meat and product parts. We define the plastic multiple use crate, the wooden crate and the cardboard crate.
Multiple use plastic crates weigh 1.5 – 2 kg and have a volume of 40 liter according to Burggraaf (1997). A
wooden crate with the same volume weighs about 2.2. kg [PFK, 1997]. Crates made from corrugated board
weigh only 0.4 kg but can only be used once while the plastic and wooden crate can be used 100 and 30 times
respectively [PFK, 1997, Wiemers, 1996, Pitt, 1996].
The category transport boxes differs a lot from the category ‘crates’. This is due to the fact that folding boxes
made out of corrugated board are the standard. Transport boxes, as we defined them, are normally used to pack
other boxes. To estimate the amount of boxes a standard box should be defined. Boxes are used in many sizes
depending on the products that are packed. We will define a standard corrugated box with a volume of 40 liter
and a weight of 800 gram [BV, 1996]. In 1995 the consumption of corrugated materials in Western Europe
amounted 11.7 Mtonnes which leads to a total consumption of 14.6 billion boxes.
Several options are available to make a more efficient use of transport packaging. Improved gluing techniques,
changes to the shape of boxes, removal of top flaps and standardization of packaging has shown that less corrugated
board is needed to fulfill the same packaging service [SVM, 1992-1996]. We assume that 20% less corrugated
board can be used in the future for the same service. Besides these efficiency options also substitution is
possible. The most promising are the returnable transport crate and the use of industrial shrink films. Earlier in
this paper we have described these options.

2.8 Pallets
The last packaging category concerns pallets. In many industrial and trade sectors pallets are used intensively
for internal and external transport of products. In Europe the production of pallets amounts to 280 million pallets
per year [van Belkom, 1994]. About 96% of these pallets are made out of wood. Not too long ago, almost
all pallets were used for one single trip and discarded afterwards. Due to environmental legislation an obvious
trend is visible towards the use of multiple-use pallets. In Germany for example, the Verpackungs-Verordnung
states that the taking back of pallets by industries is compulsory. The trend towards multiple use pallets has led
to a large increase of the number of pallets that are part of pallet pools. In 1994 about one third of the pallets
was returnable [van Belkom, 1994].
We have modeled pallets by defining several pallets made from different materials like wood, PE, recycled PE,
recycled PC, corrugated fiberboard and pressed wood fiber.
The wooden pallets can be either single or multiple trip pallets. A single trip pallets weighs about 17 kg and a
multiple use pallets weighs 25 kg. The main wood type used in production is softwood. Advantages of the
wooden pallet is that it is cheap (USD 5 - USD 20) and made from renewable resources. The multiple trip pallets
has an estimated trip number of 40 trips.
Plastic pallets are used a lot in the food industry because they are easy to clean due to the smooth surface. Furthermore,
no liquid can be absorbed by the pallets [Johnson, 1997]. The most common material for plastic pallet
production in PE but in some cases also PC is used. Pallets made out of PC are stronger than PE pallets. Just
like wooden pallets, plastic pallets can be used for single and multiple trips. A one-way plastic pallet weighs
about 14 kg and a multiple use pallet weighs around 30 kilograms [TNO, 1994, van den Berg, 1996]. The costs
of plastic pallets are higher than for wooden pallets. A PE pallet for multiple use costs around USD 75 and a
pallet for single use costs about USD 12. If the pallets are made from recycled PE, they are much cheaper. The
price for a multiple use recycled PE pallet starts at USD 25. Pallets that are made out of recycled PC are more
expensive (USD 75) [van den Berg, 1996].For the multiple use plastic pallets we will use a trip number of 75
trips.
Pallets made from corrugated fiberboard are an option to replace single trip wooden pallets. The pallet costs
about USD 6 which is cheap compared to wooden and plastic pallets [anon. 1993a]. The pallet weighs about 6
kg which makes it a very light-weight pallet. This already has been a reason for some companies to use this
pallet because it reduces the weight in the trailer [Witt, 1990].
Pallets can also be made from pressed wood fibers. The advantage of these pallets is that they can save a lot of
space if they are used for multiple-trip purposes because they use a fourth of the space of piled wooden pallets
when stacked empty. Pressed wood fiber pallets are made out low grade fibers, mostly from bark and thinnings.
The fibers are molded into a pressed wood pallet with the use of synthetic organic resins. The average costs of
these pallets are about USD 5 and the average weight of the pallet amounts to 16 kg [anon., 1993a]. We will
use a trip number of 5 trips per pallet.
3 Costs and energy consumption calculations
The introduction of new packaging (technology) is associated with a change in costs, both to the producer of
the packaging as well as to the user (packager, transport, consumer). The viability of a material efficient packaging
system will depend on the level of the costs. It is therefore, necessary to assess the costs of the various
packaging systems. Furthermore changes in material use will have direct effects on energy use during production
and transport of the packaging which can either strengthen or diminish the effects of different packaging
system. To estimate these effects also the energy consumption of packaging making and transport should be
calculated.

3.1 Cost calculations
The costs of a packaging system can be subdivided in capital costs for packaging making and filling facilities,
packaging material costs, labor costs during packaging making, filling and transport, maintenance costs during
production and maintenance cost of the package in case of reusable packages (collection transport etc.).
The material costs are often an important part of the total costs of packaging. These costs are calculated by the
MARKAL model depending on the production technologies in different years. Investment costs for packaging
making and filling facilities are based on data of several facilities built recently [Packaging Week, 1992-1997].
For returnable packaging we assume doubling of the investments and labor costs (during production) because of
the extra equipment (and space) necessary for intake and cleaning. Labor costs of filling lines are estimates
based on [van Vugt, 1992, 1993]. Labor costs for packaging making are estimated at 20% of total production
costs. For transportation we differentiated between the costs for moving and for loading the products. For multiple
use packaging we took loading costs into account for returning the package and storage costs.
3.2 Energy calculations
Energy use for packaging is generally very small, compared to the energy content of the materials and the energy
use for production. We used energy consumption figures for packaging making as determined in Buwall
(1991, 1996). For some types of packaging these are negligible compared to material production, e.g. corrugated
box making and glass blowing [Buwall, 1996]. For returnable packaging we took the energy consumption
for cleaning into account. For bottles these are estimated at 1 GJ heat and 126 MJel per 1000 bottles. Finally we
took the transport energy into account for the return trip, from store to factory, of multiple trip packaging. The
trip from factory to store we attributed to the packed products instead of the packaging material.
4 Results
The data collection has led to information about the characteristics of packaging used today, possible alternatives
for the future, and production costs and energy consumption for all options. In order to model these aspects
in MARKAL however we need to define demand categories. These demand categories are necessary in
order to allows the model to substitute between different packaging categories. For example: if the demand for
carbonated drinks is defined, the model may choose between metal cans, PET bottles and glass bottles to satisfy
the packaging demand for carbonated drinks. In contrast: if PET bottles were defined as a demand category,
substitution by metal cans or glass bottles would not be possible. In Table 4.1 the final demand categories are
stated. The demand categories are chosen in such a way that they make optimal use of packaging characteristics.
For example: carbonated and non carbonated drinks are two separate categories because carbonated drinks
require higher CO2 barriers from the packaging material than non carbonated drinks. To estimate the magnitude
of the demand categories and the volume of all individual packaging types we made use of production statistics
of consumer goods [EC, 1997] and information on packaging material consumption [e.g., PPI, 1996,
APME, 1992, 1994, 1996].
For 2000 the division of packaging materials over the demand categories is predefined. For 2020 however the
MARKAL model is free to use other packaging options or make shifts in the current division, taking into account
maximum shares which were determined based on restriction caused by the properties of the packaging
material related to the claims of the product.
Table 4.1 shows that the MARKAL model has a lot of freedom in the choices for different packaging options in
2020. The potentials have to be seen as technical potentials. We did not take any implementation barriers (next
to technical ones) into account. The reason for this is that with these bounds the MARKAL model can calculate
the technical potential of material efficiency and later, depending on the model results, we can give packaging
options minimum bounds in order to create a more realistic scenario.
Even though the input data for the MARKAL model are stated in terms of energy consumption and costs
(excluding energy consumption and costs related to materials), in Table 4.1 the CO2 emissions and costs are
stated (related to both the materials and the packaging process). These data are first order estimates because

they are not a result of the MARKAL model. The process CO2 emissions relate to the emissions during packaging
making, filling and transport. The material related CO2 emissions are calculated by using CO2 emission
factors for the different materials based on van Duin (1997). The process costs refer to the total lifecycle cost
excluding material costs.

Table 4.1 The use of packaging for the demand categories as defined for the MARKAL model, including
possible improvement options, possible market shares in 2020, costs and CO2 emissions (CO2
emissions and costs are expressed per 1000 liter packed (for pallets per 1000 trips, for carrier
bags per 1000 pieces).
Demand Demand Share Max. Material Process Material Process
category 2000 2000 share related CO2 related CO2 costs costs
2020 emissions emissions
(%) (%) (kg) (kg) (ECU) (ECU)
beverages, 46 billion 8 100 steel bev can 128 44 45 70
carbonated liter 0 100 all steel can 143 47 50 70
0 100 ultra light steel can 103 40 36 70
8 100 aluminum bev can 250 35 64 70
0 100 ultra light alu can 232 34 59 70
37 50 PET one way 233 70 34 15
6 50 PET Refill 24 43 4 26
1 50 improved PET refill 15 43 4 26
10 100 glass large 180 24 90 23
27 100 glass small 300 24 150 76
0 100 light glass 135 24 68 23
3 100 glass refill 15 96 8 115
beverages, 66 billion 20 100 glass large 180 24 90 23
non carbonated liter 0 100 light glass 135 24 68 23
54 80 liquid board 60 21 21 15
26 50 PET one way 233 70 34 15
0 30 PET Refill 24 43 4 26
0 30 improved PET refill 15 43 4 26
0 50 pouch 20 25 3 18
0 50 PC bottle 20 28 2 20
dairy products, 19 Mtonne 84 100 PS cup 164 62 26 44
no milk 16 100 PP cup 168 57 17 44
0 90 glass jar 180 24 90 45
0 90 light glass jar 144 24 72 45
wet food 4 billion 50 100 glass jar 180 24 90 45
light glass jar 144 24 72 45
liter 50 100 steel food can 154 49 142 46
honeycomb can 108 49 99 46
dry food, 663 billion 8 80 cardboard box 18 13 26 19
non susceptible liter 2 100 cardboard box + bag 20 13 30 15
1 100 PVC box 140 13 24 39
30 70 LDPE-film 19 8 3 18
44 80 PP-film 19 8 2 18
14 90 metallocene 11 8 2 18
1 60 paper 4 7 4 18

Table 4.1 (continued) The use of packaging for the demand categories as defined for the MARKAL model, including
possible improvement options, possible market shares in 2020, costs and CO2 emissions
(CO2 emissons and costs are expressed per 1000 liter packed (for pallets per 1000 trips, for carrier
bags per 1000 pieces).
Demand Demand Share Max. Material Process Material Process
category 2000 2000 share related CO2 related CO2 costs costs
2020 emissions emissions
[%] [%] [kg] [kg] [ECU] [ECU]
dry food, 310 billion 67 100 PP-laminate 22 8 2 18
susceptible liter 8 100 PET-laminate 20 8 3 18
9 100 metallocene -lam. 14 8 2 18
4 10 PP-metalised 11 8 1 18
11 10 PET-metalised 11 8 2 18
1 10 metallocene -met. 7 7 1 18
non-food liquids 10.5 billion 100 100 HDPE bottle 500 161 72 44
liter 0 100 Rec. HDPE bottle 250 161 72 44
0 100 pouch 50 27 7 36
0 50 liquid board 14 24 21 44
dry non-food 168 billion 14 80 cardboard box 18 13 26 19
liter 10 100 PVC box 140 62 24 39
0 20 cardboard blister 18 13 26 39
37 80 LDPE-film 37 8 3 18
27 80 PP-film 27 8 2 18
11 80 metallocene film 22 8 2 18
1 60 paper 8 7 4 18
carrier bags 21.5 billion 85 100 PE bag 100 6 14 3
bags 0 100 recycled bag 60 6 14 3
10 100 paper bag 28 0 26 3
5 100 multiple use bag 12 1 168 2
industrial bags 200 Mtonne 82 82 PE bag 21000 1223 3 1
9 9 paper bag 2620 0 2 1
9 40 FIBC one way 10000 582 1 0.3
0 40 FIBC returnable 2500 29 0.4 0
transport 1671 billion 16 100 plastic crates 3 17 0.4 17
packaging liter 1 20 wooden crates 2 0 0.5 15
1 20 corrugated box 10 0 6 93
32 50 shrink foil 3 0 0.4 3
50 100 cardboard crate 5 0 3 93
0 100 improved corr. Box 8 0 5 93
pallets 5.6 billion 3 100 wood one way 17000 0 4250 2667
trips 60 100 wood returnable 625 20 156 378
37 100 PE returnable 2000 159 288 649
0 100 PE one way 70000 4861 10080 12666
0 100 PE recycled 1000 139 288 649
0 100 PC recycled 1000 139 412 649
0 100 Corrugated fi- 3000 0 1800 2000
0 100 pressed wood 3200 0 800 142
pallet wrapping 1.0 billion 42 80 shrinkcovers 5000 291 720 760
trips 58 100 stretchfilm 3000 175 432 760

5 Factor 2 or factor 10?
Is it possible to reduce the CO2 emissions related to packaging use with 50% (factor 2) or is it maybe possible
to reduce the CO2 emissions with 90% (factor 10)?
To answer these questions we need the results of the MARKAL model. However, based on the results as stated
in Chapter 4 we can try to foresee what may be the result of the implementation of various options. In such an
exercise we can not take into account all the effects that different options have on each other.
If we go through the packaging options as stated in table 4.1 (and background information in Chapter 2 and
Hekkert et al. (1998) we can see that in general the following options are possible:
- Lighter packaging. This can be the result of using thinner material (foils, cans), using a different packaging
shape (cans, bottles, boxes), using more efficient packaging and filling machines (boxes). To implement
these options very often only small changes are necessary in the current production methods. For
many packaging types that are used today, material reductions of 10 - 20% are possible.
- Reusable packaging. This option has a large potential in material efficiency improvement. Reusable packaging
is possible for packaging of liquids (PET bottles and refillable non-food bottles) and in the field of
industrial transport packaging (industrial bags, crates and pallets). Trip numbers of 20 are not uncommon
which shows the large potential of this option. The savings in material consumption (and the related energy
consumption) need to be compared to additional energy requirements due to washing and extra transport).
Table 4.1 shows that the reduction in material requirement outweighs the increase in CO2 emission
due to cleaning and extra transport. Furthermore the reduced material costs are larger than the increase in
process costs.
- Material substitution. The effects of substitution options are very hard to describe in general because many
different types of material substitution are possible and it strongly depends on other variables (e.g. energy
consumption in material production, possibilities of material reuse etc.). In general two possible trends are
visible. First it is possible to use much more natural organic materials that are renewable. These materials
may be CO2 neutral and therefore attractive to use from a CO2 point of view. On the other hand these materials
are often in competition with plastic packaging that can be reused which may diminish the positive
effects of natural organic materials. Secondly current packaging can be substituted by light-weight alternatives
(PE pouch, foils). The effects of this type of substitution depend strongly on the original material. Table
4.1 shows that both options are very useful to improve the material efficiency. Only reusable packaging
seems to be a better option.
- Use of recycled material. In food packaging this option has limited possibilities due to regulations related
to hygiene considerations. However, smart technologies make it possible to use recycled material even
within this sector (e.g. multi layer PET bottles). In other sectors the use of recycled material has large
potentials (pallets, crates, carrier bags). Taking into account that for plastics the feedstock may account for
as much as 67% of the total energy input, energy reductions with a factor 2 are in some cases not unlikely.
When we add the possibilities of the options stated above we may conclude that a reduction in CO2 emission
with a factor 2 should definitely be possible for the product category packaging. We need to take into account
that this only holds if we assume an equal consumption level because a strong growth in consumption will also
lead to a strong growth in packaging consumption.
A reduction in CO2 with a factor 10 seems not very realistic with the current identified packaging options. Even
a total shift toward returnable packaging will not be enough to reach this goal. Large changes in society are
necessary to reach this goal. It is beyond the scope of this paper to get into options that may result in a 90% reduction
of CO2 emissions.
6 Policy Consequences
The packaging industry is a very dynamic and innovative sector. The sector is always trying to make packaging
as attractive, easy to open and strong as possible. Their first concern is that packaging needs to protect products

or keep it fresh and their second concern is related to marketing considerations. Material efficiency improvement
is only attractive if it saves money. Fortunately, the cost of packaging is for a substantial part related to
the material input. Therefore some measures are likely to be implemented voluntarily like weight reduction of
steel and aluminum cans and plastic and glass bottles.
From current European policy as stated in the Packaging Directive we may expect that the goals are reached at
some point in the future. The goals as stated in the packaging directive (e.g. 25-45% recycling of packaging
waste [EU, 1994]) are not very likely to reach a 50% CO2 reduction before 2020. The second Packaging Covenant
in The Netherlands goes further than the European directive. It aims for a total reduction of the amount of
packaging waste to 940 ktonnes in 2000 compared to 1314 ktonnes in 1995, or a reduction of 28%. For the
different packaging materials sub-covenants have been made for more detailed goals per material. Even though
large waste savings may be the result of this covenant, a reduction in CO2 emissions with a factor 2 is not likely
to be the result.
In order to reach CO2 emission reductions with a factor 2 or more in the field of packaging and to use the potential
of available options a more stringent policy is needed. It is not within the scope of this paper to argue for
special measures or even to indicate where to focus on. However, it seems that the effects of 'factor 2' policy on
the packaging industry in The Netherlands will not be as great compared to the average European situation.
This is mainly due to the fact that in The Netherlands the use of returnable packaging is much more common
than in other parts of Europe. Furthermore it is very likely that 'factor 2' policy will not affect the packaging
industry as much as the materials producing industries. The packaging industry will use their innovative
strength to produce different packaging products (as became visible in The Netherlands during packaging
covenant I) but the plastics, paper and board and the iron and steel industries will be forced to produce less than
before. However, the production high value added products (e.g. ultra thin steel, high tech foils) may weaken
this trend.
7 Conclusions
The packaging sector is an important sector from a material consumption point of view. The sector is responsible
for about 3.3 percent of the European CO2 emission. The most important materials used in this sector in
terms of CO2 emission are plastics, metal and paper and board.
The use of packaging materials is closely related to the use of consumer goods. The latter is likely to go through
a strong growth the coming decades and therefore a growth in packaging consumption may be expected. Even
though European policy by means of the Packaging Directive is aiming at increasing the recycling targets a net
increase of virgin packaging materials may be expected.
Many options are available to reduce the amount of packaging material. This paper shows that these options
may lead on first sight to a reduction of CO2 emission by the packaging sector with at least a factor 2 by the
year 2020. The options will de used as input data for the MARKAL model in order to calculate the total CO2
emission reduction possible.
The options dealt within this paper are specified per packaging category and by packed product category. However,
in general the options can be classified by four categories: lighter (more efficient) packaging, reusable
packaging, material substitution and use of recycled material.
Making more efficient use of packaging materials often leads to cost effective savings of 10 - 20% packaging
material. The use of reusable packaging has great potentials. Returnable or reusable pallets, crates and PET
bottles have trip numbers varying from 5 to 100 trips which has large effects on the amount of packaging material
needed per packaging service. The influence of material substitution on CO2 emissions is not clear. The
MARKAL model should create more insight in this matter. The use of recycled materials is in some parts of the
sector (industrial transport) very promising and in others it is more problematic (food packaging).

In terms of implementation and policy our study shows a potential of about 20% is available to reduce the
amount of packaging material and related CO2 emissions without large changes in consumer behavior or large
changes in production methods. Furthermore much higher reductions can be reached with small behavioral
changes. Packaging options of this type are the use of reusable and refillable packaging and the use of recycled
materials.
References
[1] Abbott, R., 1995, The ultimate can, Packaging today, June 1996
[2] Anonymous, 1993a, Selection guidelines for pallets and slipsheets, Modern Materials Handling, Nov.
1993
[3] Anonymous, 1995, The role of PET, Plastics Bulletin.
[4] Anonymous, 1996, European Glass Recycling in 1995, Glass Gazette, September 1996
[5] APME, 1992, Information system on plastic waste management in Western Europe; 1990 data, Brussels,
Belgium.
[6] APME, 1994, Information system on plastic waste management in Western Europe; 1992 data, Brussels,
Belgium.
[7] APME, 1996, Information system on plastic waste management in Western Europe; European overview;
1994 data, Brussels, Belgium.
[8] Ayoup, R., The importance of the paper bag to the cement industry, World cement, January 1997.
[9] Bührmann-Vromen Verpakking, 1997, Personal communication with BVV, 8 December 1997.
[10] Burggraaf, 1997, Personal communication with Mr. Burggraaf, van der Windt verpakking, Honselersdijk,
The Netherlands dd. 8 December 1997.
[11] BUWAL, 1991, Oekobilanz von Packstoffen, Stand 1990, Schriftenreihe Umwelt Nr. 132, Bundesamt für
Umwelt, Wald und landschaft (BUWAL), Bern, Germany
[12] BUWALL, 1996, Okoinventare für Verpackungen, Band 2, Schriftenreihe umwelt n. 250/II, Bundesambt
für Umwelt, Wald und Landschaft, Bern, Switzerland
[13] BVV, 1995, Packaging Catalogue, Buhrmann-Vromen Verpakking, Zutphen, The Netherlands
[14] Couwenhoven, E., 1992, Intergamma pakt consumenten- en transportverpakking aan, Verpakken.
[15] Couwenhoven, E., 1996, Innovaties in melk- en vruchtensapverpakkingen nog niet ten einde,
verpakkingsManagement, December, 1996.
[16] 1992, Jaarverslag 1992, Comissie Verpakkingen, Amersfoort, The Netherlands, 1992
[17] Depijper, J., 1996, Milieu vriendelijke succesproducten, packaging 2000,
[18] Donkers, 1997, AEP Borden Flexible Packaging Apeldoorn, The Netherlands, Personal communication, 9
December 1997
[19] Dijk, J., 1992, Supermarkt wil groenten en fruit in karton, Verpakken, Dec. 1992
[20] EC, 1997, Panorama of EU industry 97, European Commission DG III, Brussels, Belgium
[21] Ent, J. van der, 1995, Glas niet stuk te krijgen, Missets pakblad, Juni 1995
[22] Ent, L. van der, 1995, Metalen verpakkingen lichter en lichter, Missets pakblad, December 1995
[23] Van der Ent, L., 1995a, Europese doorbraak SiOx lijkt in aantocht, Missets Pakblad, October 1995
[24] EU, 1994, Directive 94/62 on Packaging and Packaging Waste, European Union, Brussels, Belgium.
[25] Gielen, D.J., 1997, Technology Characterisation for Ceramic and Inorganic Materials, Input data for
Western European MARKAL, The Netherlands Energy research Foundation., Petten, The Netherlands
[26] Goddard, R., 1994, Refined approach to loosing weight, packaging week, nov 10 1994
[27] Hekkert, M.P.,1995, “Energetische en Economische optimalisatie van de Nederlandse
afvalverwerkingsstructuur in 2010 door Inzet van Nieuwe Verwerkingstechnieken”, Report No. 95061,
Dept. of Science, Technology & Society, Utrecht University, Utrecht, The Netherlands.
[28] Hentzepeter, V., 1996, Einde successtory PET nog lang niet in zicht, Foodmanagement, 14/10.
[29] Johnson, L., 1997, Moving Matters, Packaging Today, August 1997.
[30] Kort, T. van der, 1992, Kartonnen blister goed alternatief, Verpakken, September, 1992
[31] Van der Kort, T., 1995, Laminated worden steeds dunner, Verpakken, December 1995?.
[32] Meert, B., 1995, het honingraatblik, Intermediair, Sept. 1995

[33] Packaging Week 1992-1997, many issues of Packaging Week in the period 1992-1997, especially the
machinery review section.
[34] PFK, 1997, Personal communication with PFK, Heinenoord, The Netherlands dd. 8 December 1997.
[35] Phylipsen, G.J.M., 1993, Functiegerichte optimalisatie van kunststof gebruik, Een voorstudie voor een
computer model, Report no. 93047. Department of Science, Technology and Society, Utrecht University,
The Netherlands.
[36] Pitt, P., 1996, A case for crate expectations, Packaging Week, September 19 1996.
[37] Plasthill, 1997, Plasthill Hillegom, The Netherlands, Personal communication, 10 December 1997
[38] PPI, 1996, Liquid Packaging: Land of milk and money, Pulp and Paper International, August 1996
[39] PPI, 1997. International Fact & Price Book 1997, Pulp & Paper International, Miller Freeman Inc., San
Francisco, US.
[40] SPG, 1994, Glasbak vervult voorbeeld functie, Nieuws glasbak, Stichting Promotie Glasbak, December
1994.
[41] Stijn, A. van, 1996, Meer mans met metalloceen-PE, Missets Pakblad 9, September 1996
[42] Stijn, A. van, 1997, Metalloceen-PP voorbestemd voor de foodsector, Missets Pakblad 1, January 1997
[43] SVM, 1992, Packaging developments ‘92, Implementation of the packaging covenant illustrated, Stichting
Verpakking en Milieu.
[44] SVM, 1993, Packaging developments ‘93, Implementation of the packaging covenant illustrated, Stichting
Verpakking en Milieu.
[45] SVM, 1993a, Verslag van Stichting Verpakking en Milieu aan de Commissie Verpakkingen, July 1993.
[46] SVM, 1994, Packaging developments ‘94, Implementation of the packaging covenant illustrated, Stichting
Verpakking en Milieu.
[47] SVM, 1995, Packaging developments ‘95, Implementation of the packaging covenant illustrated, Stichting
Verpakking en Milieu.
[48] SVM, 1996, Packaging developments ‘96, Implementation of the packaging covenant illustrated, Stichting
Verpakking en Milieu.
[49] TNO centrum voor verpakking, 1994, Missets pakblad, Nov. 1994
[50] Van Belkom, A., 1994, Kunststof pallets in opmars, Kunststof en Rubber, Dec. 1994
[51] Van Stijn, 1996, Blik(vangers) in topvorm, Missets pakblad, August 1996
[52] Vugt, H.H.A. van, 1993, Nieuwe bottellijn vppr PET-flessen bijj Coca Cola Dongen, VMT, August 1992.
[53] Vugt, H.H.A. van, 1993, Nieuwe bussenafvullijn voor Heineken Zoeterwoude, VMT, March 1993.
[54] Well, van, 1997, Personal communication with Mrs. van Well, Ligtermoed B.V. Flexible Industrial
Packaging, Ridderkerk, The Netherlands.
[55] Wiemers, C., 1996, Nieuw CBL uitpakkrat kostenbesparend, Verpakkings-management, July, 1997. Witt,
C.E., Pallets, wood isn’t the only answer, material handling engineering, Nov. 1990
[56] Zoethout, 1997, Plasthill, Hillegom, The Netherlands, personal communication, 23 December 1997
[57] Worrell, E., A.P.C. Faaij, G.J.M. Phylipsen, K.Blok, 1995, An approach for analysing the potential for
material efficiency improvement, Resources, conservation and recycling 13 (1995) 215-232

Energy use reduction potential of passenger transport in Europe and consequences for
CO2 emission
M.E. Bouwman and H.C. Moll
Center for Energy and Environmental Studies IVEM, Groningen University,
Nijenborgh 4, 9747 AG Groningen, The Netherlands
tel. +31 50 3634609; e-mail M.Bouwman@fwn.rug.nl
Abstract
In order to contribute to a sustainable society, considerable reduction in the energy use and CO2 emission
should be achieved. This paper presents a methodology to explore the energy use reduction potential of the
transportation sector. Three types of options are defined emphasising on technological, infrastructural and
behavioural change. With technological and infrastructural options, an energy reduction of over 50% by 2050
can be realised. In order to achieve further energy reductions, also options with large behavioural impact should
be implemented. This results in an 80% energy reduction potential in the transportation sector by the year
2050. The reduction potential for CO2 emission can be increased compared to the energy use reduction
potential by introducing a fuel mix with a low carbon content. If a growing mobility demand is incorporated in
the analyses, the total savings are smaller.
Introduction
The present use of energy and materials and the emission of greenhouse gases of Western Europe contribute
substantially to global resource depletion and global climate change. To contribute to global sustainability, i.e.
stabilisation at 1990 emission level by the year 2050 (Houghton et al., 1995) and leaving room for development
outside the Western world, considerable reductions of the use of energy and materials are required, resulting in
the required reduction of greenhouse gases (mainly CO2).
There is a variety of reduction options such as technological improvements, infrastructure modifications and
behavioural change. Technological improvements aim at increasing the energy and material efficiency and at
substitutions, resulting in a lower overall energy and material consumption. Some technological options
demand large behavioural adaptations e.g. shifting from private to public transportation. Some consequential
substitutions require also a new or a heavily modified infrastructure e.g. to sell a new energy carrier, or to
provide the infrastructure for increased use of public transportation. All such options can be evaluated
individually on their potential to reduce CO2 emissions.
Transport i.e. passenger and freight transport by road and railways contributes for about a quarter (OECD/IEA,
1997) [14 EJ/a] to the total energy consumption and CO2 emissions in OECD Europe. This share justifies a
specific assessment of reduction options in the transportation sector. Moreover, a substantial growth (RIVM,
1992) is foreseen for the performance - expressed in passenger kilometres and ton kilometres - with possibly a
related growth in energy requirements and CO2 emissions.
To stabilise the energy requirement and emissions by transport, at least the impact per unit of performance has
to be halved in he period 2000 - 2050. An equitable distribution of CO2 emissions on a global level in 2050
requires an absolute reduction to one fifth of the current energy consumption and emission levels for OECD
Europe (cf. Mulder and Biesiot, 1998). Combined with the expected growth of transport this long term
reduction objective implies a reduction of the impact per unit of performance to 10% of the present one.
This paper calculates the reduction potential of total energy use of and emissions from passenger transport in
OECD Europe. The general aim is to achieve a 50% reduction by 2020, and a 90% reduction in 2050.
Technological improvements, infrastructure modifications and behavioural change are all considered,
regardless of their implementation problems. This approach yields an indication of the reduction potential. The
reduction potential is calculated for both the individual category of options, and the combined total.

GER values
Direct energy use
(MJ/vkm)
Indirect energy use
(MJ/vkm)
Occupancy rate
Direct energy ue
(MJ/pkm)
Indirect energy use
(MJ/pkm)
ERE value (MJ/MJ)
Total primary energy
use (MJ/pkm)
Methodology
The reduction potential of passenger transport is calculated in relative terms, resulting in a reduction percentage
per unit of transport. The unit to compare the situations in 1990 and 2050 is the transportation of one person
over a kilometre distance (pkm). This unit comes closest to the demand for transports, which is usually not
a demand for vehicle kilometres (vkm), but a demand to cover a certain distance. The magnitude of the demand
is largely influenced by the spatial settings and the activity patterns of each individual.
For a reduction potential analysis, the motorised transport demand is of importance. We discern individual and
public transportation systems. In Europe, fourteen per cent of the transport demand is covered with public
transport, 86 % with individual passenger cars (Schol and Ybema, 1998). We do not include soft modes like
walking and cycling in our analysis, as their energy use and emissions are minimised already. Information
about the overall share of soft modes in Europe is not available. Of the total mobility demand in the Netherlands
in 1994, about 10 per cent is covered with soft modes (CBS, 1995). This share is likely to be relatively
large compared to the European average, as the share of cycling is relatively large in the Netherlands.
The reference energy use in 1990 is the energy use per passenger kilometre in a passenger car. This energy use
is calculated, based on both the direct and indirect energy use of the vehicle, and the occupancy rate.
Figure 1 gives an overview of the calculation of the reference energy use. The direct energy use per vehicle
kilometre comprises fuel used for transportation and depends on engine technology, car size, etc. The indirect
energy use comprises the energy use associated with production and maintenance of the vehicle. The value of
the indirect energy use is influenced by the vehicle production energy use, the energy needed to produce car
materials (Gross energy requirement of GER value of the materials), the annual use of the vehicle and the total
lifetime. In order to convert the direct and indirect energy use per vehicle kilometre to the value per passenger
kilometre, the values are divided by the average occupancy rate of the vehicle. The next step in figure 1 is the
conversion of the direct energy use to the primary energy use. For doing so, the direct energy use is multiplied
by the energy demand for fuel winning (ERE value, or Energy Required for Energy). The goals set in the introduction
on the reduction potential needed for transportation are based on
overall reductions. The total primary energy use per passenger kilometre implies the energy requirements of all
processes associated with the use of a passenger car.
Calculating this value for the average OECD Europe passenger car and occupancy rate provides an energy use
of 0.4 MJ indirect and 1.9 MJ direct per passenger kilometre. The summed total of these two values gives a total
primary energy use of 2.3 MJ per passenger kilometre in 1990. This paper assesses the possibility of reducing
this value with 50% by 2020 to 1.2 MJ per pkm, and with 90% by 2050 to 0.2 MJ per pkm.
Figure 1 Calculating the total primary energy use per passenger kilometer
CO2 emission is directly related to energy consumption. To attribute to sustainability, reducing CO2 emission is
an important goal. The exact amount of CO2 emission can be calculated based on the total energy consumption.
However, the amount of CO2 emission per unit energy use also depends on some external factors, like the pro-

duction system of fuels. For this reason, our analysis is performed for the energy use of transport. Calculation of
the energy use reduction potential excludes the reduction potential of external factors and in this way, the contribution
of the transportation sector is highlighted. At the end of this paper, some attention will be given to the
interpretation of the energy use reduction potential to CO2 emission reduction potential.
The energy use of 2.3 MJ per passenger kilometre is the value for the standard 1990 passenger car. Over time,
passenger car characteristics also change. For this reason, the development in car characteristics are implied in
the analysis. The development in vehicle weight is the most important characteristic for our analysis, as this
relates directly to the fuel economy of a vehicle. A ten per cent weight increase generally results in a five per
cent decrease of fuel economy (Hughes, 1993).
Vehicle weight has increased steadily during the last decades. Two factors influence this weight increase: the
demand for ever larger cars and the addition of more and more safety features and gadgets. Figure 2 displays
the development in Dutch passenger car weight, a good representative of the developments of the average
European car.
Figure 2 shows a clear increase in vehicle weight for both the weight of newly sold cars and the average weight
of the vehicle fleet. This trend is not likely to change in the coming years. Therefore, it is assumed that the average
vehicle weight increases from 1000 kg in 1990 to 1220 kg in 2020 and 1270 kg in 2050. Figure 3 displays
the expected trends in vehicle weight for Europe between 1990 and 2050. The weight increase causes a 17
per cent higher fuel use by 2050 compared to 1990 at constant engine performance (Bouwman and Moll, 1997).
Improvement options
Reducing the energy use per passenger kilometre is possible by changing the magnitude of the direct energy
use, the indirect energy use and the occupancy rate (see figure 1). Changing the ERE values is not fully included
in the present analysis; since reduction in the ERE values requires changes in the energy producing
sector.
Not all possibilities to reduce the energy use per pkm are equally easily to be introduced. Therefore, we make a
subdivision in four categories, each with increasing problems for implementation. The technological improvement
options are derived from the database gathered by (Binsbergen et al., 1994). In (Ybema et al., 1995) a
selection of these options is made, and efficiency improvement figures for 2000, 2015 and 2030 are calculated.
We will use the figures of this database to calculate the reduction potential, supplemented with figures on the
material substitution (Bouwman and Moll, 1997) and some non-technological options. The various categories
of improvement options are described in more detail below. Table 1 lists a quantitative overview of the expected
efficiency improvements.
1. Options emphasising technological change
Technological options generally have the lowest implementation problems. Options in this category will
not influence the functionality of the vehicle. The options included in category 1 comprise the improved
internal combustion engine (IIC), improved tyres and aerodynamics (ITA), continuous variable transmission
(CVT) and modified frame (MF). The IIC includes many different options, like the lean-burn technol-

ogy for igniting the engine, valve steering management, turbo-charging, direct fuel injection, electronic
engine control etc. The term improved internal combustion engine combines the relevant options for both
diesel and petrol vehicles.
Improved tires and aerodynamics offer small savings, but they are also relatively easy to implement. Continuous
variable transmission offers a continuous range of gears to be used. In this way, the engine use is
optimised, resulting in a higher fuel economy of the vehicle. The modified frame comprises all material
substitution options to reduce the vehicle weight. Light weight construction of vehicles is interesting, because
it improves vehicle economy. The most common material in the passenger car is steel, which has a
high density. Substituting a volume unit of steel by lower density materials results in a weight advantage.
Material substitution is not very easy, since a great diversity of requirements is demanded from the construction
material, and steel adequately fulfils most of these requirements. For example, the construction in
aluminium (density about 1/3 of steel) requires thicker parts, as the strength of aluminium is lower than the
steel strength. Material substitution can follow several directions. Generally, new high strength steel alloys,
aluminium and plastics are the most common materials. An extensive review of the possibilities of each of
these materials pointed out aluminium as the most promising (Bouwman and Moll, 1997).
2. Options emphasising infrastructural change
This category comprises the alternative fuel vehicles. The adaptations to be made in this situation are not
individual, but require changes in the infrastructure resulting from the distribution of new fuels. This category
includes the introduction of the electric vehicle, the hybrid vehicle and the fuel cell passenger car.
An electric vehicle is powered by a large battery. The energy use of electric vehicles is very low compared
to the standard internal combustion vehicles. However, large ERE-values for the generation of electricity
(1.7 MJ/MJ in the Europe in 1990) cause a relatively large primary energy demand for electric vehicles.
Electric vehicles are equipped with regenerative braking (RB), a system that stores braking energy. The
hybrid passenger car combines the advantages of both an internal combustion engine (unlimited range) and
electric propulsion (low emissions and noise). Hybrid passenger cars use the electric part in urban traffic
and change to the standard engine when a long range is required. The disadvantage of this system is the
relatively high vehicle weight due to the combination of the two systems. This results in a higher energy
use when driving in the internal combustion configuration compared to the standard vehicle.
A fuel cell passenger car also runs on electricity, but uses fuel which is converted directly into electricity in
the car. The technology is not yet very mature. No vehicles with fuel cell technology are on the market yet.
Implementation before 2020 is highly uncertain, and at present the costs associated with this technology
are very high.
3. Options emphasising behavioural change
These options require important behavioural adaptations. Examples cannot be found in the database used
for the technological options. This category comprises a change in modal split (increased use of public
transport), increased vehicle life, driving small passenger cars and increasing the average occupancy rate
of passenger cars. These category 3 improvement options require societal adaptations. The infrastructural
options are assumed to be easier to implement, as the necessary changes can be influenced with policy
more directly.
For the change of the modal split, it is assumed that the share of public transport in passenger mobility
doubles from 14% to 28% by 2020. This implies that this option is only valid for a small share in the total
mobility demand. Replacing a passenger car kilometre by a public transport kilometre has an average energy
advantage of 61% in 2020 and 73% in 2050, compared to the 1990 value (reconstructed out of Ybema
et al., 1994; Schol and Ybema, 1998). Increasing the vehicle life from 12.5 to 15 years affects the indirect
energy component. Doing so requires major maintenance investments from the various users of a car. The
effects of a longer lifetime are not very large. The gain from dividing the production energy of the vehicle
(15% of the lifetime energy (Moll and Kramer, 1996)) over a larger number of kilometres is partly lost by
the need for increased maintenance which has an energy requirement of 2 GJ/a (Moll and Kramer, 1996).

A decrease of the vehicle size may also contribute to energy savings, because lighter cars have a higher fuel
economy, especially for frequent stop and drive situations. Introducing small passenger cars requires large
behavioural adaptations, as this counteracts the current trends in the development of vehicle weight (see
figure 2). Driving a 850 kg passenger car in stead of the standard 1000 kg passenger car increases the fuel
economy with 9 per cent (Bouwman and Moll, 1997). For this small vehicle, no weight increases are assumed.
As the weight difference in vehicle size between the standard and the 850 kg passenger cars increases
over time, the advantage in fuel economy also increases over time.
Another method to decrease the energy use per passenger kilometre is to increase the occupancy rate of a
vehicle. By a doubling of the occupancy rate, the energy use per passenger kilometre halves. Since this option
requires large adaptations from the individual user, this option will be hard to realise. Implementing
this option should probably be accompanied by an increased share for public transport, to cover the individual
trips which cannot be combined with other individual trips by passenger car.
4. Mobility reducing options
Next to the options mentioned above, which do not influence the total demand of passenger kilometres,
measures can be taken to reduce the individual mobility demand. A higher population density may reduce
commuter travel and the construction of new roads may result in shorter routings. An increase of the use of
soft modes may also be used as such an improvement option. Since the absolute reduction potential is calculated
per passenger kilometre, the reduction of the number of kilometres is irrelevant. The changes in individual
mobility demand are considered as scenario parameters and will be used for the final calculation of
the absolute reduction potential for the passenger transport sector. Table 1 gives an overview of the expected
energy efficiencies. Only the effect of individual options is shown. Various options may be combined,
to improve the total reduction.
Table 1. Energy reduction of various improvement options
Improvement option 2000 2015 2030 2050
Improve internal combustion Otto 5 % 17 % 23 %
engine a
Improve internal combustion diesel engine a 5 % 20 % 30 %
Improve tyres and aerodynamics a
3 % 4 % 5 %
Introduce continuous variable
3 % 5 % 7 %
transmission a
Introduce modified frame b
3 a
Introduce the electric vehicle
Introduce regenerative braking a
4 a
Introduce the hybrid passenger car
Introduce the fuel cell passenger car a
Change modal split c
Increase vehicle lifetime b
b, c
Drive 850 kg vehicles
Double occupancy rate c
Source: ( a Ybema et al., 1995; b Bouwman and Moll, 1997; and c calculations described in text)
14 % 16 % 17 % 19 %
47 % 59 % 67 % 70 %
0 % 3 % 5 %
3 % 33 % 35 %
NA NA 60 %
61 % 5
73 %
2 % 2 % 2 % 2 %
15 % 20 % 21 % 22 %
50 % 50 % 50 % 50 %

Table 1 gives an overview of improvement options which might be relevant for the reduction of the total energy
use and emissions from passenger transport. The division in four categories gives some indication of the difficulties
that may occur in implementing these options. This paper will not analyse thoroughly these implementation
barriers, nor the costs associated with them.
Calculation of the reduction potential
The information presented in table 1 can be used to calculate the energy reduction potential of passenger transport.
Our goal is to calculate values for 2020 and 2050. To calculate the 2020 reduction potential, the 2015 values
are used. For the calculation of the 2050 reduction potential, only limited information is available. The values
presented in (Ybema et al., 1995) have 2030 as a time horizon. For these options, no extra assumptions are
made on extra saving potentials, which means that the 2030 value is used if no 2050 reduction potential is
given. The total reduction potential can be calculated by multiplying the efficiencies of the various techniques.
In case of a 850 kg passenger car with a modified frame, the efficiency results for 2000 in (1-0.14)*(1-0.15) =
0.73. This means that the combination of these two options results in a 27 % reduction of the energy use.
This section presents the reduction potential per category of improvement options. Reduction percentages are
based calculated compared to the 2.3 MJ per passenger kilometre of the standard 1990 petrol passenger car.
Implementing category 1 improvement options.
Table 2 shows the results of the implementation of the category 1 improvement option. As costs are not a consideration,
all options in the category are implemented.
Table 2. Reduction percentages compared to 1990 in 2020 and 2050 with category 1
improvement options
Passenger car type 2020 2050
Petrol passenger car with MF, ITA, IIC, CVT 6
22 % 29 %
Diesel passenger car with MF, ITA, IIC, CVT 30 % 39 %
Table 2 shows that technological options result in a considerable energy reduction potential, without requiring
major behavioural adaptations. The results in table 2 are corrected for the assumed weight increase. Without
the weight increase, reduction percentages of 37% would be possible in 2050 with the petrol passenger car with
MF, ITA, IIC and CVT.
Implementing category 1 and 2 improvement options
The category 2 improvement options contain the more complicated options. Three new vehicles are introduced.
The immature fuel cell technology will not yet be available in 2020, but will be a very promising option for
2050. The electric vehicle is somewhat more difficult to implement than the hybrid or fuel cell passenger car,
because of its limited driving range. It is therefore assumed that the electric vehicle can contribute to 70% of
the total vehicle fleet.
For both the hybrid and the electric vehicle, the ERE value of electricity is of great importance for calculating
the reduction potential of the vehicle. Therefore, these vehicles are showed twice in table 3, once with a high
ERE value (2.0 MJ/MJ) and once with a low ERE value (1.2 MJ/MJ). Table 3 gives an overview of the reduction
percentages possible with both category 1 and 2 improvement options.
Table 3. Reduction percentages compared to 1990 in 2020 and 2050 with category 1 and 2
improvement options
Passenger car type 2020 2050

Fuel cell passenger car with MF, ITA, RB Not yet available 54 %
Hybrid passenger car with MF, ITA, RB (high ERE value) 31 % 32 %
Hybrid passenger car with MF, ITA, RB (low ERE value) 36 % 37 %
Electric vehicle with MF, ITA, RB (high ERE value) 35 % 44 %
Electric vehicle with MF, ITA, RB (low ERE value) 54 % 59 %
With the new passenger cars, new possibilities to reduce the energy consumption are introduced. The ERE
value for electricity dominates the outcome of the calculation. In a situation with a low ERE value (1.2 MJ/MJ),
the electric vehicle is the most promising option. A reduction of 54% can be achieved with this vehicle by 2020.
With a high ERE value, the fuel cell passenger car is the most promising option by 2050. By 2020, when this
technology is not yet available, the electric vehicle performs slightly better than the hybrid and diesel passenger
car. Considering both cases, about 55 % energy reduction is possible by 2050 (the electric vehicle having a
maximum share of 70 %).
Implementing category 3 improvement options
The preceding two subsections showed that with only technological options, halving the energy use is possible
in 2050. This is still far away from the goal of 90% reduction in this year. To achieve such an energy reduction,
more stringent options should be included.
Next to the options in category 1 and 2, there are category 3 improvement options. Implementing these options
not only decreases the energy use, but in most cases also diminishes the costs associated with transport. Table 4
lists the results of the implementation of the non technological improvement options.
Table 4. Effect of non technological improvement options
Set of options 2020 2050
Without non-technological options 37 % 54 %
With increased vehicle life (+ 2.5 year) 39 % 56 %
With 850 kg passenger car 46 % 60 %
With average occupancy of 3 passengers 68 % 77 %
Doubled public transport share 40 % 57 %
With all non-technological options combined 72 % 80 %
With all non-technological options except doubled public transport share 74 % 81 %
Table 4 shows that especially the doubling of the average occupancy rate in passenger cars substantially influences
the average energy demand. When this high average occupancy rate is combined with small vehicles and
an increased vehicle life, the energy use per passenger kilometre in a passenger car is even lower than in public
transport. Therefore, adding a doubled share for public transport results in a lower reduction. However, such a
high occupancy rate may be very hard to realise when not all single persons trips are made by public transport.
Adding the non technological options results in an 80% energy reduction in 2050.
Implementing category 4 improvement options
With all possible options to reduce the energy consumption of a given mobility demand implemented, a reduction
of about 70 % in 2020 and 80 % in 2050 can be achieved. This means that halving the energy use in 2020
seems possible, but reducing the energy in 2050 with 90% cannot be achieved with the options presented. To
achieve a 90% energy reduction of passenger transport, mobility demand per person should be halved. Several
options can contribute to such a decrease, for example by changing spatial settings, reducing commuter travel
or introducing telecommuting at home. Changes in activity patterns may also contribute to a reduction of the
individual mobility demand. Next to that, a shift to an increased use of soft modes is a possibility to reduce the
energy consumption of transport.

Implementing growth of mobility demand
Next to the relative energy reduction potential, the absolute potential can be calculated. For doing so, scenarios
should be used which describe the developments in passenger mobility. Two scenarios are used, which are described
in more detail in (Ybema et al., 1997). The first scenario, called Rational Perspective (RP) is an ecologically
driven scenario. New technologies are able to penetrate quickly due to a policy shift which facilitates
implementation. Transportation growth is limited, and a shift to increased use of public transport is assumed.
In the Market Drive (MD) scenario, the market mechanism only allows new technologies to compete on prices.
Consumption grows more rapidly as in the RP scenario, and so is the mobility demand. No modal shift changes
are assumed. Table 5 gives an overview of the growth in mobility demand for the two scenarios.
Table 5. Growth in mobility demand for two scenarios (1990 = 100)
Mobility demand Rational Perspective Market Drive
2020 2050 2020 2050
Passenger car kilometres 142 190 175 240
Bus kilometres 181 328 135 182
Train kilometres 243 589 135 182
Total mobility demand 151 226 170 232
Source: (Schol and Ybema, 1998)
With these mobility demand figures, the reduction percentage in 2020 and 2050 can be recalculated. Table 6
shows the results of this calculations for category 1 and 2 options.
Table 6. Reduction percentages compared to 1990 in 2020 and 2050 with category 1 and 2
improvement options and increased mobility demand
Year Rational Perspective Market Drive
Energy use Change
Energy use Change
(1990 = 100)
(1990 = 100)
1990 6.8 EJ 100 6.8 EJ 100
2020 6.5 EJ 95 7.6 EJ 111
2050 5.3 EJ 77 5.5 EJ 81
Table 6 shows calculations in which it is assumed that the fuel cell passenger car is totally penetrated in 2050.
In 2020 this technology is not yet available. With an assumed average ERE value of 1.7 MJ/MJ, the electric vehicle
has a share of 70 % in the total individual mobility supply. The increase in fuel economy is counteracted
by the growth in mobility demand. With technological options, only a reduction of about 20% by 2050 is possible.
The situation in 2020 about equals the situation in 1990, because all efficiency improvements are neutralised
by the increased mobility demand.
In table 7, the non-technological options are also implemented. The most important non-technological options
are the doubled occupancy rate and the 850 kg passenger car. A change in modal split is included in the scenarios.
Table 7. Reduction percentages compared to 1990 in 2020 and 2050 with category 1, 2 and 3
improvement options and increased mobility demand
Year Rational Perspective Market Drive
Energy use Change
Energy use Change
(1990 = 100)
(1990 = 100)
1990 6.8 EJ 100 6.8 EJ 100
2020 3.2 EJ 46 3.4 EJ 50

2050 2.9 EJ 42 2.5 EJ 37
Table 7 shows the same effects as table 4. Because of the high occupancy of individual transport systems, the
public systems score worse than the individual systems. Therefore, a higher reduction potential is achieved in
the Market Drive scenario, in which a larger share of the transport demand is provided with individual transport
systems, than in the Rational Perspective scenario with the high public transport share. Once again, it
should be noted that this high occupancy of individual systems is more realistic in a system with a large public
transport share.
CO2 emissions related to passenger transport
The total energy saving potential for passenger transport also depends on external factors. The ERE value for
electricity is of great importance for the results. Next to that, improvement options in the manufacturing industry,
as well as in the GER value of materials could contribute to energy savings associated with transport. The
CO2 emissions of transport are closely related to the energy use for passenger transportation. As shown in figure
4, the carbon content of a energy carrier determines the CO2 emission.
Primary energy use
of carrier i
Figure 4. Calculation scheme of total CO2 emissions
* =
Carbon content
of carrier i
CO 2 emission
of carrier i
In order to achieve a 90% reduction in the CO2 emission, not only the energy use, but also the carbon content of
the energy carriers should be taken into account. This extra factor also offers an extra opportunity to reduce the
emissions.
Along with the ERE value of electricity, the carbon associated share of electricity varies. When the share of non
fossil fuels used for electricity increases, the ERE value for electricity decreases, and so is the associated carbon
content. Next to the improvement options mentioned in this paper, also changes to other fuel mixes like biofuels
or hydrogen are conceivable. This will not result in an energy reduction, but may result in a reduction of the
CO2 emission, provided that the new fuels have more favourable carbon contents or are produced sustainably
(with closed carbon cycle or out of electrolysis driven by photovoltaic electricity).
Regarding the reduction of CO2 emissions it may be concluded that analogue to the energy consumption an
80% reduction in 2050 is possible. With a change to other fuels this reduction percentage may be even higher.
In order to achieve a 90% reduction in 2050, the external factors influencing the CO2 emissions should be
halved.
Discussion and conclusions
The calculations in this paper show considerable energy saving potentials within the transportation system.
With only technological options, a 50% reduction is possible in 2020, and a 60 % reduction in 2050. When
non-technical options are added, requiring major behavioural adaptations, 80% reduction can be achieved in
2050.
Implementing the technological options will result in higher costs in the transportation sector, since the various
options require extra research and production facilities. Partly, these extra costs will be compensated by the fuel
savings achieved by the introduction of the new technology. In general, the first savings are the cheapest to realise,
more expensive options will only be implemented if the cheapest options are already in use. With the
subdivision made in this paper, this general decrease in costs is not so obvious. The options emphasising on
behavioural change generally have lower costs per kilometre. Doubling the occupancy rate about halves the
costs per passenger kilometre. So the implementation of the technological improvement options result in higher
costs, whereas the more categories of options are implemented, the lower the costs of transport will be.

Realising the 80% reduction potential by 2050 requires a considerable modification of the infrastructure.
Changes in infrastructure are not included in this analysis, but it may be expected that these changes bring
about substantial costs.
The achieved high energy consumption reduction in 2050 requires two changes of vehicle types. On the short
term, the shift will be made to the electric vehicle, while on the mid and long term fuel cell vehicles will become
the new standard. It is not very likely that on a 60 years term two major changes of vehicle type will occur.
The analysis of the transportation sector reveals clearly that both energy and material options contribute to the
total saving potential. Modifications in the car construction (originating from/affecting the material production
sectors) like the modified frame or the small passenger car offer a saving potential of 37% by 2050. Modifications
in the automotive system (originating from/affecting the energy production system) have a saving potential
of the same magnitude (39%). Options which make more efficient use of the transportation system (like
doubling the occupancy rate) have the largest saving potential with 51%. In order to achieve the maximal saving
potential in the transportation sector, both material and energy options should be included in the analysis.
The analysis made in this paper may possibly contribute to the current discussion about reduction factors. In
current discussions about sustainability the reduction factor approach is often mentioned. The in this paper calculated
50% reduction by 2020 equals a factor 2 reduction in this year, whereas the 90% reduction by 2050
equals a factor 10 reduction by 2050.
In the general discussion about sustainability also other reduction factors are proposed: factor 4 - e.g. by the
Wuppertal institute (Weiszäcker et al., 1997) - to be realised about the year 2025 and factor 20 - proposed in
the Dutch DTO program (DTO, 1997)- to be realised in the year 2040. The factor 4 is based on halving the
energy consumption at a doubled welfare level, this requires a reduction of 75 % in energy use per service. The
calculations in this paper show that such a reduction is possible when introducing a number of nontechnological
options. Without these options, the maximal reduction is a factor 2.
According to the Dutch DTO program (DTO, 1997) a reduction by a factor 20 is possible for the year 2040.
Three strategies are presented to realise this far reaching reduction: the introduction of renewable fuels, a new
transportation system, and the introduction of an external drive system for vehicles. The introduction of renewable
fuels comprises for example methanol derived from agricultural crops and wastes or hydrogen produced
with photovoltaic electricity. The new transportation system integrates individual and collective passenger
transport means, while the development of an external drive systems for vehicles reduces substantially the vehicle
weight to be transported. The DTO study does not present a detailed implementation scenario and does not
assess the relative contribution of each of these strategies, so a quantitative comparison with the DTO results
and our findings is not possible. We suppose that the DTO strategies of a new transport system and external
drive systems have about an equal effect (about factor 5) as the combination of options considered in this paper
(such as a change of the modal split, doubling of the occupancy rate, and the introduction of the small car and
the fuel cell or electric car). So DTO must assume an additional factor 4 reduction by means of renewable fuels
to attain a total reduction of factor 20. Such an reduction implies a 75% transition to carbon-free energy sources
and will have far reaching effects for the energy and the agricultural system.
The total transportation related energy use is roughly equally shared by passenger and freight transport.
Achieving a factor 2 c.q. factor 10 reduction in the total transportation sector therefore also requires a same reduction
potential in the freight transport. The calculation of the reduction potential of freight transport can be
made using the same methodology as for passenger transportation. Not all options for passenger cars can be
implemented for freight transport. In general, most options seem to have a smaller energy reduction potential
for freight transport than they have for passenger transport. For example, the modified frame results in about
twenty per cent energy reduction for passenger cars, but only in ten per cent energy reduction for trucks
(Bouwman and Moll, 1997). It seems that achieving an 80% energy use reduction in freight transport is much

more difficult to achieve than the same reduction in passenger transport. The factor 10 reduction potential will
not be very likely to be achieved.
Literature:
[1] Binsbergen, A. J. van, A. Erkens and B. Hamel (1994): Long-term energy efficiency improvement for
transport, technology assessments. TU-Delft, Department of Infrastructure
[2] Bouwman, M. E. and H. C. Moll (1997):Status quo and expectations concerning the material composition
of road vehicles and consequences for energy use. Groningen, the Netherlands: RuG/IVEM
[3] Centraal Bureau voor de Statistiek (1995): Onderzoek verplaatsingsgedrag 1994. Voorburg/Heerlen:
Centraal Bureau voor de Statistiek
[4] Centraal Bureau voor de Statistiek (1997):De mobiliteit van de Nederlandse bevolking in 1996.
Voorburg/Heerlen: Centraal Bureau voor de Statistiek.
[5] DTO (1997): Vision 2040 - 1998, Technology key to sustainable prosperity
[6] Den Haag: Interdepartementaal onderzoeksprogramma duurzame technologische ontwikkeling (DTO),
uitgeverij ten Hagen & Stam
[7] Houghton, J.T., Meira Filho, L.G., Bruce, J., Hoesong Lee, Callander, B.A., Haites, E., Harris, N., and K.
Maskell (eds.) (1995): Climate change 1994; Radiative forcing of climate change and an evaluation of the
IPCC IS92 emission scenarios. Cambridge: Cambridge University Press
[8] Hughes, P. (1993): Personal transport and the greenhouse effect. London: Earthscan publications Ltd
[9] Mulder, H.A.J. and W. Biesiot (1998): Transition to sustainable society. A backcasting approach to
modelling energy and ecology. Cheltenham: Edward Elgar
[10] OECD/IEA (1997): Energy balances of OECD countries 1994-1995. Paris: OECD, IEA
[11] RIVM (1992): The environment in Europe: a global perspective. Bilthoven, the Netherlands: RIVM
[12] Schol E. and Ybema J.R. (1998) (to be published): CO2 abatement in the Western European transport
sector. Petten: ECN
[13] Weizsäcker E. von, Lovins, A.B., and L. Hunter Lovins (1997): Factor four: Doubling wealth, halving
resource use. London: Earthscan publications Ltd
[14] Ybema, J. R., P. Lako, D. J. Gielen, R. J. Oosterheert and T. Kram (1995): Prospects for energy
technologies in the Netherlands. Volume 2 Technology characterizations and technology results. Petten:
ECN
[15] Ybema, J. R., P. Lako, I. Kok, E. Schol, D. J. Gielen and T. Kram (1997): Scenarios for western Europe
on long term abatement of CO2 emissions, Petten: ECN

CO2 Emissions of Metal Production Technologies in Relation to External
Factors.
B.W. Daniels and H.C. Moll
Center for Energy and Environmental Studies IVEM, Groningen University,
Nijenborgh 4, 9747 AG Groningen, The Netherlands
tel. +31 50 3634609; e-mail B.W.Daniels@fwn.rug.nl
Abstract
Enhanced radiative forcing (greenhouse effect), caused by antropogenic CO2 emissions, is considered a major
environmental problem for the 21st century. This paper discusses the possibilities to reduce the specific CO2
emissions of the base metal industry (represented by steel and aluminium, largest by production volume and
emissions) to 50% and 10% of the current values. By process analysis, the paper investigates the CO2 emission
reduction potential for important current and future production routes of steel and aluminium, for some scenarios
representing external parameters. Subsequently, the paper investigates the effects of recycling, applying
the technologies with the lowest emissions for each scenario. The reduction potential of metal production
strongly depends on the external parameters. Favourable external conditions greatly enhance the reduction potential
of production, and enable the desired reductions of CO2 emissions.
Introduction
The present western Europe 7 consumption of energy and materials and the accompanying emissions of greenhouse
gases substantially contribute to global resource depletion and global climate change. To contribute to
global sustainability (requiring global stabilisation at 1990 emission level by the year 2050, according to scenario
IS92a [13]), while leaving room for development outside the western world, substantial reductions of
greenhouse gase emissions in the western world are required (mainly CO2). Currently, the total production of
metals contributes for about 7%, or 250 Mt/yr, to the total CO2 emissions in western Europe. The total emissions
are likely to grow in the future, as aluminium consumption will probably grow substantially in the next
decades, and for steel a gradual growth is foreseen.
This paper will systematically investigate the emissions of current and future metal production technologies, in
relation to external factors. It will investigate whether it is possible to attain 2 and 10-fold reductions of CO2
emissions per tonne metal with these technologies.
An equitable distribution of CO2 emissions on a global level in 2050 requires a reduction to one fifth of the current
energy consumption and emission levels for OECD Europe [20]). It is the long term objective with which
to compare the reduction possibilities. Taking into account the expected growth of production, this implies a
ten fold reduction per tonne of metal. A twofold reduction per tonne of metal been is the intermediate goal,
which also indicates the efforts required to keep total emissions constant.
The paper discerns options that can be freely chosen by the individual companies and options for which the implementation
depends on external factors, as they require external infrastructure. For a number of scenarios, the
paper will present cost structure and CO2 emissions of selected metal production routes, of the subsectors iron
and steel, and aluminium. A general analysis, included in the discussion, focuses on the reduction potential of
the sector as a whole, taking into account changes in recycling percentages. This section will also compare the
possibilities for emissions reduction with the factor 2 and 10 reduction objectives. Finally, the paper will shortly
discuss the reduction potential not included in the present analysis.
Methodology: Production Routes and Scenarios
The costs and emissions analysis included 8 production routes for steel and 3 for aluminium, all having the liq-

uid metal as the product; it did not include the subsequent, product-specific, less energy intensive processes for
casting, rolling and finishing, mainly important for their material efficiency.
Calculations of emissions and costs of the production routes took place by process analysis and aggregation of
the individual processes into production routes. First, individual process data were collected, including consumption
and production and data on investment and operation costs [1-9,11,12,14-18,21-23]. Stacking of the
processes into production routes allowed calculation of the consumption and production data per unit product.
Subsequent combination of these data with prices [9], carbon contents and CO2 emission factors resulted in
costs and CO2 emissions for each product.
As always in process analysis, system boundaries presented a problem. As all selected production routes use
electricity or produce excess energy, the CO2 emissions to be attributed to the production routes depended on
the technology applied outside the base metal industry. Therefore, the emissions were calculated for two CO2
emission factors of the electricity production, of 0 and 0.1 tonne CO2 per GJe. Nuclear, solar, wind and water
based power plants or fossil power plants with CO2 removal results in a (near) zero emission factor, while 0.1
tonne CO2 per GJe represents the current European average. Conversion of excess energy emerging from a production
route to electricity at an efficiency of 40% allowed subtraction of the avoided CO2 emissions in the
electricity sector from the emissions by steel production. Steam production may also have varying emissions.
There is no separate scenario: 0.06 t CO2 per GJ steam emission factor accompanies the high electricity emissions
factor; zero emissions for steam accompany the zero electricity emissions factor.
For steel, two additional, partly external parameters were used to define scenarios. For both scrap addition and
CO2 removal two additional scenario parameters were defined, resulting in a total of 8 scenarios. The steelmaker
decides on scrap addition, but the possibilities for scrap addition strongly depend on external factors.
Likewise, CO2 removal is a technique applied by the steel producer, but transport and storage require external
infrastructure, involving external decision makers.
Steel Production
The selection of steel production routes, listed in table 1, represents the four main types of liquid steel production,
namely the blast furnace, smelting-reduction, direct reduction and scrap based routes, with two of each. In
the EC, 1993, the conventional blast furnace route accounted for about 88 million tonnes crude steel, and scrap
based electric steelmaking for about 44 million tonnes [10], with application of best practice techiques resulting
in over 150 and 9 Mt CO2 emissions, respectively.
Table 1 Important characteristics of the production routes
Route Energy input Energy output Number of
processes
CO2 removal
Proces Type GJtot * GJe* Gas* Steam* Total CO2 rate ECU/t
removal** % CO2***
BF150 coal 19.6 3.1 5 1 (p) 80 32
BF250 coal 16.7 2.5 5 1 (p) 90 30
COREX coal 28.3 10.8 4 1 (p) 90 30
CCF coal 18.8 3.5 5.5 3 1 (p) 90 30
MIDREX gas, electricity 14.3 1.6 3 1 (s) 90 10
Circofer coal,
electricity
16.1 1.6 3 1 (s) 90 10
ACEAF electricity 1.7 1.4 1
DCEAF electricity, coal 1.7 0.9 1
* Values for low scrap addition, GJ per tonne steel. **(p) indicates that removal is possible, (s) that removal is part of
tion. *** total of removal, transport and storage.
normal process opera-
A thorough description of all processes in detail would be too elaborate, but it is essential to provide some
background information on the analysed production routes. Table 1 lists some characteristics of the the produc-

tion routes that are important for the emissions. Figure 1 gives a schematical representation of each production
route.
The air blast furnace route represents the current state of the art primary steel production The oxygen blast furnace
is in an experimental stage. Currently, several COREX plants are commercially operational; CCF still is
in an experimental stage. MIDREX is the largest direct reduction process by production volume, while Circofer
is on the edge of commercial introduction.
The scrap based production routes use the electric arc furnace to melt and process steel scrap. The AC furnace
[9] represents the major part of existing production capacity. The DC furnace and comparable EAFs are commercially
operational, but the AC furnace still is by far the most common electric steelmaking process. Due to
contaminants in the scrap, scrap based steel seldom is a fully fledged alternative to primary steel.
Scenario results
Figures 2 and 3 present the costs and emissions for the base case, assuming 0.1 tonne CO2 per tonne of GJe
generated, 12% scrap addition in primary oxysteel processes, zero scrap addition in DR routes. and no CO2 removal.
Following figures will show the influence of the emission factor of the electricity sector and CO2 removal.
Declines in the electricity emissions factor have double effects. The emissions of electric steelmaking decrease,
while the net CO2 emissions of the oxygen steelmaking routes increase, as the emissions avoided by converting
excess energy to electricity decrease. Especially CCF and, even more so, COREX, with their large excess energy
productions are at a disadvantage, while the blast furnace routes are hardly affected. The DRI and scrap
based routes profit from the low emission factor of electricity generation; the AC EAF route even has near zero
emissions.

Blast furnace routes
Smelting reduction routes
Direct reduction routes
Scrap based routes
1 2 1 2
5* 4 6*
4
3
7*
4 8*
4
11 11
3
Air blast furnace route Oxygen blast furnace route
11 11
COREX route CCF route
Midrex route Circofer route
9* 10*
12 12
AC EAF DC EAF
12 13
CO2 removal [11] is only attractive in processes that produce gas volumes with sufficiently high CO2 concentrations.
The figures 4 and 5 display emissions in case of CO2 removal. CO2 removal amount 80 % from the
air-based blast furnace at 32 ECU/t CO2 and 90 % from the oxygen based blast furnace, COREX and CCF at 30
ECU/t (see table 1). CO2 removal takes place by the water gas shift reaction which converts most carbon to
CO2, and is followed by removal of the CO2. The overall removal processs consumes steam and electricity.
MIDREX and Circofer already include CO2 removal; additional costs of compression, transport and storage
amount to 10 ECU/t CO2.
4

From figures 4 and 5, CO2 removal evidently emerges as a key technoloy in achieving emission reductions
within primary production routes. However, the emission factors of electricity and steam are very important for
the ranking of the primary
routes.
As shown in figure 4, with
emission factors of 0.1 for electricity
and 0.06 for steam, Corex
is the outright winner, thanks to
its large excess energy production,
which in this scenario prevents
CO2 emissions by the
electricity generation. Of the
primary routes, CCF has second
lowest emissions, followed by
MIDREX and Circofer. The
blast furnace routes profit least
from the removal, due to the
large share of processes without
removal, and the small excess
energy production.
With zero emissions for electricity and steam (figure 5), the picture changes profoundly. Going from Circofer,
MIDREX, CCF to COREX, the emissions rise, the exact opposite of the previous ranking. CCF and COREX do
no longer profit from their excess energy productions. The ranking of the blast furnace routes does not change,
they have the highest emissions. Overall this scenario offers the highest possibilities for emissions reduction,
without in any way compromising the steel quality by the addition of extra scrap. With regard to the costs, CCF
is the outright winner in all included scenarios.
Tables 3 and 4 give an overview
of the emissions and costs of the
production routes in the various
scenarios, including scenarios
with high, 30%, scrap addition
in primary production routes,
not shown in the figures. When
slight decreases of steel quality
are acceptable, further reduction
of emissions results from maximum
scrap input.
tonne CO2/tonne liquid steel
ECU/tonne
2.5
2
1.5
1
0.5
0
200
180
160
140
120
100
80
60
40
CO2 emissions of steel production
0.1 t CO2/GJe, low scrap
bf 150 bf 250 corex ccf midrex circofer ac eaf dc eaf
Production route
Costs of steel production
Low scrap
bf 150 bf 250 corex ccf midrex circofer ac eaf dc eaf
Production route

The table allows an overall view
on the influence of the chosen
scenario on the CO2 emissions.
The emissions of primary production
routes decrease strongly
with CO2 removal. COREX is
the most sensitive, also with regard
to the electricity emission
factor. The table confirms the
role of CO2 removal as a key
technology for steel production
with low CO2 emissions.
Though the influence of maximum
scrap addition is clear, it
does not change the ranking of
the production routes with regard
to emissions. Maximizing
scrap addition reduces the emissions
of the primary routes by
about 20 to 30 %, as table 3
shows.
With regard to costs, COREX again is very sensitive to CO2 removal. The CCF route always is cheapest,
though its lead diminishes in the removal scenarios. High scrap addition results in slight increases in costs in
the cheap production routes, and
decreases in the more expensive
routes and, in some cases, in
routes which have become more
expensive by CO2 removal. In
all scenarios the electricity price
is the same, however, a higher
electricity price in the scenarios
with a low electricity emission
factor is very likely and would
harm the competitiveness of the
routes using EAF and of the
routes with high CO2 removal
rates, such as COREX. On the
other hand, profits from excess
energy production would increase.
tonne CO2/tonne liquid steel
1
0.8
0.6
0.4
0.2
0
bf 150 bf 250 corex ccf midrex circofer ac eaf dc eaf
Production route
tonne CO2/tonne liquid steel
1
0.8
0.6
0.4
0.2
0
CO2 emissions of steel production
0.1 t CO2/GJe, low scrap
CO2 emissions of steel production
0 t CO2/GJe, low scrap
bf 150 bf 250 corex ccf midrex circofer ac eaf dc eaf
Production route

Table 3 Emissions per tonne liquid steel for the production routes in scenarios with regard to scrap use, CO2
removal and the electricity emission factor
scrap removal CO2/
GJ
BF150 BF250 COREX CCF MIDREX CIR ACEAF DCEAF
L no 0.1 1.73 1.43 2.24 1.41 0.91 .52 0.16 0.17
L no 0 1.78 1.46 2.49 1.62 0.67 1.25 0.01 0.07
L yes 0.1 0.88 0.61 0.13 0.24 0.32 0.34 0.16 0.17
L yes 0 0.82 0.45 0.29 0.21 0.08 0.07 0.01 0.07
H no 0.1 1.36 1.17 1.77 1.12 0.70 1.14 0.16 0.17
H no 0 1.40 1.16 1.97 1.28 0.48 0.90 0.01 0.07
H yes 0.1 0.69 0.47 0.1 0.19 0.28 0.29 0.16 0.17
H yes 0 0.65 0.35 0.23 0.16 0.06 0.05 0.01 0.07
In the zero electricity emission factor scenarios, the emissions for steam generation are zero, too.
Table 4 Costs in ECU per tonne liquid steel for various scenarios with regard to scrap use and CO2 removal.
scrap removal BF150 BF250 COREX CCF MIDREX CIR ACEAF DCEAF
low no 155 142 149 124 197 165 173 174
low yes 176 162 193 152 200 170 173 174
high no 157 147 152 133 203 179 173 174
high yes 174 163 187 155 204 183 173 174
Table 5 Aluminium production routes
Type Production route Included processes
Primary Hall-HJroult Mining; Bayer-process; Hall-HJroult
Bipolar cell Mining; Bayer-process; Bipolar cell
Scrap based Scrap melting Scrap melting (& refining)
Aluminium Production
The aluminium production routes include two primary production routes and one scrap based. The primary
production routes include mining, alumina production [1,2] and electrolysis of alumina to aluminium. The
Hall-HJroult based electrolysis [1,2] uses carbon based electrodes, the consumption of which results in additional
CO2 emissions 8 . The bipolar cell [5] has the advantage of inert electrodes, which apart from the absence
of additional CO2 emissions, allow a far more efficient plant design. However, implementation of the bipolar
still faces severe technical problems. The scrap based production applies several energy sources for smelting of
the scrap.

Scenario results
The high electricity consumption for primary aluminium production results in a strong influence of the
electricity generation on the CO2 emissions, as shown by figure 6. Thanks to the almost complete absence of
process CO2 emissions, the bipolar electrode cell benefits more from a low electricity emission factor than Hall-
HJroult does. The difference between primary and secondary routes is much smaller in case of a low electricity
emission factor.
A shift from the current
Hall-HJroult cell to the Bipolar
cell, accompanied by a zero
electricity emissions factor,
would mean an eight-fold reduction
within primary production.
In the same circumstances,
a shift to secondary production
means a thirty-fold reduction of
CO2 emissions.
Reduction Potential of the Sector
The intermediate factor two reduction
goal is attainable for
steel, even with unfavourable
external conditions. The
CO2/t
8
6
4
2
0
Aluminium production
CO2 emissions
hall_heroult bipolar secondary
Production route
0 t CO2/GJe
0.1 t CO2/GJe
MIDREX based production route nearly realises 50 % reduction relative to the standard BF150 route, with a
high electricity emissions factor and low scrap addition; high scrap addition further lowers emissions, beneath
the 50 % goal. A factor 2 reduction in aluminium production is only attainable with a low electricity emissions
factor.
On the long term, steel production can meet the factor 10 reduction goal, if external parameters are favourable.
CO2 removal is essential to attain the factor 10 reduction without changing the ratio between primary and scrap
based production routes. The electricity emission factor and the scrap addition rate may be helpful for further
reduction of emissions, but their influence is mainly to change the relative order of the production routes. On
the long term the required changes of the infrastructure to enable CO2 storage are very probably attainable,
though a social acceptation of the storage in aquifers and empty natural gas fields may raise difficulties.
Compared with steel, aluminium lags somewhat behind; within the primary production routes a factor 8 is attainable
with a low electricity emissions factor. Further reductions of emissions require a shift to scrap based
production. If this is not attainable, extra reductions in steel production may compensate the deficiency in aluminium.
Besides external factors, internal factors are likely to influence the choice of a production route. Of the production
routes with considerable reduction potentials, COREX and CCF based routes present a logical step for
primary steel producers, while DRI based routes are entirely different from the current primary production
structure. On the other hand, DRI based routes might be a logical step for the current scrap based producers.
MIDREX, Circofer and CCF could be regarded as Asafe bets@ as they always result in emission reductions,
regardless of external parameters, while COREX requires CO2 removal to realise reductions at all. The CCF
route has the advantage of its low costs, even in the removal scenarios.

Table 6 CO2 emissions per tonne liquid steel for scenarios with regard to scrap use, emissions for power generation,
and CO2 removal, compared with those of current primary production.
removal CO2/ scrap BF150 BF250 COREX CCF MIDREX CIR ACEAF DCEAF
no
GJe
0.1 low 1.00 0.83 1.30 0.82 0.52 0.88 0.10 0.10
no 0.1 high 0.79 0.68 1.02 0.65 0.41 0.66 0.10 0.1
no 0 low 1.03 0.85 1.44 0.94 0.39 0.73 4E-3 0.04
no 0 high 0.81 0.67 1.14 0.74 0.28 0.52 4E-3 0.04
yes 0.1 low 0.51 0.35 0.08 0.14 0.18 0.20 0.10 0.10
yes 0.1 high 0.40 0.27 0.06 0.11 0.16 0.17 0.10 0.10
yes 0 low 0.47 0.26 0.17 0.12 0.05 0.04 4E-3 0.04
yes 0 high 0.37 0.21 0.13 0.09 0.04 0.03 4E-3 0.04
A low emission factor for electricity invariably implies a low emission factor for steam generation. A bold lettertype
indicates the lowest emissions primary production route with the lowest emissions in a scenario. Light
grey shading indicates an emission reduction factor between 2 and 10, compared with the base case BF150,
dark grey shading indicates an emission reduction factor higher than 10
Because of the uncertainty in
the reduction potential within
the sector and its dependence on
(partially) external parameters,
it is still important to pay attention
to other possibilities to reduce
the emissions of the base
metal sector. Moreover, changes
outside the sector may help to
reduce the necessity of expensive
measures within the sector.
Increase of recycling offers important
possibilities to reduce
the throughput through the energy
extensive primary production
routes.
tonne CO2 per tonne steel
2
1.5
1
0.5
CO2 per tonne liquid steel
Scrap based steel and emissions
0
0 0.2 0.4 0.6 0.8 1
Share of scrap based production
0.1 t CO2/GJe, no removal
0 t CO2/GJe, no removal
0.1 t CO2/GJe, removal
0 t CO2/GJe, removal
Current production
Current emissions level
Figures 7 and 8 show the relation
between the share of scrap
based production and the average
CO2 emissions per tonne liquid steel and aluminium, respectively. Each line represents the average emissions
per tonne of liquid metal in relation to the share of scrap based production. The primary and scrap based
production routes have the lowest emissions for each senario.
For steel, and even more so for aluminium, scrap based production offers important opportunities to reduce
emissions. For steel a reversal of the current two thirds-one third ratio between primary and scrap based production
would reduce the emissions by nearly a half, using current production technology. For aluminium, the
reductions by scrap based production are even larger.
However, very high recycling rates may require adjustments outside the base metal sector, because of the generally
lower quality of scrap based metal. Current application of scrap based steel production is probably limited
by an absolute scrap shortage, as Europe exports substantial amounts of scrap. In case of aluminium, the strong,
continuing growth of consumption results in an absolute shortage of scrap; high recycling rates are only
achievable after stabilisation of aluminium consumption. Then, new refinement technologies for scrap based
Factor 2
Factor 10

aluminium may allow production of high quality scrap based aluminium, be it with probably higher specific
emissions than those of current secondary production.
Conclusions
The results clearly indicate that, even with high emission factors for electricity, the factor 2 and even the factor
10 emission reductions are readily attainable in the steel sector, when CO2 removal is applied. In this case,
changes in the ratio between primary and scrap based production are not necessary; in some scenarios, some
primary production routes have even lower emissions than scrap based production routes. It is also important to
note that for the realisation of the factor 10 emission reduction, it is not necessary to call for processes that are
still in an experimental stage.
The reduction potential of primary aluminium production lags somewhat behind, but with a low emission factor
for electricity, a factor 8 reduction is possible, by a shift to the new bipolar electrode process. A moderate
increase of recycling would further decrease the emissions of aluminium production. If this would not be possible,
extra reductions in the steel sector could compensate.
The influence of external factors, especially that of the possibilities for CO2 removal, is strong, not only on the
total emissions reduction potential, but also on the choice of the production route. The choice for the best production
route requires extensive knowledge of the developments outside the base metal sector.
Because of this sensitivity of the results for developments outside the sector, and the fact that CO2 removal is
imperative for attainment of the factor 10 reduction in primary steel, it is advisable to investigate the possibilities
for increasing scrap based production. If a shift to secondary production is possible, the reduction potentials
for steel, and especially aluminium, further increase.
Two developments, mainly driven by forces outside the (primary and secondary) metal sector may further influence
the sector thoroughly, namely dematerialisation and the implementation of integral cycle management.
Both trends may substantially decrease the demand for primary produced (ferrous) materials and may also decrease
the related CO2 emissions.
Dematerialisation consists out of several phenomena, such as miniaturisation, substitution by light-weight materials,
resulting in a net decrease of the material intensity of the Gross Domestic Product. Especially the ferrous
metals production sector has already been affected by the dematerialisation trend. Therefore, the growth
rates in this sector have been lower than the GDP growth rates. A high dematerialisation trend in the future
may even stop the growth in the ferrous metals production sector, while it may favour the aluminium and plastic
production sectors.
Integral cycle management aims at the closing of material cycles in the society and the prevention of material
losses. This strategy consists out of society measures to increase the quality of materials derived from post consumption
waste and industry measures to increase the content of secondary materials in the production sectors.
Examples are design-to-dismantle and design-for-recycle practices in industry, standardisation of materials for
specific applications, separate collection and treatment of usable products, parts and other material streams. A
systematic approach can result in an optimal material cascade that saves more than 50% of the primary inputs
and decreases the energy requirements for materials with a similar percentage without compromising the quality
of materials and products [19].
List of abbreviations
ACEAF Alternating Current Electric Arc Furnace
BF150 air Blast Furnace, 150 kg/tonne coal injection
BF250 oxygen Blast Furnace, 250 kg/tonne coal injection
BOF Basic Oxygen Furnace
CCF Cyclone Converter Furnace
Cir Circofer

DCEAF Direct Current Electric Arc Furnace, continuous charge
DR Direct Reduction
DRI Direct Reduced Iron
EAF Electric Arc Furnace

Literature
[1] Anthony Bird Associates (1990): Aluminium Production Costs 1990, 2- Plant Details. Anthony Bird
Associates, 1990
[2] Anthony Bird Associates (1990): Aluminium Production Costs 1990, 1- Main Report. Anthony Bird
Associates, 1990
[3] B`hm, C., Eberle, A. (1997): Latest developments in COREX 7 technology and New Project Experience.
VOEST-ALPINE Industrieanlagenbau Gmbh / Linz, Austria, February 1997
[4] Brascugli, G., Perni, G., Repetto, E. (1997): Evolution of the electric arc steelmaking route. In: MPT
International, February 1997, p. 62-83
[5] Burkin, A.R. ed. (1987): Production of Aluminium and Alumina, Critical Reports on Applied Chemistry,
Volume 20. The Society of Chemical Industry, 1987
[6] BUWAL (1996): koinventare fhr Verpackungen, Band I. BUWAL NR. 250/1. Bundesamt fhr Umwelt,
Wald und Landschaft (BUWAL), Bern, 1996
[7] BUWAL (1996): koinventare fhr Verpackungen, Band II. BUWAL NR. 250/II. Bundesamt fhr Umwelt,
Wald und Landschaft (BUWAL), Bern, 1996
[8] DaniNls, B.W, Moll, H.C. (1997): The Base Metal Industry: Technological Descriptions of Processes and
Production Routes; Status Quo and Prospects, IVEM OR 92. IVEM, Groningen, The Netherlands, 1997
[9] Eurofer (1997): Eurofer Survey, no publication
[10] Eurostat (1994): Iron and Steel, yearly statistics. ECSC-EC-EAEC, Brussels-Luxembourg, 1994
[11] Gielen, D.J., Van Dril, A.W.N. (1997): The Basic Metal Industry and its Energy Use, Prospects for the
Dutch energy intensive Industry. ECN, March 1997
[12] Husain, R., Hirsch, M., Eichberger, H. (1996): Circored and Circofer, The Case for Lurgi DR
Technologies in the Context of Market Developments ijn the Asian Pacific Region. Paper presented at
Metal Bulletin=s 4th Asian Iron and steel conference, Seoul, South Korea, 11 - 13 November 1996
[13] Houghton J.T. et al. (ed.), (1995) Climate Change 1994: Radiative forcing of climate change and an
evaluation of the IPCC IS92 emission scenarios. Cambridge University Press, 1995
[14] IISI (1996): Statistics on energy in the Steel Industry, 1996. Update. International Iron and Steel Institute,
Commitee on economic studies, Brussels, 1996
[15] IISI, (1987): Sinter and Pellets, production and Use, Capacities (State: 1987). International Iron and Steel
Institute, Commitee on raw materials, Brussels, 1987
[16] IISI (1993): World Cokemaking Capacity. International Iron and Steel Institute, Commitee on raw
materials, Brussels, 1993
[17] Lankford, W.T. Jr; Samways, N.L.;Craven, R.F.; McGannon, H.E. (1985): The Making, Shaping and
Treating of Steel, 10th edition. United States Steel, 1985
[18] Meijer, H.K.A., Flierman, G.A., Teerhuis, C.P., Bernard, J.G., Boom, R. (1994): The Cyclone Converter
Furnace (CCF), A Hot Metal Route Avoiding the Coal Carbonization and Ore Agglomeration Stages,
Ironmaking 2000, 18th Advanced Technology Symposium. October 2-4, 1994
[19] Hoogovens Groep BV Corporate Research Laboratorium, IJmuiden, The Netherlands.
[20] Moll, H.C., (1997): The Design of Material Cascades in the Economic System - Theoretical
Considerations and some Applications, pp215-220 From Paradigm to Practice of Sustainability.
Proceedings of the ConAccount workshop 21-23 January, 1997 Leiden, The Netherlands, published by
Wuppertal Institute for Climate, Environment, Energy
[21] Mulder, H.A.J., Biesiot, W., (1998): Transition to sustainable society; a backcasting approach to
modelling energy and ecology. Edward Elgar, Cheltenham, 1998
[22] Scott, D.H. (1994): Developments affecting metallurgical uses of coal. IEA coal research, IEACR/74,
London, September 1994
[23] VOEST-ALPINE (1997): COREX 7 Status Report. VOEST-ALPINE Industrieanlagenbau Gmbh, Linz,
Austria, February 1997
[24] VOEST-ALPINE: COREX, Revolution in Ironmaking. VOEST-ALPINE Industrieanlagenbau Gmbh,
Linz, Austria, DEUTSCHE VOEST-ALPINE Industrieanlagenbau Gmbh, Dhsseldorf

Verslag van de workshop commentaren en de forumdiscussie Factor 2/Factor 10, 2 April
1998, Utrecht
ECN-Beleidsstudies
Concept 15/4/1998
Verpakkingen
Referent dhr. Clement, VROM
Commentaar kan geleverd worden vanuit een wetenschappelijke of een beleidsmatige visie. De wetenschap gaat
uit van een model, maar de praktijk is anders. Het hergebruik van flessen, standaardisatie van flessen is bijvoorbeeld
in de praktijk bijna onmogelijk. De theorie moet daarom getoetst worden aan de praktijk. Het volgende
commentaar is een beleidsmatige visie op de analyse, waarbij niet zozeer voorop staat wat VROM wil,
maar wat gegeven de huidige politieke/maatschappelijke opvattingen mogelijk is gebleken.
Zelfregulering staat centraal in Nederland (via de convenanten-benadering). De opties in de analyse zijn technisch
wellicht haalbaar, maar zijn ze ook politiek haalbaar? Is de overheid/de politiek bereid om bijvoorbeeld
heffingen te leggen om materiaalsubstitutie te bereiken? In 2020 wil men 350 kt verpakkingen besparen, dat is
een reductie van 20%. Kan men een dergelijk absoluut getal wel opgeven? Zijn er wel grote innovaties op dit
vakgebied ? De ervaringen van de laatste jaren tonen dat er weinig echte innovaties zijn ondanks een grote betrokkenheid
van alle partijen in het verpakkingenconvenant. Clement heeft geen optimistische kijk op substantiële
mogelijkheden. Marketing staat voorop bij het gebruik van verpakkingen. Het gaat daarbij primair om
zaken zoals imago, prijs, klandizie, etc. De inspanningen zijn juist gericht op een toename van de consumptie,
het gebruiksgemak van de consument staat voorop. Prijzen lijken slechts een ondergeschikte rol te spelen. Polycarbonaat
melkflessen zijn bijvoorbeeld duurder, maar de klant koopt ze toch omdat hij het ‘mooi’ vindt. Ook
de maatschappelijke ontwikkelingen leiden tot een hogere consumptie van verpakkingsmaterialen: er komen
kleinere huishoudens door vergrijzing en door de tendens eerder zelfstandig te gaan wonen. De ontwikkeling
bij verpakkingsmaterialen is in de richting van meer CO2-intensieve kunststof en minder herbruikbare verpakkingen.
De overheid is alleen in staat om deze ontwikkelingen tegen te houden wanneer er een breed politiek
draagvlak is voor harde maatregelen. De conclusie is daarom dat een factor 2 technisch wel haalbaar lijkt, maar
beleidsmatig/politiek moeilijk te bereiken is.
Wat zijn de nieuwe elementen in deze studie ?
Hekkert: De studie laat zien dat er heel veel mogelijk is op het vlak van efficiënter gebruik van verpakkingsmateriaal
in Europa. De studie richt zich niet op het te voeren beleid c.q. of er overheidsingrijpen
gewenst/mogelijk is. Echte innovaties zijn niet eens echt nodig. Grote besparingen kunnen bereikt worden met
het doorvoeren van bestaande concepten (product hergebruik, materiaal substitutie).
Is er rekening gehouden met de barrieres die de heer Clement opvoert ?
Hekkert: De doelstelling van het model is optimalisatie, geen simulatie. De opties worden in het MARKAL
model ingevoerd. Het model kan vervolgens uit deze opties kiezen voor het realiseren van een emissiedoelstelling.
De opties zijn technologisch van aard. Ze schetsen dus ook alleen wat er technisch mogelijk is. Het
model moet uitmaken wekle opties interessant zijn wanneer je de exogene kosten van CO2-emissies in rekening
brengt. Verder voert de VU binnen het Matter-project nog een deelstudie uit naar implementatiebarrières, die in
een later stadium aan het model zullen worden toegevoegd. Voor wat betreft de CO2-intensieve kunststoffen:
juist deze materialen lenen zich goed voor product- en materiaalrecycling, waardoor de CO2-emissie per eenheid
verpakking sterk gereduceerd kan worden.
Clement: Vanaf 1991 is beleid gevoerd, het bedrijfsleven voelt dat het geld gaat kosten. Wat nu gehaald is was
relatief makkelijk.

Worrell: Stel dat het klimaat beleid wringt, dan gaat de bereidwilligheid van de partners omhoog om iets te veranderen
aan het materiaalgebruik. Bestaande engineering rules leiden vaak nog tot overmatig materiaalgebruik.
Kleine bedrijven met kleine R&D afdelingen leiden ertoe dat bijvoorbeeld veel boterkuipjes overdesigned
zijn. Er zijn toch nog wel grote mogelijkheden tot verdere besparingen.
De heer Clement heeft indruk dat het bedrijfsleven al veel heeft gedaan en verwacht niet veel substantiële veranderingen,
behalve als men innovaties bedenkt. Twijfelt of het potentieel groot is.
Worrell: De aandacht voor het verminderen van verpakkingsmateriaal dateert vanaf het verpakkingenconvenant
1 in 1991. Dit is een relatief korte periode van 7 jaar, als je het vergelijkt met termijnen die in het algemeen
nodig zijn bij het bedenken en doorvoeren van innovaties. Hierbij denk je al snel aan een periode van 20
jaar. Voor verpakkingen zijn grote innovaties daarom niet uitgesloten.
Brezet: Bij grote bedrijven is men nu pas aan het werk om databases over verpakkings-innovaties tussen
bedrijfsonderdelen uit te wisselen. Als je bedenkt dat het daarbij om innovatieve bedrijven gaat dan geeft dat
aan dat er nog wel veel mogelijk is.
Bergsma: Vraagt of CO2 heffing zou helpen? En stelt nog een vraag aan VROM: Zou een alternatief voor het
poldermodel heffingen zijn?
Hekkert: Als in het model een CO2-tax wordt ingevoerd dan zullen duurdere maatregelen interessant worden.
Welke verschuivingen precies op zullen treden is vooraf moeilijk te voorspellen.
Gielen: Uit de voorlopige resultaten blijkt dat het aandeel kunststof-verpakkingen toeneemt in de t.g.v. de goede
prijs/milieuperformance verhouding
Zijlstra: Gelooft niet in heffingen, vindt het markant dat hij bij het bedrijfsleven vrijwel geen verschil ziet met
heffingen.
Clement: De verpakking is een klein deel van de totale prijs, heffingen vormen daarom geen incentive.

Transportmiddelen
Referent dhr. Smokers, TNO-wegtransportmiddelen.
De analyse is gebaseerd op back-of-the-envelope berekeningen. Daarbij zijn een aantal fouten gemaakt
(bijvoorbeeld bij de CVT). De analyse is gebaseerd op de data uit de Syrene studie van enkele jaren geleden. De
technieken die daarin beschreven zijn lopen inmiddels een jaar of 5 achter. Er is de laatste jaren veel vooruitgang
geboekt op technologisch gebied. Het verbeterpotentieel van de conventionele auto wordt in de analyse
onderschat. Kijk bijvoorbeeld naar de 3 liter auto die momenteel in ontwikkeling is. De ontwikkelingen bij
brandstofcellen lijken veel sneller te gaan dan voorspeld. Een ander voorbeeld is de snelle ontwikkeling van de
hybride auto (combinatie van aandrijftechnieken). Al deze ontwikkelingen kunnen wel tot een factor 4 verbetering.
De uitkomst ten aanzien van factor 2/factor 10 hierdoor trouwens niet beïnvloed; de correcties naar
boven en naar beneden heffen elkaar op.
Ten aanzien van de kosten moeten de conclusies genuanceerd worden. MARKAL is een macro economisch
model. De ontwikkeling van de kwaliteit van de vraag is exogeen en dus niet aan te passen. Consumenten kunnen
in de toekomst echter veel meer besteden. In 2050 mogen we verwachten dat de consument tussen de f
100.000 en f 200.000 voor een auto zal besteden. Deze prijs biedt grote mogelijkheden om ook veel meer geld
te besteden aan milieumaatregelen. We moeten nu niet terugschrikken voor hoge kosten voor milieumaatregelen
op zeer lange termijn.
Referent dhr. Annema, RIVM
1. Is de analyse van de huidige en toekomstige emissies en/of het huidige en het toekomstige energiegebruik
naar uw mening reëel?
Het artikel is duidelijk geschreven vanuit oogpunt van reductie potentieel. Implementatieproblemen om het potentieel
te bereiken zijn bewust (zie p.2) er buiten gelaten. In die zin zijn de uitkomsten dan ook niet reëel. De
auteurs concluderen zelf dat het vrij onwaarschijnlijk is dat het door hun geschetste maximale potentieel verwezenlijkt
zal worden. Het zal in ieder geval zeer moeilijk worden.
2. Zijn er belangrijke technologische opties ter reductie van broeikasgasemissies over het hoofd
gezien?
Voor personenauto’s zijn de meeste nu ‘bekende’ opties mijns inziens meegenomen. Het enige wat mist bij personenauto’s
zijn: gebruik van biomassa, lagere snelheid en voertuiggeleiding (zie DTO-studie). Wat ik mis bij
andere vervoerwijzen is een beschouwing over het collectieve transport: ook daar zijn waarschijnlijk technologische
opties die tot energiereductie kunnen leiden en die dit soort vervoer aantrekkelijker maken, met name in
het stadsgewestelijke vervoer (mogelijk deels ondergronds). Volledig buiten beschouwing gelaten is het
vliegverkeer; toch een zeer belangrijke vervoerwijze (waarschijnlijk met name in 2020 en 2050) als je naar geheel
Europa wilt kijken. Vrachtverkeer wordt geaddresseerd, niet uitgewerkt, wat ik begrijp want het is een
lastig onderwerp. Misschien een keer een aparte studie hiernaar?
3. Wat vindt u van de schattingen voor het emissiereductiepotentieel en kosten van de individuele
opties?
• Bij de schattingen van het emissiereductiepotentieel van de opties heb ik een paar opmerkingen:
• ik vind door het hele verhaal niet zo duidelijk waar de reducties betrekking op hebben. Bijvoorbeeld tabel 1:
ik neem aan dat het gaat om reductie in energiefactoren t.o.v. 1990. Graag dit soort zaken meer aangeven.
• Ik vind de verhouding tussen aandacht voor energiegebruik en CO2-emissies nogal scheef. Ik pleit voor
meer aandacht voor effecten op CO2-emissiefactoren en minder op energie, wat mijns inziens toch een
beetje een minder interessant aspect is. Het gaat dus ook om de vraag welke energiedragers in de toekomst
kunnen worden ingezet om ons voort te bewegen. Ik vind dat dit onderwerp er nogal bekaaid af komt.

• Ik vraag me af je altijd mag veronderstellen dat opties onafhankelijk van elkaar zijn: m.a.w. of de methode
van het vermenigvuldigen van de resterende fracties altijd toepasbaar is.
• Ik mis een beschouwing over de relatie tussen techniek en volume. Eenvoudig gezegd: wanneer auto’s
zuiniger worden, worden ze goedkoper in gebruik, waardoor een deel van de CO2-winst zal weglekken in de
vorm van meer verkeer.
• Een beschouwing over kosten ontbreekt nagenoeg. Wat er over wordt gezegd, is misschien wat al te kort
door de bocht: ik geloof niet dat als er meer van de voorgestelde opties ingezet wordt er een goedkoper mobiliteitssysteem
is (of ik begrijp het verkeerd). Er zijn enorme investeringen nodig in vervoermiddelen,
mogelijk een waterstofinfrastructuur, mogelijk andere weg- en railinfrastructuur, e.d. Je kan ook niet zonder
meer stellen dat opties die tot gedragsverandering moeten leiden lagere kosten met zich mee brengen: verlies
aan reistijden, verlies van vrijheid en/of comfort kunnen mijns inziens in een brede kostendefinitie wel
degelijk als ‘kosten’ worden beschouwd.
4. Klopt naar uw mening de conclusie dat een factor 2/10 al dan niet haalbaar zou zijn in de komende 25
jaar?
Het potentieel is er. Daar ben ik het mee eens. De paper spreekt zich niet echt uit of het ‘haalbaar’ is, maar
tussen de regels proef ik twijfel. Dat klopt, ik denk ook dat het twijfelachtig is. Het hoeft natuurlijk ook niet
echt haalbaar te zijn. De waarde van dit soort studies ligt om aan te geven welke wegen kansrijk zijn om tot
CO2-reducties te komen. Ik mis een beetje in de conclusies dergelijke richtingwijzers:
• aan de technische kant meer onderzoek en ontwikkeling (en mogelijk demoprojecten) op hybride, electrisch,
brandstofcel, etc.
• meer inzetten op beter relatie ruimtelijke ordening-beleid, infrastructuur (OV), e.d
Bouwman
1. Binnen de analyse gepresenteerd in de paper is de CO2 emissie van transport relatief onderbelicht. Ook
mogelijkheden om niet zozeer op energie te besparen, maar wel de emissies van CO2 terug te dringen,
komen niet uitgebreid aan de orde. Dergelijke verbeteringen vinden namelijk niet zozeer plaats binnen het
transportsysteem, als wel binnen het energiesysteem (bijvoorbeeld bij de overstap naar alternatieve
brandstoffen). Dergelijke interacties tussen de verschillende systemen worden juist met behulp van het integrale
MARKAL model duidelijk gemaakt.
2. De genoemde gedragsopties worden door de auteurs niet voorgesteld als meest veelbelovende en goed te implementeren
opties, maar worden genoemd om duidelijk te maken in welke orde grootte de maatschappelijke
aanpassingen liggen die nodig zijn om een factor 10 reductie binnen personenmobiliteit te kunnen realiseren.
3. De factor 10 benadering rekent met een reductiepercentage per gereden kilometer. Een deel van die reductie
zal verloren gaan door de mobiliteitsgroei. Binnen MARKAL wordt voor de diverse scenario’s een mobiliteitsvraag
gedefinieerd, waaraan door de inzet van voldoende transportmiddelen voldaan moet worden.
Deze scenario’s houden echter geen mogelijkheden met eventuele techniek-volume interacties.
4. Daarnaast zit er flinke groei in het vliegverkeer. Er zijn geen technologische verbeteropties gemodelleerd
voor het vliegverkeer. Bij een toename hiervan zal dit manco steeds zwaarder gaan tellen.
Tempelman
Ten aanzien van de gewichtsreductie: autobouwers mikken op gewichtsreductie in het frame van de auto. De
meeste andere onderdelen (95%) worden ingekocht. Dus niet alleen naar het frame (ongeveer 200 kg) kijken.
Eerst kijken naar de 800 kg. Hoe ga je al die toeleveranciers zo ver krijgen om bijvoorbeeld een goed belastingsmodel
te maken (bijvoorbeeld voor een stoel). Hij vraagt zich af of dit beleidsmatig te beïnvloeden is.
Bouwman
Binnen de in de paper gepresenteerde analyse is ook het indirect energiegebruik van mobiliteit meegenomen,
voor zover het de voertuigen betreft (infrastructuur wordt meegenomen in de deelstudie “gebouwde omgeving”).
Materiaalsubstitutie heeft invloed op dit indirect energiegebruik. Daarnaast wordt er in het model uitgegaan
dat de standaardauto in de loop van 1990 tot 2050 toeneemt in gewicht door upgrading.

Gebouwen en Infrastructuur
Referent dhr. Lanser, VNC/ENCI
1. Analyse van de huidige. en toekomstige emissies/energieverbruik
De gegevens van tabel 3 (productie, emissies zijn wat mij betreft moeilijk te verifieren. De twee genoemde
bronnen verwijzen niet rechtstreeks terug naar de bron van alle bronnen: de Europese branche-verenigingen
van de betreffende bouwmaterialen-producenten Dat is jammer, want deze organisaties houden allemaal jaarlijkse
productiestatistieken bij en doen dat op een erkend-betrouwbare manier
De cementproductie bedroeg in 1997 in EU+EFTA-landen 178 Mt. De bijbehorende CO2 emissies zijn 72 Mt
voor decarbonatatie van kalksteen, 49 Mt voor brandstofgerelateerde CO2 en 18 Mt t.g.v. electriciteitsgebruik
Bij elkaar 138 Mt.
Wat in tabel 3 onmiddellijk opvalt is dat er 17 bouwmaterialen worden genoemd en één halffabrikaat. En dat
ene halffabrikaat is nou net cement.
Als we in de tweede kolom van deze tabel in plaats van halffabrikaat cement het bouwmateriaal beton opvoeren,
dan is er weer eenheid in de eenheid. Maar dan praten we in plaats van over 185 Mt cement inmiddels
wel over 1200 Mt beton. Dat is zes keer zover als alle andere bouwmaterialen in tabel 3 tezamen
Na enige herberekening kan men uit deze gegevens ook de conclusie trekken dat 85% van een massastroom
(lees- beton) leidt tot minder dan 50% van de totale CO2
De resterende 15% van de totale bouwmaterialenstroom - niet zijnde beton - vertegenwoordigt de andere helft
van de uitstoot van kooldioxide.
Met andere woorden-. beton is helemaal zo gek nog niet
In de overzienbare toekomst zal in Westeuropa geen significante, afzetgroei voor cement meer plaatsvinden.
Dat is bijvoorbeeld wél liet geval in communistisch China.
De techniek is daar sterk verouderd. En de emissies navenant. De Volksrepubliek China produceert meer cement
dan West-Europa in zijn geheel. De CO2-emissie per ton cement is daar bijna twee maal zo hoog als in
West-Europa. Met de huidige stand der techniek ligt de factor twee in China dus gewoon voor het oprapen.
2. Volledigheid von de gepresenteerde technologische opties
Een verdienste van de heer Gielen is dat hij niet alleen de bouwmaterialenproductie beschouwt, maar ook alle
vóór- en achterliggende schakels in de productketen. Een ketenbenadering dus. Ook de opdeling naar 'minder'
en 'vervangen' is in principe een' verstandige, want doelmatige keuze.
De realiteit is echter aanmerkelijk ingewikkelder en weerbarstige dan een academisch technisch model kan
beschrijven, En dan spreek ik nog maar- niet over de vergaande technische en economische implicaties van de
betreffende opties.
Een deel van de opties in tabel 4 ziet ENCI daarom als tamelijk problematisch.
Anderzijds wordt een deel van de opties vandaag de dag door ons al in praktijk gebracht Ook is er nog een
aantal niet genoemde opties, dat tamelijk specifiek is voor onze tak van sport: de cement- en betonindustrie

Om nu wat meer gevoel te krijgen. voor zijn denkmodel en voor de context waar binnen wij moeten handelen,
zou ik de heer Gielen vanaf deze plaats van harte willen uitnodigen om hierover op een nader overeen te komen
datum, tijd en plaats met ons nader van gedachten te komen wisselen. Bij dezen dus!
3. Schattingen het reductiepotentieel en de kosten
Over potentieel kan men lang praten.
Als de politiek vandaag besluit dat reductie van CO2-emissies voor de komende twintig jaar duurzaam bovenaan
de agenda zal blijven staan dan kan de industrie zich daar op richten. In dat geval wordt het beschreven
reductiepotentieel waarschijnlijk redelijk uitgenut
Wel zal de overheid - goedschiks danwel kwaadschiks - een dramatische omslag in de aard van de bouwvraag
moeten bewerkstelligen. Ze zal zéker moeten ingrijpen in de bestaande marktwerking.
Als gedachte is de oplossingsrichting om beton door hout te vervangen bijvoorbeeld best reëel. Maar voor de
realisatie van die gedachte is enige kennis van de achtergronden waarom juist hout in de laatste honderd jaar in
de bouw geleidelijk maar wel - grotendeels is verdrongen door beton - dunkt mij - hoogst onontbeerlijk
Het gebrek aan technisch achtergrond is een schreeuwende omissie in het verhaal van de heer Gielen.
4. Haalbaarheid van de factor 2 en de factor 10
Met de voeten op tafel wil de Europese cement- en betonsector wel filosoferen over de mogelijkheden van een
reductie van de CO2-emissie niet een factor 2. En wie stelt dat het niet bij filosoferen mag blijven heeft natuurlijk
liet grootste gelijk van de wereld
Nu komt liet bedrijfsleven komt gewoonlijk pas echt in beweging als er haalbare -. doelstellingen worden geformuleerd
Laten we de factor 10 daarom voorlopig maar even vergeten.
Voor de Nederlandse cementindustrie ligt de situatie toch wat moeilijker
ENCI heeft zich vijf jaar geleden vrijwillig tot de overheid verplicht om in een periode van 11 jaar 20% energie-efliciency-verbetering
te realiseren.
ENCI heeft die extra efficiency-slag inmiddels grotendeels achter de rug. In 1997 produceerde ENCI cement
niet gemiddeld portlandklinkeraandeel van,48 %. Onze Westeuropese collega's zitten nog tegen de 80% aan..
De uitstoot van CO2 per ton Nederlands cement is alles bij elkaar weinig meer dan de helft van de gemiddelde
uitstoot van Europees cement. Dan zit er - helaas voor ons - ook nog eens relatief weinig cement in Nederlands
beton
Ten opzichte van liet Europese gemiddelde hebben wij de factor 2 dus al bijna te pakken. In ons eigen tuintje
valt daarom ook niet meer zoveel te halen,
Samen niet onze afnemers in de betonsector zijn we daarom drie jaar geleden een groot LCA-project begonnen.
Met LCA's kan je voor één of meer milieuparameters bijvoorbeeld de CO2-emissies, de gehele productketen
doorrekenen. Ook kan men de bijkomende effecten van bepaalde opties in kaart brengen, bijvoorbeeld een hoger
energieverbruik ten gevolge van de wens tot lagere emissies van NOx, Of omgekeerd.
Met het grootst denkbare respect voor alle keuzes die een gekozen volksvertegenwoordiging de nationale industrie
oplegt, wil ik eindigen met de stelling dat een integrale benadering de basis zou moeten zijn voor de
beoordeling van technologische opties ter reductie van de CO2-uitstoot. Met andere woorden: wat doet de optie
in de keten en wat zijn de bijkomende effecten van de optie?
De Nederlandse cementindustrie is gaarne bereid om daarbij een handje te helpen.

Referent dhr. Gouwens, TNO-Bouw
1 Rapportage
Er is reeds veel gedaan in de bouw op gebied van modelontwikkeling en milieubelasting. In Nederland vormt
het EcoQuantum model van IVAM een belangrijk instrument. Laat deze studie aansluiten bij het bestaande
werk op gebied van duurzaam bouwen en bij de LCA vereniging in de bouw. In de bouw is de afgelopen jaren
reeds veel gedaan om de milieubelasting terug te dringen. Een factor 2 is nu reeds bereikt. Een factor 10 is in
de toekomst wellicht haalbaar. Huidige gebouwen hebben naar schatting een verhouding tussen indirect en direct
energiegebruik van 1:2 tot 1:3.
2. Aanbevelingen
Materialen vormen een kleine vis in het totale energiegebruik. De grote winst kan gehaald worden in de
bestaande voorraad. Er zijn veel gebouwen die energetisch uitermate slecht in elkaar zitten. Indien je toch iets
aan de materialenkant zou willen doen dan zijn er met name knelpunten op te lossen bij:
• architectuur (smaak, verkeerd ontwerp -> zorg voor een duurzame architectuur)
• onvoldoende zorg voor kwaliteit van de gebouwen
• flexibel bouwen als oplossingsrichting (geldt vooral voor industriële gebouwen, kantoren, scholen)
• onvoldoende aandacht voor LCC (life cycle costing -> een groot probleem is dat kwaliteit in het algemeen
niet betaald wordt. Wat het in onderhoud kost, daar wordt niet op gelet.
De ruimtelijke-ordenings component verdient ook meer aandacht (werk-gebouwen zodanig situeren dat aanvoer/afvoer
energiezuinig kan gebeuren + gunstige oriëntatie)
3. Beleid
Regulering leidt tot dikke pakken papier (en werkt dus niet). Het is beter afspraken maken op vrijwillige basis.
Zo is bijvoorbeeld het Nationaal Pakket Duurzaam Bouwen opgesteld, iets dergelijks is ook voor de GWWsector
in ontwikkeling. Voor beleid is het ook belangrijk eindelijk een eenduidig LCA instrument te krijgen.
Verder kan het informatie-instrument beter ontwikkeld worden. De industrie heeft met de overheid afgesproken
om Milieu Relevante Product Informatie (MRPI) te leveren. De NVtB gaat dit coördineren en laat dit verifiëren
door een instituut (mogelijk TNO).
Gielen
De data in deze studie zijn gebaseerd op de informatie van de Europeses brancheorganisaties. Deze vergen
echter vaak nog een vertaalslag van volume-eenheden en stuks naar tonnen. De 12 Mt hogere schatting van de
emissies van de heer Lanser zal in het model op worden genomen. Het is natuurlijk waar dat beton een cruciaal
bouwmateriaal is. Dit laat echter onverlet dat de emissies een relevant deel van de totale CO2 emissies vormen,
waarvoor een oplossing moet worden gevonden. Hout is (in tegenstelling tot beton) een bouwmateriaal waarvan
de consumptie de laatste drie decennia gegroeid is, een ontwikkeling die nog versneld zou kunnen worden. Het
is aan te bevelen inefficiënte productieinstallaties in China te saneren, maar dat moet geen excuus vormen om
in Europa achterover te leunen. Ten aanzien van de technologische data zijn er twee achtergrond-rapporten
beschikbaar.
MARKAL of LCA ?
De analyse is opgezet vanuit 1 milieu probleem. Het voordeel van MARKAl ten opzichte van LCA is dat technologische
ontwikkelingen endogeen zijn. Deze ontwikkelingen kan over de levensduur van een gebouw van
belang zijn. De toekomstige afvalverwijdering is waarschijnlijk heel anders dan de huidige afvalverwijdering.
Dergelijke ontwikkelingen zijn endogeen in MARKAL, maar worden in LCA is het algemeen verwaarloosd.
Als methode verdient het daarom de voorkeur de opties eerst met MARKAL door te rekenen en daarna een
LCA uit te voeren om te kijken of er geen ongewenste neven-effecten optreden.
Is een factor 10 haalbaar bij bouwmaterialen ?
Gouwens: Een factor 10 lijkt niet haalbaar, maar het vormt ook maar een kleine vis met 20% van het totale energiegebruik.
Er is veel meer te halen aan de kant van de besparing op het directe verbruik.

Moll: Om een factor 10 te halen is het wel degelijk van belang naar het indirecte energiegebruik te kijken.

Metalen
Referent dhr.de Jong, Hoogovens
Er zijn geen gemakkelijke besparingen, in het algemeen is energiebesparing een zeer kostbare zaak. Op dit
moment is er een driedeling van de besparingen:
1 good housekeeping
2 energiebesparinsprojecten (o.a. WKK)
3 procestechnologische vernieuwingen
De mogelijkheden voor energiebesparingsprojecten zullen op termijn afnemen. Nieuwe technologieën moeten
na 2000 die besparingen opleveren. Maar de ontwikkeling van deze technieken is een lang en duur traject. De
ontwikkeling en implementatie van CCF kost bijvoorbeeld miljarden en duurt 30 jaar (indien succesvol).
De processen die na de ruwstaal-productie komen vereisen ook veel energie (4 GJ/t, met een stijgende tendens
vanwege de steeds verfijndere nabewerking). Hier zijn ook nog veel besparingen mogelijk (bijvoorbeeld
dungieten). Dit energiegebruik en de mogelijke verbeteringen dienen in de beschouwing mee te worden
genomen.
Ten aanzien van koleninjectie in de hoogoven wordt de besparing overschat (kijk ook naar energiegebruik voor
drogen en malen etc.). Ten aanzien van de ontwikkeling van CCF geldt dat Hoogovens daar momenteel volop
mee bezig is. De energiebesparing ten opzichte van de hoogoven-route loopt in de tientallen %. Evenals voor
COREX en Midrex is een gevoeligheidsanalyse echter belangrijk. Deze ontbreekt nu nog, de besparingscijfers
worden hard gesteld, terwijl er nog grote technologische barrières te nemen zijn. De kosten voor CO2-opslag
worden te laag ingeschat. Concluderend lijkt bij een succesvolle introductie van CCF over ca. 30 jaar enkele
tientallen % besparing mogelijk. Factor 10 lijkt voorlopig totaal niet haalbaar gezien de zeer hoge kosten.
Referent dhr. Mannaerts, CPB
Er bestaat een duidelijke discrepantie tussen wat technisch mogelijk is en wat economisch gerealiseerd kan
worden. Er is niet alleen sprake van een stijgende vraag ten gevolge van welvaarts ontwikkelingen. Ook de
aard van de vraag in de zin van de kwaliteitseisen aan de producten verandert. Deze ontwikkeling beperkt de
mogelijkheid voor substitutie tussen de processen (i.c. primair staal vs. secundair staal). Ten gevolge van de
marktwerking zullen prijzen veranderen waardoor een nieuwe situatie ontstaat. De kosten en het potentieel van
emissie-reductie worden hierdoor beinvloed, het is daarom moeilijk een cijfer aan te geven. Algemeen geldt dat
de marginale kostencurve voor CO2 opslag zal omhoog gaan bij een zeer brede toepassing van deze strategie.
Deze curve staat niet in MARKAL.
Afgezien van de CO2 opslag lijkt schroot het aantrekkelijkst om CO2-emissies te reduceren, maar er is veel
schroot nodig om aan de gehele vraag te voldoen. De marginale. kostencurve voor schroot stijgt snel bij een
brede toepassing van de recyclingroute. Deze prijsstijging van schroot zal een verschuiving naar elektrostaal
(recycling) blokkeren. Het geschetste potentieel in 2050 houdt nog geen rekening met de ontwikkeling van geheel
nieuwe technologieën, die het besparingspotentieel zullen vergroten.
Daniëls
1. (n.a.v. De Jong: processen volgend op de staalproductie zijn niet opgenomen in de analyse) De keuze voor
vloeibaar staal als voorlopig eindproduct is gemaakt om een goede vergelijkingsbasis te bieden voor de verschillende
productieroutes; bij het opnemen van de bewerkte eindproducten zou het energiegebruik meer
specifiek zijn voor het betreffende product dan voor de productieroute. Bovendien was vanwege de beperkte
tijd een analyse van de nabewerkingsprocessen niet mogelijk.
2. (n.a.v. niet reële externe parameters) Voor het meenemen van de interactie met externe parameters in de
analyse is MARKAL nodig. Mijn presentatie betreft een verkennende studie waarin de invloed van externe
factoren op emissies en kosten in een sector bekeken wordt. Voor een bredere analyse is meer tijd nodig.

3. (n.a.v. Mannaerts: nieuwe technieken in 2050 en De Jong: factor 10 niet haalbaar, technologische en
economische barrières) Er is gekeken naar het potentieel op de lange termijn; dit betekent dat het potentieel
niet op basis van korte termijn beperkingen beoordeeld kan worden. Anderzijds geldt inderdaad dat technieken
in de ijzer- en staalindustrie een lange looptijd hebben. Het is niet waarschijnlijk dat technieken die
nu nog niet in ontwikkeling zijn in 2050 al een grote rol voor het emissie-potentieel zullen hebben.
4. De keuze voor nieuwe technologieën is sterk afhankelijk van de marktsituatie op het moment dat een keuze
gemaakt moet worden. Op dit moment lijkt de uiteindelijke afdanking van de cokesovens de belangrijkste
drijvende kracht achter de ontwikkeling van nieuwe technieken: het is onwaarschijnlijk dat er vergunningen
zullen worden verleend voor nieuwe cokeovens.
Kram (n.a.v. kritiek Mannaerts): Denkt dat veel van de prijsvorming die de markt gaat bepalen endogeen is in
MARKAL. Een aanbodcurve van staalschroot is reeds in het model opgenomen. Een opslagcurve voor CO2 is
technisch in het model op te lossen. Verschillende toepassingen kunnen gescheiden worden gemodelleerd. Het
is evident dat het niet verstandig is te veel te vertrouwen op secundaire routes, er zijn ook factoren die het potentieel
beperken. De levensduur van glas en metaal in de economie is bijvoorbeeld heel verschillend, Bij recycling
van staal moet je er rekening mee houden dat staal ongeveer 30 jaar meegaat.
Daniels
Een hoger recyclingspercentage binnen Europa lijkt nog mogelijk. Dan moet wel grondig naar de kwaliteit van
het schroot gekeken worden.
Van de Gaag
Beveelt aan 50-50 (DRI-Schroot) door te rekenen. Schroot is niet te duur. DRI laat wereldwijd een aanzienlijke
groei zien. Dit biedt aanvullende mogelijkheden voor besparingen.
Daniels
Dit lijkt een interessante strategie. Deze ontwikkeling is momenteel al ingezet in de VS

Algemene vragen en forumdiscussie
Deelnemers aan het panel:
• van der Berg (Ministerie van Economische Zaken, DGE)
• Idenburg (RIVM, LAE)
• de Hoog (VROM, IBPC)
• Vijverberg (VROM, SP)
• Kok (NOP-MLK)
• Kram (ECN-BS)
Stellingen
1 Er moet een apart DG Materialen - en milieubeleid (MMB) komen
Vijverberg: liever coordinatie/aansturing vanuit VROM
v/d Berg: is het ermee eens dat er meer coordinatie moet komen
de Hoog: wellicht is er wel een grotere aparte plaats nodig voor materialen binnen het VROM-beleid
Idenburg: materialenbeleid moet een belangrijke positie innemen binnen het milieubeleid, ook bij het RIVM
Zijlstra: Overheid zit dan wel op de stoel van de ondernemer.
Kram: Denkt dat er tijdens deze dag geconstateerd is dat het allemaal niet automatisch tot succesvol beleid zal
leiden. Materialen moet je een aparte plek geven in je organisatie. Men heeft te maken met verschillende
beleidsterreinen. Hij pleit daarom voor een centralere rol voor productie en materialen.
Groenewegen: Mogelijkheid die VROM bij NMP zag was niet de goede insteek, volgens hem natte vinger
werk. Het is nodig een serieuzer antwoord te verzinnen. Zijn conclusie is dan ook dat er behoefte is aan een
centrale afwegings instantie. Wil de overheid wel materialenbeleid of wordt dit ondergesneeuwd door andere
doelstellingen ?
v/d Berg: Materialenbeleid en materiaalbesparingen kan de overheid niet zelf doen. Dit vereist altijd een goede
samenwerking met de industrie. De onderzoeks-instituten zouden zich op dit terrein meer moeten profileren.
De Hoog: Denkt dat aansturing op hoger niveau gebracht moet worden binnen bedrijven
Idenburg: De MARKAL studie levert belangrijke inzichten. Zaken die uit dit onderzoek naar voren komen dienen
door het beleid opgepakt te worden. Er is immers beleid ten aanzien van ontkoppeling en dematererialisatie.
Zij vindt het echter niet zinvol om het bestaande institutionele apparaat overhoop te gooien.
2 Het onderzoek van de GTI’s en NOP’s moet meer aandacht besteden aan
materiaalefficiency
v/d Berg: Met materialenbeleid raak je hart van de ondernemer. Hij vraagt zich af of de GTI’s voldoende in het
netwerk van de industrie zitten. De beleids-implementatie is een probleem.
Kok: Er is meer focus nodig op implementatie naast technologische potentieelschattingen.
Zijlstra: Materiaalkeuze is de verantwoordelijkheid van de ondernemer. Kan ISO 14001 wellicht een handvat
vormen voor materialenzorg ?
Mannaerts: Er is een concept nodig voor integraal beleid, we zijn er nog lang.niet.

Vragen
1 Er is een lange-termijn, consistent beleid nodig, anders is het voor de industrie moeilijk iets te
bereiken. Kan de overheid een dergelijk beleid vormgeven ?
De Hoog: Om consistentie te bereiken kunnen afspraken in convenanten worden vastgelegd. Er kunnen eerst
bescheiden doelen worden vastgesteld die gaandeweg worden bijgestuurd. Instrumenten gaandeweg concretiseren.
v/d Berg: Consistentie is belangrijk voor de doelgroepen. 30 jaar is voor het beleid echter te lang. Volgens de
studie van Wolfson zijn in Nederland en in Europa de heffingen op dit moment al hoger dan in de VS en Azië.
Erg hoge heffingen zijn niet erg realistisch i.v.m. de internationale concurrentie. Zo zal een EU-heffing van
250 gulden/t CO2 leiden tot een grote toestroom van producten van buiten Europa, indien de transportkosten
lager zijn dan 250 gulden/t CO2.
Gielen: Uit de MARKAL berekeningen blijkt dat men om een factor 4 emissie-reductie te bereiken hoge heffingen
zou moeten introduceren. De berekeningen zijn geen pleidooi om deze heffingen in te voeren. Men dient
zich wel te realiseren dat er een kloof gaapt tussen enerzijds ambitieuze emissie-reductiedoelstenningen en anderzijds
de beperkte instrumenteerbaarheid.
De Hoog: 200 ECU/ton zou geweldig zijn, de uitdaging voor beleidsmakers is juist om innovaties te gaan gebruiken
en/of de doelstelling te bereiken
Bergsma: Probeer het geld van de heffingen direct terug te sluizen en de heffingen meer in de sectoren te
houden.
Groenewegen: Naast heffingen is het ook belagrijk om te kijken naar de omstandigheden: wanneer is er wel en
wanneer is er geen geen potentieel voor innovaties. De mogelijke innovaties via betere informatievoorziening
en meer R&D ondersteunen.
Annema: Vraagt zich af of alle voor alle doelgroepen een factor 2 -10 bereikt kan worden. Komt er nog een
helikopterview
Kram: De MARKAL studie geeft die helicopterview. De resultaten worden verder uitgewerkt per productgroep.
Factor 2-10 moet zeker geen voorschrift worden voor alle sectoren.
v/d Berg: het NMPIII en de nota energie besparing tonen aan dat we het in Nederland als geheel redelijk doen
voor wat betreft energiebesparing. De grote industrie doet het internationaal goed, middel grote industrie redelijk,
gebouwde omgeving loopt achter. In nota wordt ingezet op een forse beleidsinspanning. Verkeer is heel
internationaal en moeilijk aan te pakken.
2 Er is in deze studie met een Nederlandse bril naar Europese kansen en problemen gekeken. Waarom
is Nederland zoveel beter dan de rest van West-Europa ? Wat denken industrieën te kunnen doen om
de rest van Europa en de rest van de Wereld op Nederlands niveau te brengen ?
Idenburg: Nederland is klein, men zit dichter op elkaar, dan kom je eerder de problemen tegen. Vangnet in andere
landen kleiner. Aan de andere kant is er in Nederland veel zware industrie waarvoor vanuit de overheid
veel aandacht is. CO 2 en NOx hebben de aandacht gekregen.
Vijverberg: Nederland is dichtebevolkt, het beleid is daarom gericht op een vooruitstrevende aanpak van milieuproblemen.
v/d Berg: Bij de ondersteuning van grote buitenlandse industrieën speelt de concurrentie mee. Bij zaken die
niet concurrentie-gevoelig zijn kan een kennisuitwisseling plaatsvinden. Alleen of grote multinationals dat
doen?
?: Zware industrie doet dat al (bijvoorbeeld de staalindustrie).
De Hoog: De rijksoverheid heeft veel onderzoek laten doen, milieumaatregelen zijn gestimuleerd via belasting
faciliteiten. Daardoor zit de industrie op dit moment industrie op een hoog niveau. Er zijn inmiddels vergelijkbare
programma’s op Europees niveau.
v/d Berg: Er is veel kennis, maar in EU-kader wordt daar nog niet veel mee gedaan.
Zijlstra: JI is een mogelijkheid.
Brezet: Vindt dat we de vooruitstrevende positie van Nederland niet moeten overschatten.

3 Wat voor problemen zouden we als beleidsmakers hebben als we dit soort studies en dit soort modellen
niet hadden ?
v/d Berg: Het is nuttig dat verschillende instituten tot heel verschillende conclusies komen ten aanzien van dit
onderwerp.
De Hoog: Het is interessant om dingen weer een keer na te lopen. Hoe komen nu materiële grondstoffen ons
land binnen en hoe hangt dit samen met de energiehuishouding. Het is goed om daar sommen over te maken en
conclusies uit te trekken. Het is belangrijk de non-CO2 broeikas gassen ook mee te nemen. Deze zijn immers
ook onderdeel van de Kyoto doelstelling.
Kram: Het MARKAL model is inmiddels uigebreid met non-CO2 broeikasgassen.
v/d Berg: De meerwaarde van de studie is dat de extremen in beeld worden gebracht.
Smokers: De deelstudies leiden ertoe dat de sectoren beter in MARKAL worden weergegeven.
Kram: de meerwaarde ligt niet alleen in het West-Europese MARKAL model. Ook de sectorale studies vormen
op zichzelf belangrijk materiaal.
4 Is MARKAL geschikt voor het doorrekenen van energiebesparingsopties (of CO2-opties) in de transportsector
? Zo ja, waarom ? Zo nee, kan het geschikt gemaakt worden?
Kram: MARKAL is berperkt geschikt, er zijn veel exogene factoren (bijvoorbeeld belastingen, irrationeel gedrag
van de consument) die niet in het model meegenomen kunnen worden. Met name in de transport sector
wegen deze beperkingen zwaar. Doorrekenen de consequenties. Ja en nee
v/d Berg: Het ECN-SAVE model kan ook gebruikt worden voor het doorrekenen van de transportsector
Annema: Alleen die twee modellen, beleidsmatig is daar weinig mee te doen. Buiten het model om zijn er veel
betere analyses mogelijk ?
Kram: Er zijn echter niet veel direct geregistreerde waarnemingen beschikbaar. Daarom kunnen modelberekeningen
belangrijke informatie toevoegen.
5 Over een periode van 50 jaar zijn aanzienlijke veranderingen in het consumptiepatroon te verwachten.
In hoeverre wordt dit in MARKAL meegenomen ? Als het er niet inzit, zou het er dan niet in
opgenomen moeten worden ?
Kram: MARKAL houdt er bijvoorbeeld rekening mee dat autonoom het gewicht van auto’s toeneemt ten
gevolge van een stijgende welvaart.
Smokers: Juist voor consumentenartikelen zijn ten gevolge van de stijgende bestedingsruimte grote veranderingen
te verwachten. Het onderwerp verdient meer aandacht.
Vervolgtraject
Kram: Vindt dat er nuttige en leerzame dingen zijn besproken. Er zijn afspraken gemaakt voor verdere gesprekken.
De workshop vormt een goede input voor de sturing van het project. Het plan is een goed verslag van
deze dag maken en dit onder een breed publiek te verspreiden. Het commentaar zal in de detailstudies verwerkt
worden. Er zal nog een vervolg workshop georganiseerd worden waarin de nieuwe resultaten inclusief het
verwerkte commentaar gepresenteerd zullen worden.

Deelnemers
Beleid
drs. H.L. Baarbé (VROM-DGM-GV)
mr. A.J. van den Berg (EZ-DGE-EBD)
dhr. A. Bos (EZ-DGE-EBD)
drs. C. Clement (VROM-DGM)
drs. B. van Engelenburg (VROM-DGM-LE)
dhr. R. de Graaf (EZ-DGE-EBD)
drs. M.M. de Hoog (VROM-DGM-IBPC)
drs. H. Merkus (VROM-DGM-LE)
dhr. C. Veerman (VROM-DGM-A)
drs. C.H.T. Vijverberg (VROM-DGM-SP)
Industrie/belangenorganisaties
ir. J.W. Broers (Rijkswaterstaat DWW)
ir. P. van der Gaag (GHR)
dr. E. Janssen (Koninklijke Hoogovens Aluminium)
ir. A.G. de Jong (Koninklijke Hoogovens Staal)
ir. P.A. Lanser (VNC)
mw. O.C. Ortmans (VVAV)
dhr. G. Rolleman (Wavin Nederland)
mw. A. Schermer (KNP-BT)
dhr. T.J. Visser (Vereniging Milieudefensie)
mw.drs. W.M. Zijlstra (Vereniging VNO-NCW)
dhr. H. Muilerman (St. Natuur en Milieu)
Onderzoek/onderzoeksmanagers
dhr. J.-A. Annema (RIVM)
ir. G.C. Bergsma (CE)
Prof.dr. H. Brezet (TU Delft, dagvoorzitter)
ir. R.C. Dorgelo (SBR)
dhr. ir. P. Alderliesten (ECN)
drs.ir. A.W.N. van Dril (ECN)
drs. F.J. Duijnhouwer (RMNO)
Prof.ir. C. Gouwens (TNO-Bouw)
ir. G. van Grootveld (DTO)
mw.ir. M.G.M. Harmelink (RIVM)
mw.drs. J. Hoekstra (NOVEM)
mw.dr.ir. A.M. Idenburg (RIVM)
dhr. J. de Koning (TNO-MEP)
drs. M. Kok (NOP-MLK)
drs. H. Mannaerts (CPB)
drs. D. Scheele (WRR)
dr.ir. R. Smokers (TNO-Wegtransportmiddelen)
dhr. E. Tempelman (TUD-TA)
mw.dr. E. van der Voet (CML)
MATTER deelnemers
dr. R. Benders (IVEM)
mw. drs. M. Bouwman (IVEM)
drs. B. Daniëls (IVEM)

ir. R. van Duin (Bureau B&G)
drs.ir. D. Gielen (ECN)
mw. drs. T. Goverse (VU Amsterdam)
dr. P. Groenewegen (VU Amsterdam)
drs. M. Hekkert (NW&S)
ir. L. Joosten (NW&S)
ir. T. Kram (ECN)
dr. H.C. Moll (IVEM)
dr. E. Worrell (NW&S)