Ecological Economics 42 (2002) 43 – 57

This article is also available online at:

www.elsevier.com/locate/ecolecon

SPECIAL SECTION

The contemporary European copper cycle: waste

management subsystem
M. Bertram a,b,*, T.E. Graedel a, H. Rechberger a,1, S. Spatari a
a
Center for Industrial Ecology, School of Forestry and En6ironmental Studies, Yale Uni6ersity, New Ha6en, CT 06511, USA
b
Department of Industrial Sustainability, Institute of Technology Cottbus, D 03050 Cottbus, Germany

Abstract

A comprehensive copper mass balance for waste management in Europe has been carried out, including municipal
solid waste, construction and demolition waste, wastes from electrical and electronic equipment (WEEE), and
end-of-life vehicles (ELV). The recycling efficiency of the current waste management system in Europe was quantified
and the sources of copper scrap used for secondary copper production were determined. Additionally, an assessment
of copper losses to the environment from incinerators and landfills was undertaken. As a final step, select parameters
were varied to test the sensitivity of copper waste generation results to the uncertainties in the data. The total flows
of copper into the European waste management system consists of 920 Gg/y domestic copper waste and of 300 Gg/y
imported old scrap, of which 740 Gg/y are recycled and 480 Gg/y are landfilled. In Europe 2 kg per capita of copper
waste is generated annually. WEEE and ELV are the most important domestic waste streams from the perspective of
copper contents. They contain 67% of the total copper throughput, but only make up 4% of the mass of total waste
generation. Because WEEE is the fastest growing waste category, this finding emphasizes the need for efficient WEEE
recycling strategies. The overall recycling efficiency for Europe for copper in all types of waste, excluding prompt
scrap and scrap imports, is 48%, with a range of 5–58% depending on the country. This shows further potential for
increased recycling activities in the future. Emissions of copper to the environment are under 5 Gg/y but several new
sources for emissions are not yet quantified. Uncertainties in waste generation rate and composition for some waste
categories (WEEE, C&D) are high, and additional analysis is needed to confirm the above findings. © 2002 Elsevier
Science B.V. All rights reserved.

Keywords: Copper; Emissions; Europe; Material flow analysis; Recycling; Waste management

1. Introduction

Two main goals of modern waste management

* Corresponding author. Tel.: +1-203-432-5675; fax: + 1- are to protect human health and the environment
203-432-5556 and to conserve resources such as materials, en-
E-mail address: marlen.bertram@yale.edu (M. Bertram).
1
ETH Swiss Federal Institute of Technology Zurich, Chair
ergy and space. Material flow accountings (Eu-
of Resource and Waste Management, CH 8093 Zurich, rostat Statistical Office of the European
Switzerland Communities, 1997; Spatari et al., 2002 —here-

0921-8009/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 1 - 8 0 0 9 ( 0 2 ) 0 0 1 0 0 - 3
44 M. Bertram et al. / Ecological Economics 42 (2002) 43–57

after Paper II) are one means to determine how copper stocks and flows (STAF) in the European
far a selected waste management system realizes economy which frames the budgeting approach
these practices with regard to above objectives, (Graedel et al., 2002), and characterizes the over-
since they help to determine emissions from in- all ‘STAF-Europe’ copper cycle (the 15 countries
dustrial and technological systems to the environ- of the European Union plus Poland, see Paper II).
ment, to detect potential resources (stocks and In this paper we aim to quantify both generation
flows) within that system, and to reveal and un- rates for wastes and their copper contents in order
derstand non-optimal waste management prac- to identify the relevance of waste streams within
tices. Non-optimal management can refer to the copper life cycle. Recycling of each waste
either the generation of waste streams having category is assessed and the overall recycling effi-
unfavorable material composition or simply the ciency for Europe is determined. As far as possi-
unfavorable management of waste. ble, investigations were carried out on the country
Wastes of unfavorable composition are either level. Emissions of copper to the environment
too contaminated for direct and safe disposal and from incinerators and landfills are assessed and
must therefore be treated, or have concentrations other sources for losses are briefly discussed.
of valuable substances too low to serve as a
resource (e.g. municipal solid waste (MSW)).
Such wastes can only be prevented by changes in 2. Method and materials
product design at the beginning of the life cycle or
under certain circumstances through different col- 2.1. System analysis
lection and/or separation strategies. Because
products can have residence times of several years Material flow analysis is used to establish the
to decades, current and emerging design for envi- copper budget for the European waste manage-
ronment (or design for disposal) activities do not ment system. The temporal and spatial system
have an immediate effect on waste management. boundaries are defined in Paper II; as far as
The same applies to prevention strategies. On the possible, data for 1994 have been chosen. The
contrary, the vast stocks of materials that have system definition guarantees that more than 90%
accumulated over the past several decades in pri- of European waste is covered by the cycle. The
vate households and infrastructure suggest a rise term solid waste comprises the following cate-
in waste quantities in the short-to-medium-term gories: MSW; construction and demolition waste
future. (C&D); WEEE; end-of-life vehicles (ELV) includ-
An example of suboptimal management of ing passenger cars, buses and trucks; sewage
waste categories is the direct landfilling of wastes sludge (SS); hazardous wastes (HW); and indus-
that have resource potential but are not yet recov- trial waste (IW). The above categories are chosen
ered (e.g. wastes from electrical and electronic because they follow ‘classical’ data and manage-
equipment (WEEE)). This example illustrates that ment approaches. The material is collected by the
combined logistical and technological actions are municipality or by private companies, mechani-
still appropriate and necessary for medium-term cally sorted, separated for recycling purposes (the
solutions in waste management. Overall, detailed recycled stream is known as old scrap), and usu-
material balances for economically and ecologi- ally either incinerated or directly landfilled. The
cally important substances can highlight the defi- chosen definition of waste excludes wastes from
ciencies in existing waste management systems, the primary production of copper in mills,
and therefore are quite valuable for decision mak- smelters, and refineries such as tailings and slag,
ing with respect to optimal waste management which are treated separately in Paper II. Gordon
strategies and technologies. (2002) discusses the generation rates, composition,
This paper investigates the flows of copper as- and recovery potential of these wastes with regard
sociated with the European waste management to copper. Besides the above-mentioned waste
system. The study is part of a research study of categories, prompt scrap is also investigated: it
 M. Bertram et al. / Ecological Economics 42 (2002) 43–57 45

comprises all copper waste that arises during the trade is mainly bound for recycling. Hence, it is
manufacturing of finished products and is imme- safe to assume that copper flows via waste for
diately recycled within the production facility disposal are insignificant for the budget with re-
without entering the classical waste management gard to import/export.
system. Hence, prompt scrap has a much shorter (b) Reported statistical values for transconti-
lifetime than old scrap. Fig. 1 illustrates the posi- nental scrap trading are assumed to be exclusively
tion of classical waste management within the old scrap since prompt scrap is mainly re-melted
‘STAF-Europe’ system shows the waste streams and returned to the production stage of semi-
and losses to the environment that were products within Europe.
investigated. Waste generation rates, recycling, and disposal
The following assumptions are made for model- quantities are determined by a literature survey,
ing the system: and losses from treatment and disposal facilities
(a) Apart from copper scrap, imports and ex- into the environment are estimated. Fig. 2 illus-
ports of wastes to and from Europe are not trates the relevant processes and flows represent-
considered. The copper concentration in wastes ing European copper-waste management circa
for disposal (e.g. MSW, HW) is small compared 1994. The first step for the seven identified cate-
to wastes for recovery and recycling (old scrap), gories of waste is collection. Portions of the waste
and therefore only high quantities of cross- streams can be mechanically treated and sorted,
boundary waste streams could be relevant. The and copper scrap may be separated and returned
import/export of wastes for disposal and recovery to the economy. The reuse of copper-containing
to non-EU member states is regulated in the EC products and product parts is not counted in the
Waste Management Regulation and the Basel recycling flows, since the reuse extends the lifetime
Convention (UBA, 2000). This legislation strongly of products and therefore takes place in the use
restricts the import/export of wastes for disposal. phase. Hence product reuse is not inventoried in
A study on trade of non-ferrous metal-bearing this waste management MFA.
waste (Johnstone, 1998) indicates that the major- Waste that is not recycled is either incinerated
ity of waste trade (700 Gg/y) is intra-European. or directly landfilled. The partitioning between
There is a HW trade (120 Gg/y) to developing incineration and landfill is determined for each
countries, but as Johnstone (1998) cites, this waste waste category on a country level. The loss of

Fig. 1. The position of ‘classical’ waste management within the ‘STAF-Europe’ system and the major investigated material flows.
46 M. Bertram et al. / Ecological Economics 42 (2002) 43–57

Fig. 2. Solid waste is collected and partly mechanically separated for recycling and recovery purposes. Copper leaves the system
either via recycling streams or is lost to the environment. The remainder is either incinerated or directly landfilled.

copper to the environment is assessed for gaseous 3.1.2. Hazardous waste

and aqueous emissions from incineration, leachate For HW an average generation of 90 kg/c y
from landfills, SS applied to agricultural land, and is documented (CEC, 2000). Since 1994, the
ashes for road construction. Directory on HW (94/904/EG) has regulated
HW in the European Union. It defines 237
types of HW. This directory is mandatory for
3. Results and discussion all member states (UBA, 1999). Nevertheless,
national totals reported to the OECD and
3.1. Solid waste generation Eurostat are still generally based on national
classifications and definitions, which makes it
Table 1 shows annual waste generation in very difficult to draw valid comparisons
‘STAF-Europe’ subdivided into MSW, HW, SS, among data from different countries, as the
C&D, IW, WEEE, and ELV. classification systems can be quite different. The
variations in HW classification can be explained
3.1.1. Municipal solid waste by the fact that the European Waste Catalogue
As shown in Table 1, the variation in the does not sufficiently describe what is regarded
generation of MSW ranges from 250 (Austria) to and reported as HW (Christiansen and Munck-
670 kg/c y [c is the unit for capita] (Luxembourg) Kampmann, 2000). Since HW are often compo-
within the European countries studied (CEC, nents of other waste groups, e.g. C&D waste
2000). MSW data usually include commercial (asbestos, lead, tars, paint and preservative
waste together with waste from private house- residues, adhesives, bonding agents and sealant
holds, as well as waste fractions (e.g. paper, metals, and certain plastics), its generation is difficult
plastics, etc.) separately collected for recycling. to identify with accurency.
 M. Bertram et al. / Ecological Economics 42 (2002) 43–57 47

3.1.3. Sewage sludge comparably higher C&D waste generation rate

Dry weight production of SS is about 19 kg/c y (670– 725 kg/c y) than countries with low popula-
(CEC, 2000) in member states of the European tion densities (e.g. Sweden, Greece, and Ireland),
Union. The percentage of the population served whose range is between 160 and 190 kg/c y (Eu-
by sewage waste water treatment facilities varies ropean Commission, 1999). This difference may
between 10% (Portugal) and 98% (Denmark) in partly be explained by differences in data collec-
countries of the European Union (EEA, 1997). tion and reporting among the countries, as well as
inaccurate estimates in countries where no official
3.1.4. Construction and demolition waste statistics exist. Research results show that, in gen-
The definition of C&D waste (European Com- eral, insufficient databases exist for the quantifica-
mission, 1999) includes waste arising from the tion of C&D waste. An MFA of the construction
total or partial demolition of buildings and/or industry for a region in Austria showed that
civil infrastructure; waste arising from the con- reported data on C&D are very likely to be too
struction of buildings and/or civil infrastructure; low, and a considerable amount of C&D is not
soil, rocks and vegetation arising from land level- yet reported (Glenck et al., 1997).
ing, civil works and/or general foundations; and
road planings and associated materials arising 3.1.5. Industrial waste
from road maintenance activities. For the purpose Data published by the European Topic Center
of determining the copper content in C&D waste on Waste and Material Flows (2001) are used for
in Section 3.4, the generation of soil and dredging the generation of IW. This term includes all
soil, asphalt, tar and tarred products is not wastes from industry and manufacturing that are
counted in this model. It is found that countries non-hazardous. The generation of non-hazardous
with high population densities (e.g. Germany, The IW ranges from 40 (Portugal) to : 1600 kg/c y
Netherlands, Belgium, and Luxembourg) have a (Sweden). A uniform system for classifying IW

Table 1
Solid waste generation in Europe (kg/c y)

Country Population (million) MSW C&D SS HW IW WEEE

Germany 81.4 503 725 28 113 1032 7

France 57.9 379 408 13 102 1813 7
UK 58.3 394 515 19 41 961 7
Italy 57.2 430 350 20 100 388 7
Spain 39.1 451 332 20 100 353 7
Netherlands 15.4 355 725 20 83 1247 7
Belgium 10.1 352 668 9 201 1327 7
Sweden 8.8 355 192 26 57 1591 7
Austria 8.0 253 588 21 95 1313 7
Denmark 5.2 380 508 32 52 442 7
Greece 10.4 354 173 20 100 48 7
Finland 5.1 348 265 28 112 294 7
Portugal 9.9 377 323 15 100 40 7
Ireland 3.6 422 158 11 64 1056 7
Luxembourg 0.4 677 750 20 358 500a 7
Poland 38.7 300 400a 8 50 800a 4a
‘STAF-Europe’ 409.5 402 427 19 90 846 7

Data source: European Commission (2001), CEC (2000), European Commission (1999), European Topic Center on Waste and
Material Flows (2001), Christiansen and Munck-Kampmann (2000), O8 kopol (1998).
a
Informed estimate.
48 M. Bertram et al. / Ecological Economics 42 (2002) 43–57

does not exist in Europe; therefore, national totals second-hand markets and repair shops. ARGUS
are based on national definitions and reporting (1996) estimates in Germany a portion of WEEE
systems. from private households that is collected in do-
mestic waste in the range 0.52 and 1.6 kg/c y. As
3.1.6. Waste from electrical and electronic for the industrial sector, we assume the theoretical
equipment WEEE generation rates correspond to actual gen-
On the basis of its high electric conductivity, eration rates since we expect industry to have
copper is the basic material of electrical engineer- short hibernation periods for retired electronic
ing technologies. Today more than half of Ger- products. An expected generation rate of 1 kg/c y
many’s copper consumption is used in mechanical by private households and 6 kg/c y by industry
apparatus and electrical engineering (DKI, 2001). for the year 1994 is assumed for all countries
The interpretation of the term WEEE differs except Poland. Regional EU differences are not
greatly among countries. WEEE covers a wide taken into account. Any attempt to measure this
spectrum of electronic products ranging from con- potential more precisely, and especially to ade-
sumer goods such as personal computers and quately take into account regional and socio-eco-
refrigerators to capital goods such as medical nomic differences, will be faced with great
equipment systems and monitoring and control difficulties.
instruments. Actual components of published
WEEE generation rates in Europe are insuffi- 3.1.7. End-of-life 6ehicles
ciently recorded. Estimates for OECD countries Quantities for passenger cars, buses, and trucks
assume a distribution pattern of 70% consumer are taken from a study by Müller (1995) con-
goods and 30% capital goods (O8 kopol, 1998). ducted for EUROSTAT using the following
WEKA (1999) indicates for Germany a distribu- equation:
tion of two-thirds consumer goods and one-third
ELV = DStock + New +Import− Export (1)
capital goods. Quantities for the domestic mate-
rial flow of WEEE in Germany range between 16 where ELV, end-of-life vehicles; DStock, previous
and 21 kg/c y in 1994 (Zhang and Forssberg, year’s vehicle stock minus actual vehicle stock;
1999; ZVEI, 1992), and for the European Union, New, first registrations of vehicles; Import, im-
16– 24 kg/c y in the late 1990s (Christiansen and ports of used vehicles; Export, exports of used
Munck-Kampmann, 2000; Welslau and Kraus, vehicles.
1998). These values are based on lifetime models The results of Eq. (1) for 1992 are presented in
incorporating production, import and export, and Table 2. There is some uncertainty associated with
consumption statistics and the average lifetime of the final geographic destination of all ELV. This
electrical and electronic products. European data is especially true in Germany where 50–60% of
collected by O8 kopol (1998) show a separate col- ELV are exported, mainly to Eastern European
lection rate of WEEE from private households of countries (ARGE-Altauto, 2000). This uncer-
1 kg/c y in the 1990s. The same study indicates tainty presents boundary allocation difficulties for
that a rather successful separate collection of STAF-Europe, because Poland is a primary im-
WEEE from private households could probably porter of retired vehicles from Western European
not catch more than an average of 4– 5 kg/c y. countries, and is included within the STAF-Eu-
The wide gap between the theoretically expected rope boundary.
generation rate of 14 kg/c y and the actual possi-
ble collection rate of 4– 5 kg/c y from private 3.2. Generation of new scrap
households suggests that a significant fraction of
retired electronic products remains in ‘hiberna- Table 3 presents the data on secondary (recy-
tion’; that is, they are not collected, but rather cled) copper production. The published data on
stored in households. Furthermore, a fraction of secondary copper includes the copper returned to
retired electronics may circulate and hibernate in smelters/refineries (International Copper Study
 M. Bertram et al. / Ecological Economics 42 (2002) 43–57 49

Table 2
ELV (passenger motor cars, buses and coaches, and trucks) in pieces

Country Passenger motor cars Buses and Vehicles for the transport of goods and other vehicles (thousand)
(thousand) coaches

Germany 3272 5127 173

France 1831 3957 358
UK 1252 3265 466
Italy 1384 2889 186
Spain 363 2069 51
Netherlands 317 503 52
Belgium 262 130 25
Sweden 215a 2.226a 14a
Austria 214a 634a 35a
Denmark 71 151 16
Greece 135 156 2
Finland 106a 780a 5a
Portugal 5 181 5
Ireland 61 41 27
Luxembourg 12 33 6
Poland n.a. n.a. n.a.
‘STAF-Europe’ 9500 19 918 1421

n.a., not applicable.

a
Import and export data are not available for these countries, so values are estimated based on historical vehicles registration
statistics (AAMA, 1996) and assumed lifetime of 13 years for cars and 20 years for busses and trucks.

Table 3
Secondary copper production and copper origin in 1994

Country Production in Gg Source in Gg

Secondary smelter/refinery production Secondary manufacturing Old scrap Prompt scrap

production

Germany 339 264 290 313

France 25 106 39 92
UK 37 88 44 82
Italy 78 314 116 275
Spain 47 24 38 33
Netherlands – 52 10 42
Belgium/Luxembo 160 – 112 48
urg
Scandinavia 24 56 28 45
Austria 44 – 31 13
Greece – –
Portugal – –
Ireland – –
Poland 24 – 17 7
Other Europe – 42 8 34
‘STAF-Europe’ 778 946 734 990
50 M. Bertram et al. / Ecological Economics 42 (2002) 43–57

Group, 1999), and the copper used directly by recycling of C&D in member states of the Eu-
manufacturers, known as re-melt scrap (World ropean Union. About 175 000 Gg of C&D, con-
Bureau of Metal Statistics, 1995). In general, sec- taining : 3% metals (mostly steel), are generated
ondary copper used in smelters and refineries is in the European Union each year. This metal
composed of 70% old and 30% prompt scrap. content corresponds well with quantities found in
Secondary copper used in manufacturing (re-melt) C&D waste treated by a sorting plant in Switzer-
must have higher purity, and therefore is com- land (Brunner and Stämpfli, 1993). Recycling and
posed of 80% prompt and 20% old scrap (Interna- reuse of C&D varies from B 5% for some coun-
tional Copper Study Group, 1999). Today almost tries (e.g. Spain, Portugal, Greece, and Ireland) to
100% of copper production waste is recycled be- 90% for others (The Netherlands), resulting in a
cause of its high purity and economic value. Only weighted average of 28% for Europe. Most coun-
galvanic sludge, chemical, and electrolysis tries with poorly developed recycling-infrastruc-
residues, whose recycling is not yet economical, tures have an active ‘informal’ recycling industry
are disposed of as part of HW. Therefore data on that is based on removing all items with economic
prompt scrap are used to describe the generation value (e.g. copper pipes) from C&D by hand
of copper production waste. before it is landfilled. A study on material and
substance flows in a full-scale construction sorting
3.3. Generation of old scrap plant gives the following four output fractions
(Brunner and Stämpfli, 1993): organic fraction
Determining the separation efficiencies for old (25%), metal fraction (3%), and two inorganic
scrap in each country is not possible due to fractions (bulky fraction 45%, fine fraction 27%).
limited information; thus, only approximations The substance balance for copper shows that 69%
can be made. We assume that the relevant waste of the copper leaves the sorting plant within the
groups for copper recycling are ELV, C&D, and metal fraction. This amount of copper could be
WEEE. theoretically recycled. The study also shows that
about 50% of the copper is separated by hand
3.3.1. End-of-life 6ehicles sorting. We assume that all countries except Italy,
Copper from ELV is gained from dismantling Spain Portugal, Greece, and Ireland have com-
parts of the engine, radiators, and electric compo- parable metal separation technologies and assess
nents, as well as from the shredding and non-fer- the recycling efficiency for copper in C&D to be
rous separation processes. Frei et al. (1993) note between 30 and 50%.
that from the average vehicle scrapping plant 40%
of the copper contained in automobiles reaches 3.3.3. Waste from electrical and electronic
the scrap market, 35% is transferred to the com- equipment
bustible light fraction (residues from shredder Waste from electrical and electronic equipment
plants, known as ASR), and 25% remains in (WEEE) is one of the most complex waste
ferrous scrap. In a French study that examined streams requiring management since the equip-
copper contamination in secondary steel produc- ment includes a substantial fraction of the peri-
tion Russo et al. (1999) found that dismantling odic table elements (MCC, 1996). Problems
steps prior to shredding processes make it possible arising with the establishment of adequate waste
to separate 90% of copper in ELV during the treatment facilities include the rapidly changing
disposal procedures. Overall, we assume that 70% nature of electrical and electronic equipment, and
of copper in the ELV stream is recovered and the negative environmental impact that results
recycled. from the treatment of these wastes due to the
presence of hazardous substances in WEEE.
3.3.2. Construction and demolition waste Though mechanical separation standards make it
The following represents a summary of data possible to recycle 80–90% of WEEE, the cost of
collected by the European Commission (1999) on recycling most equipment is higher than the value
 M. Bertram et al. / Ecological Economics 42 (2002) 43–57 51

of the recovered metal (Langer, 1994). The ab- “ Data on copper concentrations in HW and IW
sence of dedicated collection and take-back sys- are not available. We assumed 500 mg/kg for
tems for this waste group means that a significant HW (similar contents as MSW) and 200 mg/kg
proportion of WEEE in Europe is still disposed of for IW.
as part of the municipal waste stream (e.g. 0.52– With the exception of MSW and SS, data
1.6 kg/c y in Germany (ARGUS, 1996)), is stored availability for waste composition is not satisfac-
in households or illegally dumped (O8 kopol, 1998). tory. We believe that at least the order of magni-
Nevertheless, increased recycling efforts have tude of our copper content estimates are correct;
taken place over the last few years. White prod- however, in order to develop strategies for recy-
ucts (high metal content) are usually recycled in cling and optimized waste management additional
automotive shredder facilities, and the same com- and more accurate data are needed to verify the
panies that produce computers do some recycling assumptions made in the present work.
of them. A recycling efficiency of copper in The copper content of scrap varies between 2%
WEEE of :50% is assumed. (copper sludge) and 99% (offal of pure copper
semi-finished products). Electronic scrap, which is
3.4. Copper content of wastes the mechanically pre-treated WEEE, represents
the largest secondary copper resource of old
The copper contained in each of the solid waste scrap. Its copper content ranges from 15 to 20%.
flows is the product of the overall flow magnitude Copper-brass radiators from ELV have a copper
and the average copper concentration, which is concentration of 60–65% (IPPC, 2000).
determined as follows:
“ The copper concentration of MSW varies be- 3.5. Losses
tween 360–700 mg/kg in Europe. This range is
derived from several investigations carried out In this paper we consider the following losses of
in Austria and Switzerland by using incinera- copper to the environment: (a) emissions from
tors as an analytical tool to determine MSW incinerators and landfills to the atmosphere and
composition (Brunner and Mönch, 1986; hydrosphere (surface waters and groundwater),
Schachermayer et al., 1995; Morf et al., 1997; (b) application of SS in agriculture, (c) use of
Belevi and Mönch, 2000; Morf et al., 2000). incinerator bottom ash (‘slag’) in road construc-
“ Few published data exist for elemental compo- tion. Emissions from mechanical separation facili-
sition and concentration in C&D waste ties are not considered and are assumed to be
streams. The copper content of C&D is ob- negligible.
tained from a material flow analysis of a full-
scale, construction-waste sorting plant 3.5.1. Incineration
(Brunner and Stämpfli, 1993): 670 mg/kg. About 22% of the MSW stream was incinerated
“ WEEE is composed of consumer and capital in STAF-Europe in 1994 (CEC, 2000). Countries
goods. Consumer goods have an average cop- in which incineration constitutes an important
per content of 4.6%. The recorded copper con- part of waste treatment are Denmark, France,
tent of capital goods is 14.5% (WEKA, 1999). Luxembourg, Sweden, and The Netherlands. The
“ The copper content of ELV is :1.4% (Ke- incinerators in those countries are mainly state-of-
oleian et al., 1997); this corresponds to vehicles the-art facilities obeying the European limit for
built in the 1980s. The copper concentration of particulate matter emissions of 10 mg/m3 (EPC,
larger vehicles (trucks, busses) is assessed to be 2000). In order to assess the rate of copper emis-
0.5%. The copper contained in ELV was calcu- sion to the atmosphere an actual particulate mat-
lated as the weighted average of copper in cars, ter emission of 5 mg/m3 is assumed. For example,
busses, and trucks. incinerators in Austria achieve levels B 2 mg/m3.
“ The copper concentration in SS varies between The specific volume of off-gas is : 5 m3/kg
200 and 640 mg/kg dry matter (CEC, 2000). MSW. This yields a specific particulate matter
52 M. Bertram et al. / Ecological Economics 42 (2002) 43–57

emission of 25 mg/kg MSW. Copper concentra- Finnveden (1996) gives a survey of research re-
tions in emitted fine particulate matter vary be- sults on emission factors (transfer coefficients)
tween 1000 and 5000 mg/kg (Schachermayer et showing a range between [0.00001– 0.0005] for the
al., 1995; Morf et al., 1997; Rechberger, 2002). survey time period (several decades up to one
Assuming an average value of 3000 mg/kg results century). Typical values for copper in leachates of
in a specific copper emission of 0.075 mg/kg MSW landfills are 0.02–0.1 mg/l (Jensen and
MSW or a transfer coefficient of copper into the Christensen, 1999); higher values up to 0.5 mg/l
atmosphere of 0.00015 [kg Cu emitted per kg Cu may be found during the first 10 years of opera-
in MSW]. Hence, it is reasonable (an overesti- tion (Döberl et al., 2002). Assuming an annual
mate) to apply a transfer coefficient range for precipitation of 1000 mm/y, an average evapora-
atmospheric copper emissions from incinerators tion of 50%, a depth of 15 m3 (z=1.3Mg/ m3),
of [0.001–0.0001]. If incinerators are equipped and an average leachate concentration of 0.1 mg
with wet scrubbers for air pollution control the Cu/l over 100 years yields a transfer coefficient of
scrubber waters are physically and chemically : 0.0005. Because this rough estimate does not
treated. The purified wastewater is either evapo- consider any leachate treatment, a value of 0.001
rated (no aqueous emissions) or discharged into can be regarded as the upper limit for aqueous
receiving water. Emissions of copper into the hy- copper emissions of landfills over 100 years.
drosphere for state-of-the-art wastewater treat-
ment are estimated to have transfer coefficients in 3.5.3. Application of sewage sludge and utilization
the range [0.000001– 0.0001] (Schachermayer et of incinerator ashes
al., 1995; Morf et al., 1997; Rechberger, 2002). About 30% of SS was applied to agricultural
There are no data available on the co-incineration land in 1994 (CEC, 2000). The use of this disposal
of combustible wastes in (a) industrial furnaces method varies considerably among the countries
such as rotary kilns of the cement industry and studied, from 11% in Ireland to 67% in Denmark.
fluidized bed furnaces of the paper industry and Germany, the Netherlands, and Denmark use 60–
(b) coal-fired power plants. Generally, because 90% of their incinerator bottom ash as an aggre-
these facilities are not equipped with state-of-the- gate in road base material (European
art air pollution control devices, emissions may be Commission, 2000). About 90% of the copper is
relevant if wastes with composition similar to transferred to the bottom ash in incinerators
MSW are utilized for energy recovery. This uncer- (Morf and Brunner, 1998; Belevi and Mönch,
tainty regarding emissions from co-incineration 2000). Hence bottom ash represents a sink for
becomes increasingly significant since the Eu- copper. There have been attempts to use incinera-
ropean Union’s policy on waste management re- tor ashes for other purposes, for example as a
lies on co-incineration of wastes in order to secondary resource for cement and brick produc-
reduce the amount of biodegradable wastes that tion (Dhir et al., 2000). All these applications
are landfilled, a goal of the new landfill directive result in losses/dissipation of copper. Further in-
(CEC, 1999). vestigations and more analyses are necessary in
order to determine the relevance of these losses.
3.5.2. Landfill
Assessing the average copper emissions of 3.6. Copper budget for European waste
landfills is a nearly impossible task. Emissions are management
a function of (a) technology (leachate contain-
ment, collection, and treatment), (b) the climate Fig. 3 illustrates the copper flows in the ‘STAF-
(precipitation, evaporation, wind velocity, temper- Europe’ waste management system for 1994. A
ature, and humidity), and (c) the surface design total of 920 Gg/y of copper within MSW, C&D,
(permeability, porosity, capacity, mightiness, re- ELV, WEEE, SS, HW, and IW and 300 Gg/y of
cultivation). Hence, site-specific assessments of imported old scrap enter waste management.
emissions should be carried out if possible. Around 48% of domestically generated copper
 M. Bertram et al. / Ecological Economics 42 (2002) 43–57 53

waste is recycled; the remaining 52% is either half (56%) of copper scrap is re-melted directly by
landfilled or lost to the environment. Roughly brass mills, wire rod producers, foundries, and
60% of copper scrap used in European secondary ingot producers; the remaining 44% goes back to
copper production comes from prompt manufac- the smelting and refining processes. There is con-
turing scrap and 40% from old scrap. The 740 Gg siderable agreement in comparing the data on old
of old scrap is made up of 300 Gg imported and scrap (734 Gg/y) obtained from literature (Inter-
440 Gg domestic scrap. The high use of scrap in national Copper Study Group, 1999) and that
Europe is mainly within the large manufacturing calculated based on the method described in Sec-
base from which new scrap is generated. Only tion 3.3 (720 Gg/y).
23% of scrap used in domestic secondary produc- Table 4 shows annual waste generation cate-
tion comes from old non-imported scrap. About gories in STAF-Europe as well as their relative

Fig. 3. Copper flows in ‘STAF-Europe’ associated with waste management. The overall recycling efficiency is 48%. Emissions are
B 0.05%, and losses due to SS application and utilization of incinerator bottom ash for road construction are B 0.5%. All units are
in Gg/y (1000 metric tons/year).

Table 4
Waste generation and copper contents for STAF-Europe in 1994

Waste category Waste generation Mass fraction Copper content Copper content Copper fraction

Tg/y % % Tg/y %

MSW 165 21.5 0.05 0.083 9.0

C&D 184 24.0 0.07 0.129 14
WEEE 2.9 0.4 13 0.377 41
ELV 27a 3.5 0.9 0.243 26.4
SS 7.7 1.0 0.037 0.0028 0.3
HW 30 3.9 0.05 0.015 1.6
IW 350 45.7 0.02 0.070 7.6
Total 767 100 0.12 0.92 100

a
The weight of an average car in Europe is estimated to be 1000 kg (ARGE-Altauto, 2000), for busses and trucks 12 000 kg is
assumed.
54 M. Bertram et al. / Ecological Economics 42 (2002) 43–57

Table 5
Waste generation, copper content and uncertainties

Waste category Waste generation Copper concentration

Base value Uncertainty Base value Range

Tg/y % % %

MSW 165 920a 0.05 0.04–0.06a

C&D 184 +100b,a 0.07 0.05–1a
WEEE 2.9 940a 13 6–14a
ELV 27 920a 0.9 0.8–1.0a
SS 7.7 920a 0.037 0.02–0.04a
HW 30 910a 0.05 0.02–0.08a
IW 350 930a 0.02 0.01–0.05a

a
Uncertainties and ranges for each stream were judged from informed estimates.
b
Possible uncertainties for C&D waste generation are up to +200% (Section 3.1.4).

Fig. 4. Effects of uncertainties in waste generation and copper content on the copper flow entering the waste management system.
Columns represent the flow of copper entering waste management (base flow, 920 Gg/y), bars the sensitivity in the base flow due
to changes in (a) waste generation and (b) copper content of each waste stream.

importance for copper. IW, C&D, and MSW are to see whether any of our conclusions might be
the most relevant flows from a mass flow point of affected by these uncertainties. To evaluate this,
view. These wastes count for 90% of the total two key parameters (waste generation and copper
waste generation. Regarding the waste categories content) were selected to test the sensitivity of
for their potential for copper recovery, WEEE, estimates of copper entering the waste manage-
ELV, and C&D are most relevant. WEEE, with ment system on the ultimate results. The waste
its growing generation rate, is a particularly at- generation rate and copper concentration for each
tractive future resource for secondary copper. of the seven waste streams was varied individually
while keeping other parameters constant. Table 5
3.7. Sensiti6ity analysis shows the assumed uncertainty ranges based on
sources provided in Sections 3.1 and 3.4. In the
The uncertainties in this analysis are admittedly base scenario 920 Gg/y of copper (hereafter base
substantial, and are best regarded as initial char- flow) enters the waste management system (Fig.
acterization under the classification framework of 4).
Graedel et al., 2002. Nonetheless, it is important Fig. 4a shows how changing the waste genera-
 M. Bertram et al. / Ecological Economics 42 (2002) 43–57 55

tion rate over the given uncertainty range for each be recycled and that the European Union should
stream affects the base flow. Changes in the C&D attempt to attain high efficiencies for all its mem-
and WEEE streams affect the base flow the most. ber countries. With high certainty, WEEE is al-
For example, changes in the WEEE generation ready the most important source for old copper
rate can potentially change the base flow by 9 scrap. Considering the predicted annual growth
150 Gg/y. Variations in the ELV, MSW, IW rate of 5–10% for this category for the next 10
generation rates change the base flow by lesser years (Christiansen and Munck-Kampmann,
amounts; 950 Gg/y (ELV), and 9 20 Gg/y 2000), it is time to develop recycling strategies for
(MSW, IW). Changes in SS and HW generation WEEE. In other words, it would be important to
do not have a significant effect on the base flow. investigate whether, for example, separate collec-
Fig. 4b shows how variation in the copper tion and mechanical dismantling of WEEE is
content of each stream affects the base flow enter- superior to a system where consumer appliances
ing waste management. Here, the WEEE stream and MSW are collected together and sent to an
has the greatest influence (+30/ − 200 Gg/y). The incinerator with adjacent recovery of copper from
IW variation can change the base flow by + 100/ the bottom ash. To do this, environmental and
− 40 Gg/y. Variation in the C&D and ELV economic aspects, including collection, transport,
streams have a small influence on the base flow, treatment, and landfill, must be considered.
while SS and HW have essentially no influence. Should it turn out that WEEE treatment yielding
Overall, applying single-variable sensitivity analy- high recovery rates for copper and other precious
sis illustrates that assumed uncertainties in waste substances is not yet economic, WEEE should be
generation and copper concentration have a rela- at least stored in separate compartments at a
tively small effect on the total domestic copper landfill site in order to generate a stock for future
waste flow that is derived. Variation in the copper recovery activities.
concentration in the WEEE stream has the The copper content of MSW, SS, and perhaps
biggest effect on the base flow, which can decrease also of IW and HW is small and their relevance
as much as 20%. Nevertheless, it is worth men- for copper is not very significant. Nevertheless,
tioning that we believe uncertainties in C&D efforts should be made to reduce the loss of
waste generation to be as high as + 200%, which copper via these waste categories. This applies
would increase the base flow by 260 Gg/y ( + especially to wastes with high generation rates,
28%). such as IW and MSW. In order to develop sensi-
ble strategies, more information on the sources of
copper in these wastes is needed.
4. Conclusions Generally, the budget for copper shows that its
distribution pattern in the waste management sys-
The results presented here show that it is feasi- tem is quite favorable, making high recovery rates
ble to develop a copper budget for the European theoretically possible by applying appropriate
waste management system. However, these results technology to selected waste streams. Such
must be considered with caution since uncertain- changes are generally much easier and quicker to
ties in waste generation rate and composition for realize than structural changes such as the re-
some waste categories are rather high. Additional design of industrial and commercial products.
analysis is needed to confirm the findings. It
would be useful to establish a central, public data
bank for collecting all available information on Acknowledgements
composition and generation of wastes. Such a
resource would provide fundamental information This research was funded by the U.S. National
for engineers and decision-makers. Science Foundation under grant BES-9818788. H.
The recycling efforts of several European coun- Rechberger thanks the Max Kade Foundation,
tries prove that a high percentage of copper can Inc. for post-doctoral support.
56 M. Bertram et al. / Ecological Economics 42 (2002) 43–57

References European Commission, DG Environment, 2000. A study on

the economic valuation of environmental externalities from
AAMA, American Automobile Manufactures Association, landfill disposal and incineration of waste, Brussels.
1996. World motor vehicle data, 32nd ed. AAMA, Wash- European Commission, Eurostat, 2001. Statistical pocket
ington, DC. book: EU energy & transport in figures 1970 – 1999. http://
ARGE-Altauto Arbeitsgemeinschaft Altauto, 2000. 1. Moni- europa.eu.int/comm/energy – transport/en/etf – en.html.
toring Bericht gemäß Punkt 3.6 der Freiwilligen Selbstver- European Topic Center on Waste and Material Flows, 2001.
pflichtung zur umweltgerechten Entsorgung von The activity of the ETC/WMF on source of wastes e.g.
Altfahrzeugen (Pkw) im Rahmen des Kreislaufwirtschafts-/ manufacturing. http://www.etc-waste.int.
Abfallgesetzes. www.arge-altauto.de, (in German). Eurostat Statistical Office of the European Communities,
ARGUS, 1996. Stand der Entsorgung von elektrischen und 1997. Material flow accounting: experience of statistical
elektronischen Kleingeräten in der Bundesrepublik institutes in Europe, Directorate B: Economic statistics and
Deutschland. Final Report of Research Project No. 103 10 economic and monetary convergence, Luxembourg.
603 for the Umweltbundesamt Berlin, Arbeitsgruppe Finnveden, G., 1996. Solid waste treatment within the frame-
Umweltstatistik an der Technischen Univeristät Berlin AR- work of life cycle assessment. Metals in municipal solid
GUS e.V., Berlin, (in German). waste landfills. Int. J. LCA 1 (2), 74 – 78.
Belevi, H., Mönch, H., 2000. Factors determining the element Frei, M., Küng, A., Schuler, B., 1993. Kupfer-Flüsse in der
behavior in municipal solid waste incinerators. 1. Field Automobilbranche. Vertiefungsblock-Arbeit ETHZ,
studies. Environ. Sci. Technol. 34, 2501 –2506. Zürich, (in German).
Brunner, P.H., Mönch, H., 1986. The flux of metals through a Glenck, E., Lahner, T., Arendt, M., Brunner, P.H., 1997.
municipal solid waste incinerator. Waste Manage. Res. 4, Baurestmassen in Oberösterreich, Stoffbilanzen der
105– 119. Bauwirtschaft (BRIO-S), Institute for Water Quality and
Brunner, P.H., Stämpfli, D., 1993. Material balance of a Waste Management, Vienna University of Technology,
construction waste sorting plant. Waste Manage. Res. 11, Vienna, (in German).
27– 48. Gordon, R.B., 2002. Production residues in copper technolog-
CEC, Commission of European Communities, 1999. Council ical cycles. Resour. Conserv. Recy., Accepted for
directive on the landfill of waste of April 26, 1999/31/EC, publication.
Brussels. Graedel, T.E., Bertram, M., Fuse, K; Gordon, R.B., Rech-
CEC, Commission of the European Communities, 2000. Re- berger, H., Spatari, S., 2002. The characterization of tech-
port from the Commission to the Council and the Eu- nological copper cycles. Ecol. Econ., 42, 9 – 26.
ropean Parliament of the implementation of community International Copper Study Group, 1999. Copper Bulletin, 6
waste legislation (Directive 75/442/EEC on waste, Direc- (11). Lisbon, Portugal, November.
tive 91/689/EEC on hazardous waste, directive 75/439/ IPPC, International Pollution Prevention and Control, 2000.
EEC on waste oils and Directive 86/278/EEC on sewage Reference document on best available techniques in the
sludge) for the Period 1995 –1997, Brussels. non-ferrous metals industry. Institute for Prospective Tech-
Christiansen, K.M., Munck-Kampmann, B., 2000. Waste an- nology Studies, Seville, Spain, May.
nual topic update 1999, Topic report 2/2000. European Jensen, D.L., Christensen, T.h.H., 1999. Colloidal and dis-
Environmental Agency, Copenhagen. solved metals in leachates from four Danish landfills. Wa-
Dhir, R.K., Dyer, T.D., Paine, K.A. (eds.), 2000. Use of ter Res. 33 (9), 2139 – 2147.
incinerator ash. Proceedings of the International Sympo- Johnstone, N., 1998. The implications of the Basel Convention
sium on Sustainable Construction, University of Dundee, for developing countries: the case of trade in non-ferrous
20– 21 March 2000, Telford, London. metal-bearing waste. Resour. Conserv. Recycling 23
DKI Deutsches Kupfer Institute, 2001. Recycling von Kupfer- (1998), 1 – 28.
werkstoffen, (in German). Keoleian, G.A., Kar, K., Manion, M., Bulkley, J., 1997.
Döberl, G., Huber, R., Brunner, P.H., Eder, M., Pierrard, R., Industrial Ecology of the Automobile: A Life-Cycle Per-
Schönbäck, W., Frühwirth, W., Hutterer, H., 2002. Long- spective. SAE International, Warrendale, PA.
term assessment of different waste management options — Langer, B.E., 1994. Recycling von Elektronikschrott. Metall.
a new, integrated and goal-oriented approach. Waste 48 (11), 880 – 885 in German.
Management and Research. In press. MCC, Microelectronics and Computer Corporation, 1996.
EEA, European Environmental Agency, 1997, Sludge treat- Electronics Industry Environmental Roadmap, Austin,
ment and disposal: Management approaches and experi- TX, p. 195.
ences. Environmental Issues Series No. 7, Copenhagen. Morf, L., Ritter, E. and Brunner, P.H., 1997. Güter- und
EPC, European Parliament and Commission, 2000. Directive Stoffbilanz der MVA Wels. Final Report, Institute for
on the incineration of wastes of December 4, 2000/76/EC, Water Quality and Waste Management, Vienna University
Brussels. of Technology, Austria, (in German).
European Commission, 1999. Construction and demolition Morf, L.S., Brunner, P.H., 1998. The MSW incinerator as a
waste management practices, and their economic impacts. monitoring tool for waste management. Environ. Sci.
Report to DGXI, Brussels. Technol. 32, 1825 – 1831.
 M. Bertram et al. / Ecological Economics 42 (2002) 43–57 57

Morf, L.S., Brunner, P.H., Spaun, S., 2000. Effect of operating Spatari, S., Bertram, M., Fuge, D., Fuse, T., Graedel, H. and
conditions and input variations on the partitioning of Rechberger, H., 2002. The contemporary European copper
metals in a municipal solid waste incinerator. Waste Man- cycle: One year stocks and flows. Ecol. Econ. 42, 27 – 42.
age. Res. 18, 4 – 15. UBA, Umweltbundesamt Deutschland, 1999. Sonderabfall,
Müller, K., 1995. Waste from Road Vehicles in the Member http://www.umweltbundesamt.de/uba-info-daten/daten/
States of the European Union. Internal Working Docu- sonderabfall.htm (in German).
ment, Doc. ENV/95/13, Eurostat, Luxembourg. UBA, Umweltbundesamt Deutschland, 2000. Grenzüberschre-
O8 kopol, 1998. Collection targets for waste from electrical and itende Abfallverbringung, http://www.umweltbundesamt.
electronic equipment (WEEE). Report compiled for the de/uba-info-daten/daten/grenzueberschreitende-abfallver-
Directorate General (DG XI) Environment, Nuclear Safety bring.htm (in German).
and Civil Protection of the Commission of the European WEKA, 1999. Richtiger Umgang mit Abfällen, In: Praxis
Communities Contract N° B4-3040/97/000800/MAR/E3, Handbuch Plus. Fachverlag für technische Führungskräfte
Hamburg, Germany. GmbH, Augsburg, Germany, Section 3.2.11 (in German).
Rechberger, H., 2002. Informed estimate based on unofficial Welslau, G., Kraus, H.H., 1998. Elektronik Schrott (WEEE).
reports about emission tests on behalf of local authorities Europäisches Parlament, Direktion B, Abteilung für
as well as the work as an expert in environmental impact Umwelt, Energie und Forschung, Luxemburg (in German).
assessment of incinerators, in press. World Bureau of Metal Statistics, 1995. Metal Statistics 1984 –
Russo, P., Birat, J.P., Aboussouan, L., Geoffroy, X., Quelier, 1994, 82nd ed. Ware, England.
M., Backert, S., 1999. Upgrading scrap quality by improv- Zhang, S., Forssberg, E., 1999. Intelligent liberation and clas-
ing shredder operation, 4TH ASM International Congress sification of electronic scrap. Sweden Powder Technol. 105,
on Recycling of Metal, 17 –18 June 1999, Vienna. 295 – 301.
Schachermayer, E., Bauer, G., Ritter, E. and Brunner, P.H., ZVEI, Zentralverband Elektrotechnik- und Elektronikindus-
1995. Messung der Güter- und Stoffbilanz einer Müllver- trie e.V., 1992. Entsorgung von gebrauchten elektrischen
brennungsanlage. Federal Environment Agency, Monogra- und elektronischen Geräten– Leitfaden zur Entwicklung
phien Bd.56, Vienna, (in German). eines Konzeptes. Frakfurt/Main (in German).