The Copper Balance of Cities

Material management faces a dual challenge: on the one hand satisfying large and increasing demands for goods and on the other hand accommodating wastes and emissions in sinks. Hence, the characterization of material flows and stocks is relevant for both improving resource efficiency and environmental protection. This article focuses on the urban scale, a dimension rarely investigated in past metal flow studies. We compare the copper (Cu) metabolism of two cities in different economic states, namely, Vienna (Europe) and Taipei (Asia). Substance flow analysis is used to calculate urban Cu balances in a comprehensive and transparent form. The main difference between Cu in the two cities appears to be the stock: Vienna seems close to saturation with 180 kilograms per capita (kg/cap) and a growth rate of 2% per year. In contrast, the Taipei stock of 30 kg/cap grows rapidly by 26% per year. Even though most Cu is recycled in both cities, bottom ash from municipal solid waste incineration represents an unused Cu potential accounting for 1% to 5% of annual demand. Nonpoint emissions are predominant; up to 50% of the loadings into the sewer system are from nonpoint sources. The results of this research are instrumental for the design of the Cu metabolism in each city. The outcomes serve as a base for identification and recovery of recyclables as well as for directing nonrecyclables to appropriate sinks, avoiding sensitive environmental pathways. The methodology applied is well suited for city benchmarking if sufficient data are available.

Supply-restrictions of Cu in China lead to the identification of potential reservoirs of secondary Cu resources in urban infrastructure. (Zhang et al. 2012) Stockholm (SE) Pathways and stock are investigated comprehensively. (Sörme et al. 2001a;Bergbäck et al. 2001;Sörme et al. 2001b) Linkoping (SE) Stocks-in-use and hibernating stocks in the telecommunication network are estimated in view of the recovering potential. (Krook et al. 2011) Cape Town (ZA) Stocks-in-use of major appliances are linked with product lifetimes in order to predict future waste flows and recovering potentials.
(van Beers and Graedel 2003) emissions and/or waste flows Nanjing (CN) Cu impacts on road sediments are analyzed. (Zuo et al. 2012) Stockholm (SE) The pathway analysis of diffuse emissions is driven by increased concentrations in Stockholm's receiving sediments. (Sörme and Lagerkvist 2002;Sörme et al. 2001b) Urban catchment (UK) Water quality is assessed by diffuse emissions entering waste water systems. (Rule et al. 2006) Sofia (BG), New Heaven (USA) Cu discard, reuse and recovery fluxes and rates are compared between two cities. (Dimitrova et al. 2007) Villach ( AUT) The fate of diffuse emissions is determined in order to assess environmental risks. (Rebernig 2007) End-of-Pipe plants Vienna (AUT) -Waste water treatment plant: Heavy metal flow ratios on plant level demonstrate the separation efficiency. High separation rates improve environmental performance. -Incinerator: Monitoring of heavy metal in residues reveals the temporal evolution of concentrations in household waste. (Kroiss et al. 2008; Morf and Taverna 2006) Hinwil (CH) Incinerator: Residues were analyzed in order to estimate recovering potential of precious metals and rare earth elements. (Morf et al. 2013) Urban Soil Edinburgh, Dundee Historical early warning of slow-poisoning of urban soils. (Purves 1966) Oslo (SE) Urban transactions are used to demonstrate the influence of urbanization of chemical soil quality. (Reimann et al. 2011) Vienna (AUT) Estimation of geogenic background values (relevant for legislation) based on geochemical patterns in urban soil. (Pfleiderer 2011) Taipei (TW) Identification of toxic contaminants in urban top soil layers. (Jien et al. 2011) S-5  Figure S1 displays the generic stock and flow chart. Figure S1: Generic Cu flow model on 1 st level. It covers 9 city internal processes of which 6 represent mainly anthropogenic activities (dark grey boxes) and 3 stand for environmental media (light grey boxes). Exterior processes are splitted in the supply and export as well as receiving waters in the hinterland. Regarding nomenclature, the flow acronyms refer to the type of flow (first three letters), to the source process (second three letters and to the sink process (last three letters). The stock acronyms refer to the type of stock only.

City selection and city characteristics
S-6 Figure S2: Generic Cu flow model on 2 nd level. It disaggregates the "Waste Management System".

Model equations and data acquisition
This section provides a compilation of flow and stock calculations as well as insights into data acquisition. In Vienna, total deposition rate was measured with an rate of 148,7 g Cu/ha/yr on arable land close to Vienna (Spiegel 2003), the dry deposition was measured in town with an average rate of 7,7 g Cu/ha/yr (Kalina et al. 2000). City wide extrapolation on an area base results about 6.6 t Cu/yr, from which 3.5 tons are allocated to building area, 2.8 t Cu/yr to green space (ADE PBLUPV), 0.3 t Cu/yr to open water bodies (ADE PBLUHY ) and 0.9 t Cu/yr to traffic surfaces (ADE PBLTEC).
In Taipei, one research was conducted that sampled and analysed the content, source, and transportation mechanism of metals in atmospheric deposition. Dry deposition rate was 3.8 mg/m 2 /yr The import of products stands for consumer goods and pre-products. Both are finally consumed or processed by trade and skilled labor.
In Vienna, we have chosen four main products groups that ended up in the urban production sector or final consumption.
(1) Cu containing goods like sheet metals, wires, cable pies and pre-products were taken from a national Cu flow study (Daxbeck et al. 2006). To allocate the flows on different sectors (buildings, infrastructure, production industries), we used global market shares of specific Cu products (ICF 2001). The number of employers in the construction and production sector (Hauptverband der österreichischen Sozialversicherungsträger SV 2011; Statistik Austria 2009c) were used as proxy to downscale from the national to the urban level. Therefore, 4,600 t Cu/yr are allocated to the production industry and is part of the product flow (PRO EXAECO). 11,800 t Cu/yr were allocated to the construction sector (follow sec. 2.2.2.1.3).
(2) Electrical and electronic appliances (EEA) are monitored when they placed on the market and when they turn to waste. To estimate the flow, we used the difference of imports and exports on EEA from the regional, mass based trade statistic (Magistratsabteilung 05 2011) and a mean Cu content based on a S-7 basket of 11 representative goods (Truttmann et al. 2005;Oguchi et al. 2011;Hausmann 2005;Wittmer 2006). To estimate the allocation, we used the national sale statistics for household consumption, the population ratio as proxy for downscaling to the regional level (EAK 2009 In Taipei, top-down and bottom-up approaches were adopted to estimate of Cu in products. Since "import" generally happens at a national level, there was no urban statistics of imported goods into the city. Hence, the proxy of total retail sales of consumer goods and areas of factories were used to downscale to the urban level for the imported products (PRO EXAECO), as 16,000 t/yr of Cu contained. Because there is lacking statistics of products consumption within the city, we used data of recycling to estimate backward. In Taiwan, people can recycle household products with certain recycling fee as feedback. Therefore, bottom-up method was adopted to estimate the quantity of consumer goods, with the assumption that Taiwanese purchase a new product when they discard the old one. In terms of Cu-based products, two categories were selected: Waste Electrical and Electronic Equipment (WEEE) and Waste Computer Appliances (WCA). We took statistics from Recycling Fund Management Board, which is an official organization in charge of executing the recycling fund system. Estimation showed that around 5,000 t Cu/yr was consumed in household commodities (PRO ECOPHH).
In Vienna, the imported Cu as construction material including pipes for plumbing and heating, sheets for roof and outdoor applications as well as cables and wires like building wires for electric currents and telecommunications. As calculated in sec. 2.2.2.1.2, 11.800 t Cu/yr (CON EXAECO) enter the construction sector, of which 9,400 t Cu/yr accumulates in buildings and 2,400 t Cu/yr infrastructure (CON ECOTEC). Specific land use categories were used to add 6,600 t Cu/yr to the PHH (CON ECOPHH) and 2,800 t Cu/yr to industrial and business buildings. If it comes to construction waste, the total flow was categorized in three groups: (1) Demolition Material amounted 1.8 Mio t/yr of which the Cu fractions are mostly unknown. So, we picked out the top 5 construction waste flows from the regional waste statistic (Wiener Umweltschutzabteilung MA22 2011) and multiplied them with corresponding Cu contents from literature (Brunner and Stampfli 1993;Schnöller et al. 2010;Arx 2006). So, 800 t Cu/yr were estimated in demolition material. Allocation of flows to the waste generators PHH, ECO and TEC was done by a sector based consumption ratio and the land use ratio. Direct disposals were neglected after the economic benefit of recycling omits landfilling. (2) Cu Scrap was collected with an rate of 3,200 t Cu/yr (Wiener Umweltschutzabteilung MA22 2011) and is estimated to be fully recycled. Allocation followed the same routine as for the demolition material. (3) Collected cables were explicit recorded in the local waste statistics. Hence, we used measured Cu contents in cables (Skutan 2008) and the same allocation proxy as for the demolition material. The total construction waste flows were 3,000 t Cu/yr from private households (CWA PHHCTS), 1,300 t Cu/yr from industry and business (CWA ECOCTS) and 700 t Cu/yr from technical infrastructure (CWA TECCTS).
In Taipei, we determined the imported construction material from infrastructure ad buildings import in Taiwan. The economy and transportation budget for infrastructure and the newly built floor area for S-8 buildings were included to downscale statistics from Taiwan to Taipei. The two categories contributed to 45.3% of the total Cu consumed in Taipei which was 13,000 t/yr (CON EXAECO). We allocated the imported construction material to construction material consumption in economic sectors, private households and infrastructure by the proportion of land-use area. Cu consumed in households and infrastructure was then estimated to be 8,000 t/yr (CON ECOPHH) and 4,000 t/yr (CON ECOTEC), respectively. For construction waste, construction and demolition waste (C&D waste) and total demolished floor area of the year were considered. Allocation for households, infrastructure an economy sectors were determined by the corresponding area of land-use in Taipei. Results show that Cu in construction waste are 150 t/yr from private households (CWA PHHCTS), 50 t/yr from infrastructure (CWA TECCTS) and 3 t/yr from economy sectors (CWA ECOCTS). The relatively low amount of construction waste produced from infrastructure and economy sectors may be due to the longer lifespan of the infrastructure and buildings. In Vienna, the modal split in the traffic sectors allocates 1/3 of all trips to public transport, 1/3 to private cars and 1/3 to non motorized traffic like cycling and walking. So, private and public transport fleets are of equal importance for Cu calculations. To estimate the inflow, we used official registration statistics for vehicles like cars, busses, lorries and motor bikes (Statistik Austria 2008), the rolling stock alteration of the public transport provider (Lebhart 2010), and the and the average Cu content per vehicle found in the literature (European Copper Institute (ECI) 2011; Bertram et al. 2002;Struckl 2007). This results an inflow of 2,200 t Cu/yr (VEH ECOVEH). The total outflow has not been statistically reported. So, the whereabouts of EOL cars can't be determined at full scale. We carried out a national car balance in order to estimate the number of unidentified cars. Hence, the Viennese car ownership of 15.3% acts as proxy for regional scaling (Statistik Austria 2008) and an average Cu content per car was used to estimate Cu flows. This results an outflow of 880 t Cu/yr of which 240 t Cu/yr enter a shredder (WAS VEHCTS), 140 t Cu/yr t are legally exported to foreign countries (VEH VEHECO) and 500 t Cu/yr remain unidentified in terms of their whereabouts (VEH UID).
In Taipei, private vehicles dominate transportation, especially the sedans and scooters. Since statistics of newly sold vehicles were absent, Cu estimation was top-down from the national data. Vehicles which are ready to be in use were assumed to be the sales volume of cars and scooters/motorcycles of the study year. By the Taipei-Taiwan ratio of newly registered vehicles and Cu content 1.4% , sales volumes in Taipei was estimated to be 500 t/yr of Cu (VEH ECOVEH). We assumed that all end-of-life vehicles (ELV) were treated as recyclable waste. Since data of recycled ELV in Taipei was lacking, we used statistics of recycled ELV in Taiwan and Taipei-Taiwan ratio of newly registered vehicles to top down the estimation, as a result of 200 t/yr of Cu in waste vehicles (WAS VEHCTS). Amount of registered vehicles difference between 2008 and 2009 was assumed to be the exported ELV, since the change of ELV was unknown and unclear. Hence, 90 t Cu/yr in ELV (VEH VEHECO) was estimated to leave the city boundary. In Vienna, we considered three main emission sources. First, brake wears from low duty vehicles was estimated according to the fleet mileage (Holzapfel and Riedel 2011), average total wear rates (Winther and Slento 2010) and Cu concentrations in brake linings representing the Austrian fleet (Figi et al. 2010). Finally we used transfer coefficients from the Netherlands National Emission Inventory (Hulskotte et al. 2006;Oonk et al. 2005) in order to examine the flows towards ambient air, road surfaces and vehicle depositions.

S-9
Second, brakes and wheels from rolling stock (trams, trains) emit Cu too. To estimate the flows, we processed the individual network lengths (ÖBB-Holding AG 2008;Wiener Linien 2008), normalized Cu abrasion rates from railways (Burkhardt et al. 2005) and the transfer coefficients from individual sources to ambient air, railroad, surrounding and vehicle deposits (Müller et al. 2008). Third, catenary wear contains 99.5-99.9% Cu on a mass basis. To estimate the wear entering urban sinks, we used the track lengths, corresponding cross sections at the time of installation and replacement, an average lifetime of 40 years and transfer coefficients were derived from a SFA study that balanced Cu flows on 1 kilometer railroad (Müller et al. 2008). Fourth, flows from roofs were estimated according to ) with a rate of 1.5 g Cu/m 2 /yr. The Cu roof area was gathered from local tin smiths (Wocilka and Höfner 2011), the total roof area and average roof slope was delivered by city authorities who once plotted the inclinations for estimating solar energy potentials (Kubu 2011), the regional precipitation rate (Lebhart 2010), the SO2 concentration in precipitation was equally set with monitoring data in ambient air (Augustyn et al. 2010), and the pH value with 5. Allocating the single results that a) the vehicles release 5 t Cu/yr to the ambient air (PAE VEHPBL), 2.2 t Cu/yr occur as road debris (PAE VEHTEC), and b) the catenaries release 0.6 t Cu/yr to ambient air (PAE TECPBL), 2.0 t Cu/yr to surrounding urban surfaces like roads and facades. Cu deposits on vehicles were not taken into further account.
Taipei City is a rather young Asian urban system and materials used in roof are different from those in Europe. Cu flows of trains, trams and roofs were not taken into account. Airborne particulates from light and high duty vehicles were considered in break wear estimation, with mileage (Department of Statistics 2007) and the wear rate which was the same as in Vienna. Mass of Cu in tire wear were estimated through the topdown data of registered vehicles in Taipei, the tire wear factor, and Cu content in the tire wear emission. Around 6 t Cu/yr were emitted due to break wear and tire wears (PAE VEHPBL). In Vienna, the majority of collected water flows entered the mixed sewer network and the waste water treatment plant (WWTP) downstream. Those flows become part of grill material, sewage sludge or effluent. Combined sewer overflow enters the receiving water without treatment. The separated sewer network takes up parts of the surface runoff and transports it to the receiving water. We established a separate SFA model ( Figure S3) in order to balance waterborne Cu flows.
 Measured flows were available for the WWTP (Kroiss et al. 2008). Therefore, 16.2 t Cu/yr enter the plant of which 13.6 t Cu/yr (SSL WWSCTS) are transferred to the sewage sludge, 1.6 t Cu/yr to the effluent (EFF WWSHYD) and 1.0 t Cu/yr to the grill material.  Diffuse emissions enter urban surfaces via surface runoff. They calculation of particulate emissions from the transport sector (low duty cars, trams and trains, catenary), building sector (Cu roofs) and atmospheric deposition is described at sec. 2.2.2.1.5. The surface runoff from the transportation grid covers brake & tire wear and catenary with an amount of 3.3 t Cu/yr (SRO TECWWS). The surface runoff from PHH carries atmospheric deposition and roof runoff with a flux rate of 4.0 t Cu/yr (SRO PHHWWS). The surface runoff from ECO covers 2.2 t Cu/yr (SRO ECOWWS). The ratio of separated and mixed sewer network length of 19:81 (Lehmann 2011) is used as proxy to allocate the surface runoffs to the two sewer types. 1.8 t Cu/yr enter the separate sewer system which transports the Cu to the receiving water. 8.9 t Cu/yr enter the mixed sewer system which of 50% enter receiving water as spillover (Fenz 1999) with a flux rate of 3.8 t Cu/yr (SPW WWSHYD).  Two point sources are relevant. (1) Waste water in PHH covers the final use of tap water as well as Cu from anthropogenic activities. (1.1) Viennese tap water covers the geogenic Cu content and Cu corrosion from pipes. Data on local water consumption statistics excluding the losses (Daxbeck et al. 1996;Tomenendal 2011) Lampert et al. 1997). Other emission sources like residues from food preparation, washing dishes, toilet papers, washing clothes and cleaning activities count for 843 mg Cu/cap/yr (Baccini et al. 1993;I C Consultants Ltd 2001). The multiplication with Viennese population size results 2.7 t Cu/yr. In total, the flux rate is 5.0 t Cu/yr (WWS PHHWWS).
(2) Waste water data from business and industries were restricted for access. As a consequence, we calculated the flow by stressing the mass balance principle with a flux rate of 7.4 t Cu/yr (WWS ECOWWS). In Taipei, the percentage of houses connected to public sanitary sewers is 100%. There are three wastewater treatment plants in operation for Taipei: Dihua, Neihu, and Bali. Local statistics showed that Cu in untreated and treated wastewater from Neihu WWTP were not detectable (< 0.02 mg/L). To estimate the flow, we set the detection limit of 0.02 mg Cu/l as maximum concentration. The estimation turned out to be 9.8 t/yr (WWA PHHWWS) and 2.8t/yr (TWA WWSHYD). By the ratio of different land uses, Cu in surface runoff was 0.02 t/yr (SRO ECOWWS), 0.5 t Cu/yr (SRO PHHWWS) and 0.4 t Cu/yr (SRO TECWWS). Cu in sewage sludge was then estimated to be 60 t Cu/yr (SSL WWSCTS). To estimate the flow, we used the sewage sludge flow with 65,700 tons (sum of Bali, Dihua, and Neihu sewage farms) and Cu concentrations with 887 ppm from Bali Sewage Farm (Shih 2007 (Wittmer 2006). Those from business were not exclusively reported for Vienna. Therefore, the national amount was downscaled by the ratio of WEEE from private households to total WEEE flow in Vienna. Summarizing, about 1,700 t Cu/yr were generated by households (HHW PHHCTS) and 5,400 t Cu/yr by industry and business (WAS ECOCTS).
In Taipei, household waste includes the trash collected by cleaning squad and by people themselves, as well as the recyclables. Result shows that 1,700 t/yr of Cu contained in household waste (HHW PHHCTS). In Taiwan, industrial waste comprises from industries and service sectors. Since there was nearly no factory in Taipei, industrial waste was presumed to come from service sectors, whose composition is similar to household waste. Because the process ECO was defined as the industrial sector, the amount of industrial waste transported by industrial or waste management institutes to incinerators and landfill sites were considered. Cu concentration in industrial waste was assumed to be as the same as bottom ash and fly ash from incinerators with a rate of 60 t Cu/yr (WAS ECOCTS).

Exported products (PRO ECOEXA)
Exported products stand for valuable goods which cross the city boundary.
In Vienna, exported products cover three sub-flows. First, the exported vehicles are traded as goods and were estimated to be 140 t Cu/yr (VEH VEHECO). Second, the number of unregistered EOL vehicles was assumed to be exported with 500 t Cu/yr (VEH UID). Third, fabricated goods from the production industry lack of sound data after the import/export statistics are based on the headquarter approach instead of a territorial allocation. So, we estimated 2,000 t Cu/yr with the help of import/export statistics and subtracted residues from production. In total, 2,600 t Cu/yr leave Vienna in a physical product form (PRO ANTEXA).
In Taipei, the same estimation assumption was made as in Vienna, and products exported from Taipei City were estimated to be 8,000 t Cu/yr.
2.2.2.1.9 Exported waste including recyclables (WAS CTSEXA) Exported waste flows are directed to recycling or disposal facilities out of the city. They cover valuable Cu which is recovered as well unrecovered Cu in residues from incineration.
In Vienna, we combined the generated waste flows, individual information regarding the whereabouts in recycling facilities and specific Cu contents. The flow is summarized by individual flows like recyclables within construction waste, waste from industry and business, and recycled EOL vehicles to an extend of 11,100 t Cu/yr (WAS CTSEXA).
Taipei City incinerated household waste from Keelung City too and stored the bottom ash temporarily during the period of constructing new landfill in Keelung. Likewise, Keelung City would provide landfill service of waste from Taipei City and temporary depository fly ash monolith. The temporary repository of bottom ash and fly ash will be sent back to the original city after the new landfill is constructed. Statistics of Taipei City showed that in 2003 total weight of bottom ash conveyed to Keelung City was 96,000 tons. We also assumed that ELV, scrap, and recyclables were included in exported waste, result in total 2,700 t Cu/yr (WAS CTSEXA).
2.2.2.1.10 Incineration: Mixed waste, exported residues for underground storage, residues, disposable waste (WAS CTSINC, TTR INCEXA, RES INCLDF) Waste incinerators are used to treat household and toxic industrial waste by temperatures up to 1,200 °C. While microorganisms like bacteria, fungi and virus as well as organic compounds are mineralized, heavy metals like Cu are transferred to bottom ash and APC residues. The whereabouts of residues depends on national polices and varies from bottom ash as construction material to underground disposal for filter APC residues.

S-12
In Vienna, four plants secure the treatment of household waste, residues from mechanical sorting, bulky waste fractions, street litter and some minor waste fractions with an amount of 600,000 t/yr. To estimate the plant load, we used Cu concentrations for six different waste flows that stem from plant specific monitoring reports and waste flow measurements (Boller 2002;Skutan and Brunner 2006;Skutan and Rechberger 2007;Umweltbundesamt 2000;Arx 2006;Hausmann 2005). Therefore, about 1,140 t Cu/yr enter the incinerators (WAS CTSINC). Public secondary waste flow data (Wiener Umweltschutzabteilung MA22 2011) and measured Cu concentrations in a representative Austrian waste to energy plants (Skutan and Rechberger 2007;Taverna et al. 2011) result 990 t Cu/yr in bottom ash (RES INCLDF) and 16 t Cu/yr in APC residues (TTR INCEXA). The bottom ash is used as aggregate to construct concrete blocks for landfill stabilization that ends up at a Viennese landfill. APC residues are heavily contaminated with salts and heavy metals. In 2008, they were exported to Heilbronn, an underground storage facility in Germany.
In Taipei, three incinerators operate for Taipei: Beitou, Neihu, and Muzha. According to local statistics, waste transported into incinerator was 640,000 ton in 2009. Cu content was estimated from the Cu concentration in fly ash and bottom ash. Cu in municipal solid waste (MSW) was 200 t/yr (WAS CTSINC). Residues from incinerators consist of fly ash and bottom ash, which contained 200 t/yr of Cu (RES INCLDF). Due to the policy "Zero Waste" of the local government, the waste amount in Taipei has decreased, and no statistics showed that residues from incinerators were exported from the city. Hence, Cu in exported residue for underground storage is zero (TTR INCEXA) in Taipei's case.
2.2.2.1.11 Composting: Compostable waste, residues, compost as fertilizer (WAS CTSCOM, RES COMCTS, COM COMUPV) Composting degrades biogenic waste in terms of volume and mass. Quality proven compost can be used as nutrient supplier for plants and soil improver.
In Vienna, one compost plant has been operated. The inputs stem from separated biomass collection and cover materials such as greencut and uncooked vegetables from households. To estimate the Cu flows throughout the plant, we used monitored bulk flows, and derived average dry matter contents as well as Cu concentrations from the literature. The inflow covers compostable raw materials and residues from anaerobic digestion in a biogas plant (Magistratsabteilung 48 2009). Data lacks prevented full Cu balance on plant level. So we set the output as net addition to plant. We assumed a dry matter ratio of 30% and a Cu concentration of 83 mg/kg dry matter and results 1.4 t Cu/yr (WAS CTSCOM). About 33.000 t valuable compost is used for land applications in Vienna (Republik Österreich 2001b; Weinmar 2011). Legal Cu concentrations for high quality compost A+ was set with 70 mg/kg dry substance (Republik Österreich 2001a) and resulted 1.4 t Cu/yr (COM CTSCOM = COM COMUPV).
In Taipei, incinerator plants in Taipei have auxiliary spaces to process compost, which has nothing to do with incinerators themselves. Uncooked food waste scraps are collected to incinerator plants and through the process of dehydration, fermentation, etc., they become compost as product. People or companies can take the compost if they need. Since there are few livestock industries or agricultural activities in Taipei, the compost here was assumed to be the treated uncooked food waste without residues. Due to limited data, the official statistics of uncooked food waste scraps in Taipei City and compost from Muzha Refuse Incineration Plant were adopted to estimate total compost in Taipei, which included 0.04 t Cu/yr (WAS CTSCOM, COM COMUPV).

Pesticides (PES ECOUPV)
The agricultural sector applies Cu as purposeful fungicides and bactericide on fruits and vegetables. The utilization of Cu dates back to beginning of the 19 th century and was known as "Bordeaux-Brühe" It is a mixture of chalk and aquatic Cu-sulphate-dissolution and bans the evolution of mildew. In 2009, the European Commission decided to constrain Cu form agricultural appliances after 2016 (Berger et al. 2011). Member states are enforced to establish national risk assessments in order to regulate Cu use.

S-13
In Vienna, the flow was calculated based on data on Austrian pesticide and plant protection product use (Berger et al. 2011) for wine, fruits and vegetables & crops and Viennese agricultural land use areas. In total, about 2.4 t Cu/yr are released by pesticides (PES ECOUPV).
In Taipei, there are barely agricultural activities; the agricultural area is 5.31 km 2 , which is around 2% of total area of the city. Hence, estimation of Cu in pesticides was not included for Taipei's case.
2.2.2.1.13 Fertilizer (FER ECOUPV) Manure, mineral fertilizer and harvest residues are common fertilizer inputs into agricultural soils. Especially manure is known to have a significant Cu concentration. Concentrations vary from 40-300 mg Cu/kg manure depending on the type of animal and foodstuff (Zethner et al. 2007). As an example, Cu is given to piglets in high doses to protect them from various diseases in the first weeks. Mineral fertilizer and harvest residues cover minor Cu as trace element.
In Vienna, the agricultural sector plays a minor role in terms of land use and production rates. About 15% of urban area is used to produce working animals, crops, wine and vegetables for local supply primarily. The Cu release was estimated by multiplying normalized Austrian and German Cu inputs (Fricke and Höhl 2000;Umweltbundesamt 2001;Zethner et al. 2007;Berger et al. 2011) with agricultural area classified into viticulture production, fruit production, and agriculture & horticulture (Fitzthum 2009). In total about 2.2 t Cu/yr are released by fertilizer applications (FER ECOUPV).
In Taipei, there are barely agricultural activities; the agricultural area is 5.31 km 2 , which is around 2% of total area of the city. Hence, estimation of Cu in fertilizer was not included for Taipei's case.
In Vienna, three types of networks are taken into account. First, the electricity grid covers ~3.700 km of overhead lines and ~ 22.300 km of underground cables (Wien Energie 2010). The stock was estimated based on voltage classes, cable lengths, cross sections and specific masses according to the Swiss network (Wittmer 2006). In total, about 89,200 t Cu provide the public energy supply. Second, Cu in the telecommunication network is estimated by a proxy based on the number of business units. Combining data from Australia and Sydney (van Beers and Graedel 2007) with Viennese business statistics resulted about 13,300 t Cu. Third, catenaries supply the 2 nd largest tram network worldwide as wells as national and local railway tracks with electricity. The Cu stock of 1,000 t is estimated based on the track length, cross sections, Cu content, and lifetime data. Summarizing, the infrastructure covered about 103,000 t Cu. In reality, the stock is larger than estimated due to Cu cables in subway lines and power transfer stations which were not taken into account.
In Taipei, the electricity transmission system and telecommunication system were taken into account. The electricity grid included 2,100 km overhead power lines and 310 km underground power lines (Lin 2003). For the telecommunication network, estimation was made according to case of Australia and statistics of household and business number in Taipei which gave a result of 24,000 t Cu. Cu stock in infrastructure was aggregated as 32,000 t Cu. Since Cu stock in cables and wires of MRT (Mass Rapid Transit) was not included, the actual value stock maybe larger than the estimated one.

Economy and private households (ECOstock, PHHstock)
The process economy and private household covers stocks like roofs, water pipes, heating systems, telecommunication and electricity networks as well as consumer goods.

S-14
In Vienna, the building structure and housing technology is comparable to Swiss standards. Using per capita data from Switzerland (Wittmer 2006) result about 134,000 t Cu in buildings. The land use area is used as proxy to allocate 94,000 t Cu to private households and 40,000 to economy. Cu in consumer goods like washing machine, dryer, electronics equipment and minor items like keys and coins stands for 37,000 t Cu. They were allocated to private households.
In Taipei, we selected items based on products listed in recycling fund system to estimate Cu stock in consumer goods. Assumption was made that those Cu-contained products included in recycling fund system were those comparatively highly used in households in Taiwan. The lifespan of different products were presumed to be uniform distribution. These consumer goods were either electrical and electronic appliances (EEA) or computer appliances. Five items for Cu stock calculation were color TV, air conditioner, washing machine, laptop, and desktop computer. With numbers of hundred households in Taipei as 9,628.31 in 2009, Cu consumed annually in household appliances and computer appliances in Taipei City were 17,000 t Cu. In economy sectors (ECO), building materials were assumed to be the stocks in it while consumer products were all input to private households (PHH). Electricity and telecommunication wire and cables were taken into account for Cu stock of buildings. In this study, the average length of in-use line in a case study in Tainan (Ou et al. 2007) and the total floor area in Taipei were integrated to do estimation. The Tainan case was reasonable and suitable for estimating because its structure and space usage are common around Taiwan, and hence the case could represent the general housing in Taiwan as well as in Taipei. We used ratio of landuse to allocate Cu stocks of buildings to ECO and PHH. The result shows that 200 t Cu existed in the form of building materials in ECO.

Vehicles (VEHstock)
The vehicle pool covers private and commercial wheelers as well as rolling stocks.
In Vienna, the calculation is based on the number of vehicles (Statistik Austria 2008; Wiener Linien 2009) and corresponding Cu contents from literature European Copper Institute (ECI) 2011;Hoock 2008;Struckl 2007). The private and public owned fleet consist registered motorized vehicles like cars, lorries and busses as well as rolling stock such as subways and trams. In total, the Cu stock in the transport sector covers 24.000 t Cu.
In Taipei, the estimation included the number of registered cars, buses and scooters in operation. 12,000 t Cu was estimated to be placed in vehicles.

Urban Soil (UPVstock)
Urban soil covers Cu, which compromises geogenic backgrounds and former anthropogenic inputs.
In Vienna, soil sampling data were reported by the City Authority. We printed a box plot including the detected Cu concentrations in parks and playgrounds. Therefore, about 46 mg Cu/kg soil were multiplied with the area of green space and a soil depth of 30 cm. This results about 4,000 t Cu in urban soil (UPVstock).
In Taipei, there is no local report of the soil sampling data, and hence estimation was made according to EPA public health declaration that Cu concentration in soil generally ranges from 2 to 250 ppm. With the area of green space of Taipei City and 30-cm depth of soil, Cu stock in urban pedosphere was obtained to be 830 t (UPVstock).
S-15                       . That amounts 4,982 t Cu/yr. We allocated the flows to sectors based on the consumption ratio of "pre-products" first, and to the buildings in ECO and PHH with the land use area.         The National Railway Company ÖBB operates 761 km electrified track with varying catenary diameters. The catenary material is made up from 99,5 -99,9% Cu and 0,1-0,5% Ag or Cd. The catenary is replaced after 40 years; the remaining cross section is about 80%. The urban Public Transport Provider Wiener Linien operates 363 km electrified tram tracks (main tracks + station tracks). The cross section is assumed with 120mm2. Lifetime data are equally to ÖBB.        (Baccini et al. 1993) Cu from washing dishes 25 (Baccini et al. 1993) Cu in toilet paper (from recycling paper) 64 (Baccini et al. 1993) Cu from product use

(I C Consultants Ltd 2001)
Cu from washing clothes 218 (Baccini et al. 1993) Cu from cleaning living areas 436 (Baccini et al. 1993) Total 843 Note: mg/cap/yr = milligram per capita per year Cu in urin 5 (Lampert et al. 1997) Cu in skin particles 47 (Lampert et al. 1997) Cu in human off-flows 760 Notes: mg/cap/yr = milligram per capita per year S-32      (Skutan and Rechberger 2007, p. 82;Skutan and Brunner 2006, p. 236) Average Organic Waste Collection (Biotonne) 76 (Umweltbundesamt 2000) Average Plant Waste Collection (Grünschnitt) 104 (Umweltbundesamt 2000) Cu concentration in street litter 46 (Boller 2002) Cu concentration in bulky waste 2.000 (Hausmann 2005) Cu concentration in paper and plastic 65 (Arx 2006) Note: mg/kg = milligram per kilogram 2.2.2.3.11.3 Waste from industry and business      To estimate the recycling flow, we used the recycling potential of six generated waste flows, combined them with individual information regarding the whereabouts in recycling facilities.  (Zethner et al. 2007;Kühnen and Goldbach 2004) . If the plants are further harvested, this amount of Cu is actively removed from the soil. With an agricultural area of ~ 6.000 ha, the Cu flow in harvest is estimated to be ~ 0,3 t/yr. To test this values probability, detailed harvest information (Fitzthum 2009) combined with Cu take-up values for Crops, Wine and Fruit production which was published in an Austrian study (Berger et al. 2011). The Sum of Cu take-up in harvest is therefore estimated to be 0,3 t/yr (uncertainty factor: 1,34).       (Ableidinger et al. 2007). Another 36 (excluding remediation sites) old unofficial dumps and landfills, filled with household or construction waste can be found in official datasets (Umweltbundesamt Österreich 2012). The total amount of Cu stocks in Viennese landfills is estimated to be ~ 144.000 t of Cu (uncertainty factor 1,84). This value fits well with the first estimation of 100.000 t Cu in landfills (without unofficial waste dumps) quite well.  Deposited wastes include excavated soil (usually from construction sites), construction waste and treated municipal waste according to Austrian landfill regulations and furthermore some municipal solid waste is stored temporarily. Using Cu concentrations from earlier calculations for excavated soil (28-61 mg/kg), construction waste (80-670 mg/kg) and municipal solid waste (1000-2600 mg/kg) the total Cu stock was calculated. Especially for the landfills "Langes Feld" and "Rautenweg" calculations are difficult. Due to their age, it is difficult to determine a Cu content of the former input. For example, "Langes Feld" was originally built to receive construction and demolition waste from World War II. Additionally to the type of landfill, slags which are used as building material in Viennese landfills and their input since 1964 (when the first MSW incineration plan in Vienna was established) were estimated. Here, the input of 2008, calculated with ~ 1.100 t/yr as well as capacities of incineration plants since 1964 were used as a base value. Additionally to official landfills, waste dumps or suspected waste dumps with a total Volume of ~ 30 Mio m³ were included into calculations (Umweltbundesamt Österreich 2012). With an average density of 1,0-1,2 t/m³ (Fellner 2012) and contents (partly suspected, partly tested) as a mixture of untreated municipal solid waste and construction waste (with unknown percentages), Cu stocks are estimated to be between 24.000 and 90.000 t.