Annex 19 Complementary information to source category 2d Copper Production

Overview of recent revisions

PCDD/PCDF emission factors are generally confirmed for source category 2d. No data could be found on class 6 (pure primary Cu smelters with no secondary feed materials). New data have been found on PCDD/PCDF releases through the water vector as well as on PCBs.

Table III.19.1 PCB emission factors for source category 2d Copper Production

2d Copper Production Emission Factors (µg TEQ/t copper)
Classification Air Water Land Product Residue
1 Sec. Cu – Basic Technology          
2 Sec. Cu – Well controlled 5       40
3 Sec. Cu –Optimized for PCDD/PCDF control 0.3        
4 Smelting and casting of Cu/Cu alloys          
5 Prim. Cu, well-controlled, with some secondary feed materials 0.01        
6 Pure primary Cu smelter with no secondary feed materials          

PCB emission factors are provided with:

  • A medium level of confidence for class 2 and 5 air emissions, as emission factors are not based on expert judgment but are not derived from a broad geographical coverage;
  • With a low level of confidence for class 3 air emissions and class 2 residue releases, as emission factors are based on extrapolations and expert judgment.

Derivation of emission factors

Release to Air

Emissions to air from copper production seem to vary considerably depending on the process technology, the nature of the materials processed and the gas cleaning system applied. The occurrence of PCDD/PCDF is principally associated with secondary copper production.

The following data are from secondary copper facilities. A study in the US on a copper production plant using a blast furnace and fitted with afterburners and fabric filters, gave an emission factor of 779 μg TEQ/t of scrap.

Studies in Germany on several plants gave emission concentrations, which varied over a large range from 0.032 to 30 ng TEQ/Nm³ (LUA 1997). Installations for smelting and casting of copper and its alloys, e.g. brass, gave emissions between 0.003 and 1.22 ng I-TEQ/Nm³ with a geometric mean of 0.11 ng TEQ/Nm³ (German data, LUA 1997). The compilation for European plants by the IPPC Bureau reported emissions of <0.1 ng I-TEQ/Nm³ (BREF 2009). From these data, an emission factor of 0.03 μg TEQ/t of copper/copper alloy was derived. The data do not allow for further differentiation according to technology or performance.

In the cleaned gases from sulfuric acid plants, emissions between 0.01 and 0.001 ng TEQ/Nm³ have been measured (BREF 2009). The same sources report – without further specification - that processes in the melt shop for the production of semis (semi-manufactures such as alloy cast ingots, foils, sheet, strip) gave emission factors for electric furnaces of <5 μg and for shaft and rotary furnaces of <10 μg TEQ/t, respectively.

Measured PCDD/PCDF results are available from Germany (Meyer-Wulf 1996) and Sweden (LUA 1997). However, it should be noted that these plants as well as those in Canada are not “pure” primary copper smelters since they process significant amounts of recyclable materials (Copper Smelters 2004). Measured data from Germany from such a “primary” copper smelter, that uses considerable amounts of secondary materials as feed (up to 40%) in flash smelting furnaces and matte converters gave emissions between 0.0001 and 0.007 ng TEQ/Nm³ resulting in a very narrow range of emission factors from 0.002 and 0.02 μg TEQ/t of copper (LUA 1997). Meyer-Wulf (1996) reported raw gas concentrations after the primary smelter between 0.004 ng I-TEQ/Nm³ and 0.3 ng I-TEQ/Nm³ whereas the higher concentrations were obtained when PVC was present in the recycled materials. Purified gases after the H2SO4 plant were either non-quantifiable or 0.001 ng I-TEQ/Nm³. The EU Dioxin Inventory report of 1997 (LUA 1997) reports concentrations of 0.005-0.015 ng I-TEQ/m³ in the waste gases from the roasting furnace for ore desulphurization. The volume of the waste gas was 5,000 Nm³ per ton of copper produced. In addition, from a Swedish primary smelter that recycles considerable amounts of secondary materials, which produced 2,000 Nm³/t of waste gases, a concentration of 11 ng I-TEQ/m³ was reported. From the results of the measurements given above, emission factors between 0.25 μg I-TEQ/t (from German results) and 22 μg I-TEQ/t (from Swedish results) were derived. The Belgium inventory took an emission factor of 10 μg I-TEQ/t to estimate its national releases (LUA 1997). The data in the upper range reflect more classes 2 and 3.

Globally speaking, the assessment made for the 2005 version of the Toolkit was confirmed by the literature review conducted between 2007 and 2012:

  • Class 1: Concentrations of 63, 199 and 246 ng TEQ/Nm³ were measured at the outlet of a furnace (before a bag filter) in Chinese plants (Hung et al. 2009). These concentrations corresponds to emission factors of 328, 1037 and 1282 µg TEQ/t respectively, based on gas flows used in another Chinese publication (Ba et al. 2009). In the draft revised BREF on the non-ferrous metal industry (BREF 2009), the highest European concentrations reported in installations where no APC was used reached 29.5 ng TEQ/Nm³. Assuming a gas flow of 10,000 Nm³/t, such a concentration corresponds to an emission factor of 295 µg TEQ/t.
  • Class 2: Data obtained in China (Ba et al. 2009) and Taiwan (Yu et al. 2006) are very similar. The former publication provides an emission factor of 14.8 µg TEQ/t and the latter provides an emission factor of 24.5 µg TEQ/t.
  • Class 3: Data collected on a Chinese plant using a bag filter and activated carbon injection show that PCDD/PCDF concentrations can reach 0.1-0.7 ng TEQ.Nm³, corresponding to an emission factor range of 0.5-3.65 µg TEQ/t (Hung et al. 2009). In Korea, concentrations of 0.63 ng TEQ/Nm³ were measured (Kim et al. 2005), resulting in an emission factor of 6.3 µg TEQ/t, based on the assumption that the gas flow is 10,000 Nm³/t.
  • Class 5: Three publications directly provide emission factors (Iwata et al. 2008, Grochowalski et al. 2007, Yu et al. 2006): 0.43, 0.04 and 0.014 µg TEQ/t respectively. These data were collected from plants located in Japan, Poland and Korea.

So far, there only are few data on releases of PCDD/PCDF from class 6 copper plants. The majority of information is from secondary copper plants, where occasionally high PCDD/PCDF emissions were found in the stack gases. When compiling this Toolkit, no measured data of PCDD/PCDF emissions or releases from pure primary copper smelters have been submitted nor found elsewhere. In some countries, like Chile, among others, primary copper smelters use only ores and concentrates and do not mix with secondary materials. In other countries, like Germany, Sweden, and Canada, among others, primary copper smelters receive feeds that include scrap and other recycled materials that are introduced in these “primary” copper smelters at rates between 15% and 40% (COCHILCO 2004). For the pure primary copper smelters as present, among others, in the Chilean copper foundries, the probability to form PCDD/PCDF in the production of primary copper seems to be very low or not existing. These primary foundries use clean raw materials and use either the base smelting process (with furnaces like the Teniente or the Noranda) or the flash smelting (with Outokumpu furnace). The white copper or concentrates from the furnaces are converted into copper blister in an oxygen-rich atmosphere by utilizing the Peirce-Smith Converter. Typical temperatures in the smelting processes are well above the critical temperatures reported for PCDD/PCDF formation: in the Teniente, the gases are at 1,260°C in a sulphur dioxide-rich atmosphere (at 25%), the liquid white copper at 1,240°C, and the liquid slags the temperature is 1,240°C. In the Outokumpu flash furnace the temperature is around 1,260°C and the gases leave at 1,300°C-1,350°C. The Pierce-Smith Converter operates in a temperature range of 1,150°C-1,250°C. The refining of the copper blister – to remove sulphur and oxygen - takes place in rotary kilns at an operational temperature around 1,200°C. The slags still have quite high copper contents (4%-10%) and are treated in the Teniente furnace, electric arc furnaces, or slag flotation plants at temperatures above 1,200°C. Purification of gases originating from the smelting furnaces and the converters is done by rapid quench, followed by electrostatic precipitators and washing towers and wet scrubbers. The sulphuric acid plants (H2SO4 plants) apply catalytic converters (COCHILCO 2004).

New data have been identified with respect to PCB air emissions. PCBs were measured at the stack of a primary copper plant corresponding to class 5 (Yu et al. 2006). On this Korean plant, a concentration of 0.08 ng TEQ/Nm³ was measured, from which an emission factor of 0.012 µg TEQ/t was derived by the authors. PCBs were also measured on three primary smelters in Poland, where concentrations were in the range of 0.0004-0.0035 ng/Nm³. Based on PCDD/PCDF data, these concentrations would correspond to a PCB emission factor of 0.001 µg TEQ/t. Therefore, an emission factor of 0.01 µg TEQ/t is proposed for class 5. Regarding class 3, concentrations measured in Belgium and in Korea are similar, as those are ranging from 0.026 to 0.046 ng TEQ/Nm³ (Kim et al. 2005, François et al. 2005). Assuming a gas flow of 10,000 Nm³/t, these concentrations correspond to an emission factor of 0.3 µg TEQ/t on average. In two references, emission factors to be assigned to class 2 are calculated: 0.098 and 9.8 ng TEQ/t respectively (Ba 2009, Yu et al. 2006). Therefore a PCB emission factor of 5 µg TEQ/t is proposed for class 2.

New data have also been identified regarding HCB air emissions from primary copper production (Iwata et al.2008). This Japanese publication proposed an emission factor of 11,000 µg/t. This figure is assigned to class 5.

Release to Water

These may occur if effluents are discharged and the concentration is likely to be influenced by any water treatment applied. Any liquid release should be noted along with its source and treatment applied.

One set of data is available from a Swedish plant which is one of the largest copper smelter of its kind worldwide (Jansson et al. 2009). This plant processes ores and secondary raw material such as electronic scrap. Three different effluents were sampled and analyzed, with two replicate samples collected for each effluent. The effluent consisting of purified process water and water used in the production of sulfur dioxide, mixed with cooling water showed concentrations of 3.7 to 9.1 ng TEQ/Nm³. Taking into account the effluent flow on this site, this range corresponds to an emission factor range of 0.2-0.5 µg TEQ/t. The two other flows consist of cooling water, where emission factors are lower (1 to 20 ng TEQ/t). As this Swedish site processes a large range of feed materials, a common emission factor of 0.5 µg TEQ/t is proposed for all classes under category 2d.

Release to Land

No release to land is expected.

Release in Products

No releases to with the products are expected.

Release in Residues

PCDD/PCDF will be found in the solid residues from the process. The principal concern is the residues from the gas treatment equipment. Dusts and sludge collected from gas treatment may be highly enriched in PCDD/PCDF. Concentrations of up to 20,000 ng TEQ/kg have been reported (SCEP 1994).

UK data (Dyke et al. 1997) suggests approximately 2,000 t of filter dusts arise from production of 46,000 t of copper. Combined with an average concentration of 14,400 ng TEQ/kg in the dust (SCEP 1994) this resulted in the 2005 Toolkit emission factor of 630 μg TEQ/t of product. This estimate, which was originally considered highly uncertain, has been confirmed by subsequent references from China and Korea (Ba et al. 2009, Yu et al. 2006).

For high technology plants a lower emission factor of 300 μg TEQ/t was originally proposed in the 2005 Toolkit. This rough estimate has been confirmed by subsequent references. For instance, an emission factor of 116 µg TEQ/t was assessed from Korean data (Jin et al. 2009).

For class 2, a PCB emission factor of 40 µg TEQ/t is proposed from two references where 4.17 µg TEQ/kg of residues and 0.13 µg TEQ/kg of residues are reported (Ba et al. 2009, Yu et al. 2006).

It should be noted that solid residues from the copper smelters may be recycled internally or be transferred to other secondary metal reclamation plants. In such cases, the solid residues constitute an intermediate and its PCDD/PCDF release will not be taken into account in the national PCDD/PCDF release inventory.