International Aluminium Institute

Results of the 2013 Anode Effect Survey Report on the Aluminium Industry’s Global Perfluorocarbon Gases Emissions Reduction Programme

International Aluminium Institute | www.world-aluminium.org

Contents Global Aluminium Industry PFC Emissions Reduction Performance 2013 ............................. 4 Industry Trends ..................................................................................................................... 5 2013 Anode Effect Survey ..................................................................................................... 6 Adoption of revised GWPs and its consequence ................................................................... 9 Global Emissions Estimations ..............................................................................................13 Uncertainties ........................................................................................................................16 Benchmark Data...................................................................................................................17 Appendix A – Facility Emissions Calculation Methodologies.................................................23

Tables Table 1 – Aluminium smelting technology categories ............................................................ 6 Table 2 - 2013 Anode Effect Survey participation by technology ........................................... 7 Table 3 - IPCC GWP Values ................................................................................................. 9 Table 4 – Perfluorocarbon emission results from facility data reporting to the 2013 Anode Effect Survey ..................................................................................................................................11 Table 5 – Production weighted mean PFC emissions per unit production of reporting entities, 2006-2013 ............................................................................................................................12 Table 6 – Total global 2013 PFC emissions .........................................................................14 Table 7 - Slope and overvoltage coefficients by technology, including uncertainty (Source: IPCC, 2006) .........................................................................................................................24

International Aluminium Institute | www.world-aluminium.org

Figures Figure 1 –Location of primary aluminium production, 1990 & 2006-2013 (SOURCE: IAI & CRU) .............................................................................................................................................. 5 Figure 2 – Primary aluminium smelting technology mix, 1990-2013 (SOURCE: IAI & CRU).. 5 Figure 3 – Reporting production & rate 1990-2013 ................................................................ 8 Figure 4 – Reporting rates (aluminium production) per technology, 2006-2013 ..................... 8 Figure 5 – Median PFC emission rates (as CO2e) of reporting entities, per technology, 20062013 .....................................................................................................................................10 Figure 6 – PFC emissions (as CO2e) per tonne of aluminium production, 2006-2013 ..........15 Figure 7 – Absolute PFC emissions (as CO2e) and primary aluminium production, 1990-2013 .............................................................................................................................................15 Figure 8 – PFC emissions (as CO2e per tonne Al) performance of reporters, benchmarked as cumulative fraction within technologies, 2013 .......................................................................18 Figure 9 –PFC emissions performance of reporters (t CO2e/t Al), benchmarked as cumulative production within technologies, 2013 ....................................................................................18 Figure 10 - PFC emissions performance of reporters (t CO2e/t Al), benchmarked as cumulative production within technologies, 1990 & 2013........................................................................19 Figure 11 - Average anode effect frequency of reporters benchmarked by technology type, 2013 .....................................................................................................................................20 Figure 12 - Average anode effect duration of reporters benchmarked by technology type, 2013 .............................................................................................................................................20 Figure 13 - Average anode effect minutes per cell day of reporters benchmarked by technology type, 2013 ............................................................................................................................21 Figure 14 - Average anode effect overvoltage of reporters benchmarked by technology type, 2013 .....................................................................................................................................22

International Aluminium Institute | www.world-aluminium.org

4

Global Aluminium Industry PFC Emissions Reduction Performance 2013 Global aluminium industry perfluorocarbon (PFC) emissions intensity (as CO2e per tonne of production) has been reduced by more than 35% since 2006, almost 90% since 1990. With primary aluminium production having grown by over 150% over the same period, absolute emissions of PFCs by the aluminium industry have been reduced from approximate 100 million tonnes of CO2e in 1990 to 32 million tonnes in 2013. The International Aluminium Institute (IAI) has collected anode effect data directly from primary aluminium producers for the purposes of calculating sectoral PFC emission inventories for over a decade, with annual surveys carried out since 2000. The 2013 Anode Effect Survey generated data from 218 reporting entities (smelters & potlines) representing 20 million tonnes of primary aluminium production, with emissions from the remaining 30 million tonnes of global primary aluminium production (the majority in China), modelled using historic, sampled or technology average data. This survey report outlines year 2013 data collection and analysis methodologies and global results. Historically IAI has used global warming potential (GWP) values for perfluorocarbon gases as published in the IPCC Second Assessment Report (1996), to align with a 1998 decision of the Conference of the Parties to the Kyoto Protocol (Decision 2/CP.3). The 2011 decision by the Conference of the Parties serving as the meeting of the Parties to the Kyoto Protocol (Decision 4/CMP.7) to utilise revised GWPs from the IPCC Fourth Assessment Report (2007), for the second commitment period (from 2013) has necessitated recalculation of aluminium industry PFC emissions (as CO2e). This report outlines this process and, where appropriate, delivers data using both sets of GWP. However, for all future reports (and for online historic datasets and baselines), only 2006 GWPs will be employed. Current and historic PFC emissions data (utilising revised GWPs) can also be found on the International Aluminium Institute’s website http://www.worldaluminium.org/statistics/perflurocarbon-pfc-emissions/#data. As in this report, separate company or country PFC emissions data is not published, but rather is aggregated by production technology

International Aluminium Institute | www.world-aluminium.org

5

Industry Trends Growth in primary aluminium production continues to be driven by China and the GCC countries and, since the late 1990s, based on latest point fed prebake technology. Global primary aluminium production in 2013 was a record 51 million tonnes.

Primary Aluminium Production (million tonnes)

55 50 45

China

40

GCC

35

Other Asia

30

Africa

25

Oceania South America

20

CIS 15

Europe

10

North America

5 0 1990

2006 2007 2008 2009 2010 2011 2012 2013

Figure 1 –Location of primary aluminium production, 1990 & 2006-2013 (SOURCE: IAI & CRU)

Annual Primary Aluminium Production (million tonnes)

60 HSS 50

VSS NB:technology category details in Table 1

SWPB 40

PFPB CWPB

30

20

10

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

1995

1990

0

Figure 2 – Primary aluminium smelting technology mix, 1990-2013 (SOURCE: IAI & CRU)

International Aluminium Institute | www.world-aluminium.org

6

2013 Anode Effect Survey Perfluorocarbons, or PFCs, are a group of potent greenhouse gases with long atmospheric lifetimes (in the tens of thousands of years), of which the greatest volume is emitted from industrial processes. PFCs can be produced in the primary aluminium electrochemical smelting process, during events referred to as anode effects. An anode effect is a process upset condition, where an insufficient amount of alumina (Al2O3), the raw material for primary aluminium production, is dissolved in the electrolyte bath, contained in the electrolytic cells (or pots) within a smelter reduction line (potline). This causes the voltage in the pot to be elevated above the normal operating range, resulting in the emission of gases containing the PFCs tetrafluoromethane (CF4) and hexafluoroethane (C2F6). Data on anode effects generated by process monitoring systems allows one to calculate such emissions. The International Aluminium Institute has collected anode effect data directly from primary aluminium producers for the purposes of calculating sectoral PFC emission inventories for over a decade, with annual surveys carried out since 2000. The IAI Anode Effect Survey requests data from all aluminium smelting facilities around the world, via IAI member companies (http://www.world-aluminium.org/about/members/), direct correspondence with non-member producers and regional industry associations. Facilities are requested, where possible, to report by potline, and to separate data from different technologies within a single plant. As well as anode effect process data, reporters are also asked for information that allows for quality control (by comparison against other facilities and within reporters’ data over time) and for the IAI to take a snapshot and monitor over time the adoption of anode effect mitigation technologies such as prediction and automatic termination software. The reporting form and guidelines (PFC001) can be found from the IAI website (http://www.world-aluminium.org/media/filer_public/2013/01/15/pfc001.pdf). BROAD TECHNOLOGY CATEGORY

TECHNOLOGY CATEGORY Centre Worked

Prebake (anodes pre-baked)

Side Worked Vertical Stud

Søderberg (anodes baked in-situ)

Horizontal Stud

ALUMINA FEED CONFIGURATION

ACRONYM

Bar broken centre feed

CWPB

Point centre feed

PFPB

Manual side feed

SWPB



Manual side feed



Point feed Manual side feed

VSS HSS

Table 1 – Aluminium smelting technology categories

International Aluminium Institute | www.world-aluminium.org

7

Participation Rate It is significant that the 2013 survey results include data from 100% of SWPB, VSS and HSS technology production. On average, these technologies produce more emissions per tonne of aluminium produced than the CWPB and PFPB categories (see Error! Reference source not found.). As the aluminium production in China represents an increasing proportion of the industry and non-reported data are predominantly from China, the overall reporting rate shown in Figure 5 continues to decrease (40% in 2013). Outside China, 24 smelters, representing over 5 million tonnes of production (equivalent to around 10% of worldwide production), do not report data to the IAI.

TECHNOLOGY

2013 primary aluminium production (1,000 tonnes)

2013 production represented in survey (1,000 tonnes)

2013 participation rate by production

CWPB

1,412

613

43%

PFPB (Rest of World)

19,988

15,257

76 % 34%

PFPB (China)

24,936

0

SWPB

514

514

100 %

VSS

3,378

3,378

100 %

HSS

373

373

100 %

All Technologies (excluding China)

25,666

20,135

78 %

All Technologies (Including China)

50,602

20,135

40 %

0%

Table 2 - 2013 Anode Effect Survey participation by technology Note: any inconsistencies due to rounding

The high coverage of the survey data outside China (with respect to both metal production and emissions) and of the higher emitting technologies, combined with the fact that actual measurements and secondary information, means that the IAI is able to develop estimates of PFC emissions from the global aluminium industry, with some degree of accuracy.

International Aluminium Institute | www.world-aluminium.org

50

100%

45

90%

40

80%

35

70%

30

60%

25

50%

20

40%

15

30%

10

20%

5

10%

0

0%

Reporting Production

Non Reporting Rest of World

Non Reporting China

Reporting Rate (RHS)

Reporting rate

Annual Primary Aluminium Production (Million tonnes)

8

Figure 3 – Reporting production & rate 1990-2013

100% 80% PFPB CWPB HSS VSS SWPB

60% 40% 20% 0% 2006

2007

2008

2009

2010

2011

2012

2013

Figure 4 – Reporting rates (aluminium production) per technology, 2006-2013

Data Requested Annual (1 January – 31 December 2013) data required include: 

Annual primary aluminium metal production (MP), the mass of molten metal (in metric tonnes) tapped from pots in reporting period;



Anode effect frequency (AEF), the average number of anode effects occurring per cell day over the reporting period;



Anode effect duration (AED), the average time (in minutes) of each anode effect over the reporting period;



Anode Effect Overvoltage (AEO), the average cell voltage (in millivolts) above the target operating voltage, when on anode effect, over the reporting period.

International Aluminium Institute | www.world-aluminium.org

9 Overvoltage is specifically requested from operators employing Rio Tinto Alcan AP-18 or AP3x PFPB technologies and SWPB facilities using control technology that records overvoltage rather than anode effect duration. These anode effect performance data allow for the calculation, by the Intergovernmental Panel on Climate Change (IPCC) Tier 2 or Tier 3 methodology 1 , of facilities’ total annual tetrafluoromethane (CF4) and hexafluoroethane (C2F6) emissions, and hence tonnes of CO2 equivalent (CO2e) emitted per tonne of aluminium produced. F

F

It should be noted that the IPCC Tier 1 methodology of multiplying metal production by a technology-specific coefficient to estimate PFC emissions is not good practice, as the results are not derived from process data and consequently have a very high uncertainty attached to them. IAI does not use the Tier 1 methodology in any of its PFC emissions calculations.

Adoption of revised GWPs and its consequence The 2011 adoption, by the Conference of the Parties to the Kyoto Protocol, of revised global warming potentials (GWPs) for greenhouse gas emissions calculations in the protocol’s second commitment period (2013-2020) has necessitated recalculation of the industry’s PFC data (current and historic). Historically, IAI reports have used IPCC 2nd Assessment Report (1996) GWPs to calculate carbon dioxide equivalency for CF4 and C2F6 emissions, in alignment with the first commitment period recommended values. Recommended GWPs for the second commitment period are now drawn from the IPCC 4th Assessment Report 2007).

GWPs

IPCC 2nd Assessment Report

IPCC 4th Assessment Report

CF4

6,500

7,390

C2F6

9,200

12,200

Table 3 - IPCC GWP Values

As a result, IAI has recalculated and republished its historic PFC emissions data, based on the revised 2007 GWPs (http://www.world-aluminium.org/statistics/perflurocarbon-pfcemissions/#data) and from 2013 onwards will report only using these values.

1

2006 IPCC Guidelines for National Greenhouse Gas Inventories, Primary Aluminium Production, Chapter 3,Section 4.4, http://www.ipccnggip.iges.or.jp/public/2006gl/pdf/3_Volume3/V3_4_Ch4_Metal_Industry.pdf.

International Aluminium Institute | www.world-aluminium.org

10

2013 Survey Results Anode effect data was collected from 218 reporting entities (smelters & potlines) representing 20 million tonnes of primary aluminium production Results are summarised in Error! Reference source not found. below. Facilities that have made PFC measurements by which Tier 3 calculation of PFC emissions is possible account for 46% of the total reported CF4 emissions from survey participants. It should be noted that Tier 3 calculations typically carry an uncertainty of +/- 15%, with well controlled systems down to +/- 12%, while uncertainty in Tier 2 calculations can be as high as +/- 50%." The range of anode effect and PFC emissions performance within technologies is explored further in the “Benchmark Data” section below. Changes in median emission performance (in t CO2e/t Al) within technologies between 2006 and 2013 are shown in Figure 5. SWPB

VSS

CWPB

PFPB

HSS

2013

2013

2012

2012

2011

2011

2010

2010

2009

2009

2008

2008

2007

2007

2006

2006 0

1

2

3

4

5

6

7

8

9 10

0

1

SWPB

VSS

CWPB

PFPB

2

4

5

6

7

8

9 10

t CO2e/t Al

t CO2e/t Al IPCC 4th GWP

3

HSS

IPCC 2nd GWP

Figure 5 – Median PFC emission rates (as CO2e) of reporting entities, per technology, 2006-2013

Reported average (production weighted mean) PFC emissions (as CO2e) per tonne of production have been reduced by 36% between 2006 and 2012 (CF4 by 41%, C2F6 by 43%).

International Aluminium Institute | www.world-aluminium.org

11 IPCC 4th GWP Median

Technology

CWPB

PFPB

SWPB VSS HSS ALL

IPCC Tier

Median

Mean No. of Reported Total CF4 Total C2F6 CF4 C2F6 C2F6: CF4 Total PFC reporting production emissions emissions intensity intensity weight emissions2 entities (kt Al) (Gg CF4) (Gg C2F6) (kg CF4/ (kg C2F6/ ratio (kt CO2e) t Al) t Al)

2

1

324

0.005

0.001

3

1

288

0.008

0.001

2 Slope 3

64

5,753

0.151

0.018

27

5,413

0.134

0.017

Slope 2 OV

18

1,462

0.091

0.011

3 OV

6

1,761

0.082

0.007

2

6

111

0.007

0.002

3

2

403

0.107

0.032

2

67

2,620

0.390

0.021

3

9

759

0.121

0.007

2

13

162

0.022

0.002

3

4

211

0.084

0.009

-

218

20,135

1.269

0.140

IPCC 2nd GWP

Median PFC

Mean PFC

Total PFC

Median PFC

Mean PFC

intensity

intensity

emissions

intensity

intensityr

(t CO2e/ t Al)

(t CO2e/ t Al)

(kt CO2e)

(t CO2e/ t Al)

(t CO2e/ t Al)

0.022

0.003

0.15

123

0.20

0.20

105

0.17

0.17

0.023

0.003

0.12

4,040

0.20

0.27

3471

0.17

0.23

0.264

0.089

0.25

1,875

2.97

3.65

1,581

2.48

3.08

0.146

0.008

0.06

4,116

1.17

1.22

3,577

1.02

1.06

0.150

0.013

0.10

918

1.26

2.46

791

1.09

2.12

-

-

0.11

11,072

-

0.55

9,525

Table 4 – Perfluorocarbon emission results from facility data reporting to the 2013 Anode Effect Survey

Note: any inconsistencies due to rounding

2

Carbon dioxide equivalent (CO2e) emissions for survey participants are calculated by multiplying the total tonnes of each PFC component gas by the Global W arming Potential (GWP) values reported in the IPCC 4th Assessment Report (i.e. 7,390 for CF4 and 12,200 for C2F6). Calculations using GWP contained in the IPCC 2nd Assessment Report (i.e. 6,500 for CF4 and 9,200 for C2F6) are also listed here for reference.

International Aluminium Institute | www.world-aluminium.org

0.47

12

Reporting production (kt Al)

Reporting rate by production

CF4 emission factor (kg CF4/t Al)

C2F6 emission factor (kg C2F6/t Al)

Total PFC emission factor (t CO2e/t Al)

Total PFC emission factor (t CO2e/t Al)

IPCC 4th GWP

IPCC 2nd GWP

2013

20,135

40%

0.063

0.007

0.55

0.47

2012

21,006

44%

0.069

0.008

0.61

0.52

2011

22,413

51 %

0.079

0.009

0.68

0.60

2010

21,774

53 %

0.071

0.009

0.63

0.54

2009

22,184

60 %

0.069

0.008

0.61

0.52

2008

24,741

63 %

0.089

0.010

0.78

0.67

2007

23,903

63 %

0.106

0.013

0.95

0.81

2006

23,177

68 %

0.116

0.014

1.03

0.87

Table 5 – Production weighted mean PFC emissions per unit production of reporting entities, 2006-2013

International Aluminium Institute | www.world-aluminium.org

13

Global Emissions Estimations Methodology A more realistic picture of the global aluminium industry’s PFC emissions inventory should include some estimate of the non-reporting industry year on year. In fact, the IAI voluntary objective is an objective for the industry as a whole, not just IAI membership or reporting companies and so is based on such a global estimate. The IAI uses median PFC emissions performance per technology (as shown in Error! Reference source not found. above) applied to non-reporting production by technology in order to calculate the global PFC emissions inventory from aluminium production. Non-reporting aluminium production tonnage data is taken from three sources. The majority (China 2013 primary aluminium production of 24,936,070 metric tonnes) is reported by the China Nonferrous Metals Industry Association (CNIA). Around 3.5 million tonnes of production (n=13) is from other IAI surveys – primarily IAI Form 100 “Primary Aluminium Production” (http://www.world-aluminium.org/media/filer_public/2013/01/15/iai_form_100.pdf). Finally, just under 1.2 million metric tonnes of production is data kindly provided by the CRU Group (www.crugroup.com), for facilities where there is no direct IAI data collection (n=7).

Accounting for China Recent (2008-2013) PFC emissions measurements at 27 PFPB facilities in China have yielded a median emission factor of 0.80 tonnes CO2e per tonne of aluminium produced (CF4 median 0.100 kg/t Al; C2F6:CF4 weight fraction 0.046), compared with a PFPB survey reporter median performance of 0.20 tonnes CO2e per tonne of aluminium (0.023 kg CF4/t Al; C2F6:CF4 weight ratio = 0.12). This China-specific value (0.80 t CO2e/t Al) is applied to the 2013 Chinese non-reporting PFPB cohort, in place of the IAI PFPB survey median, and has also been applied to historical Chinese non-reporting production, to derive a time series that more accurately reflects Chinese smelter performance and global emissions than one based on rest-of-world averages, albeit one that remains static over time.

International Aluminium Institute | www.world-aluminium.org

14

2013 Global Aluminium Industry PFC Emissions Summing the emissions and production data from reporting and non-reporting facilities and then dividing total global PFC emissions (t CO2e) by total global production (t Al), gives a production weighted average 2013 PFC emissions performance for the global aluminium industry of 0.63 tonnes of CO2e per tonne of primary aluminium produced, as outlined in Table 6 – Total global 2013 PFC emissions

Total PFC emissions (1,000 t CO2e)

Total aluminium production (1,000 tonnes)

PFC emission factor (t CO2e/t Al)

PFC emission factor (t CO2e/t Al)

IPCC 4th GWP

IPCC 2nd GWP

Reported

11,072

20,135

0.55

0.47

Calculated from nonreporters

20,928

30,467

0.69

0.61

TOTAL GLOBAL

32,000

50,602

0.63

0.56

Table 6 – Total global 2013 PFC emissions Note: any inconsistencies due to rounding

Global PFC emissions (as CO2e) per tonne of production have been reduced by 33% since 2006, on course to meet the IAI voluntary objective of a 50% reduction by 2020 on a 2006 baseline. The 33% improvement since 2006 takes the overall improvement since 1990 to 88%. With PFC emissions per tonne cut by almost 90% since 1990 and primary aluminium production having grown by 159% over the same period, absolute emissions of PFCs by the aluminium industry have been reduced from approximate 100 million tonnes of CO 2e in 1990 to 32 million tonnes in 2013, a fall of over 68%.

International Aluminium Institute | www.world-aluminium.org

15

PFC Emissions (t CO2e/t Al)

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3

IPCC 4th GWP

IPCC 2nd GWP

Figure 6 – PFC emissions (as CO2e) per tonne of aluminium production, 2006-2013

120 Annual Primary Aluminium Production (Mt Al) 100

Total Annual PFC Emissions (Mt CO2e)

80

60

40

20

0

Figure 7 – Absolute PFC emissions (as CO2e) and primary aluminium production, 1990-2013

International Aluminium Institute | www.world-aluminium.org

16

Uncertainties Understanding sources and magnitude of uncertainty in the calculation of global industry PFC emissions is important, not only in terms of the current emissions inventory and its relationship to top-down measurements of PFCs in the atmosphere, but also with respect to quantifying the industry’s performance over time. Given that the 2013 data presented above indicates a significant reduction in total PFC emissions (as CO2e) since 1990, it is necessary to consider the uncertainties inherent in the 1990 baseline number and the 2013 performance number and to quantify the probability that the reduction has been made. Potential significant sources of uncertainty include: 

the application of average industry IPCC Tier 2 calculation factors,



use of Tier 2 factors for calculating PFC emissions for survey participants where suitable facility specific measurements are not available, and,



estimates of PFC emissions for producers that do not participate in the anode effect survey.

Uncertainty arises from the use of IPCC Tier 2 average industry factors due to the uncertainty in the mean slope and overvoltage coefficients. Additional PFC measurements will reduce the uncertainty of the mean coefficient values. However, for all technology groups there is considerable variance in the individual values of slope and overvoltage coefficients, from which the means are calculated. For this reason, calculations of PFC emissions with Tier 2 coefficients will be more uncertain than calculations made with Tier 3 coefficients, calculated from PFC measurements made using good measurement practices. Calculations of PFC emissions for non-reporters is even more uncertain where, due to lack of availability of anode effect performance, the median emission factors of reporters per technology is applied to nonreporters.

International Aluminium Institute | www.world-aluminium.org

17

Benchmark Data The IAI Anode Effect Survey provides respondents with valuable benchmark information, allowing producers to judge their performance relative to others operating with similar technology. The benchmark data are presented in this section in the form of cumulative probability graphs and calculated PFC emissions benchmark data as both cumulative probability and cumulative production graphs. The cumulative probability graphs show, on the horizontal axis, the benchmark parameter: 

PFC emissions per tonne of aluminium;



Anode effect frequency (AEF);



Anode effect duration (AED);



Anode effect minutes per cell day (AEM) and



Anode effect overvoltage (AEO).

The vertical axes show the cumulative fraction of reporting facilities that perform at or below the level chosen on the vertical axis. For facilities reporting data from multiple potlines, a data point is shown for each potline. To illustrate how the graph in Figure 8 is interpreted consider, for example, the 0.5 point on the vertical axis, at which the HSS data point is 1.26 t CO2e/t Al. The interpretation is that 50% of all potlines/facilities reporting HSS anode effect data operate at or below 1.26 t CO2e/t Al. At 1.0 on the vertical axis the HSS point is 4.64 t CO2e/t Al. The interpretation is that all HSS facilities reported anode effect data that reflected PFC emissions performance at or below 4.64 t CO2e/t Al or, in other words, the maximum value calculated for HSS operators in 2013 was 4.64 t CO2e/t Al.

PFC Emissions per Tonne of Aluminium The lowest PFC emissions per tonne of aluminium produced are produced by PFPB facilities, although with a wide range of performance. The VSS and HSS facilities show a similar distribution, but with higher average emissions factor. The highest PFC emissions per tonne of aluminium produced and the widest range in performance result from SWPB cells.

International Aluminium Institute | www.world-aluminium.org

Cumulative Fraction of Reporting Entities

18

1.0 0.9 Note: SWPB 86th and 100th percentile outliers are at 8.1 tCO2e/t Al and 12.6 t CO2e/t Al respectively.

0.8 0.7 0.6 0.5 0.4

0.3 0.2 0.1 0.0 0.0

1.0

2.0

3.0

4.0

PFC Emission Factor (t CO2e/t Al)

CWPB & PFPB

SWPB

HSS

5.0

6.0

VSS

Figure 8 – PFC emissions (as CO2e per tonne Al) performance of reporters, benchmarked as cumulative fraction within technologies, 2013

14

PFC Emissions (t CO2-eq/tonne Al)

12

PFPB&CWPB SWPB

10

HSS 8 VSS 6 4 2 0 0

5

10

15

20

Cumulative Aluminium Production of Reporting Facilities (Million tonnes) Figure 9 –PFC emissions performance of reporters (t CO2e/t Al), benchmarked as cumulative production within technologies, 2013

International Aluminium Institute | www.world-aluminium.org

19 Taking the 1990 reporting cohort and plotting it against 2013 data shows improvement both from existing facilities over this time but also, importantly, the positive contribution of new (predominantly PFPB) capacity added since 1990. 45

PFC Emissions (t CO2-eq/tonne Al)

40

PFPB&CWPB

35

SWPB

30

HSS 25

VSS

20 15 10

1990

5

2013

0 0

5

10

15

20

Cumulative Aluminium Production of Reporting Facilities (Million tonnes)

Figure 10 - PFC emissions performance of reporters (t CO2e/t Al), benchmarked as cumulative production within technologies, 1990 & 2013

International Aluminium Institute | www.world-aluminium.org

20 Anode Effect Frequency & Duration

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Anode Effect Frequency (number of AE per cell day) CWPB & PFPB

SWPB

VSS

HSS

Figure 11 - Average anode effect frequency of reporters benchmarked by technology type, 2013

1.0 Cumulative Fraction of Reporting Entities

Cumulative Fraction of Reporting Entities

The following graphs shows the distribution of anode effect frequency and duration data for reporting facilities in 2013. As can be expected from the greater degree of control capability of PFPB cells, this technology has the lowest AEF distribution of the five groups.

0.9 0.8 0.7

0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

Anode Effect Duration (minutes) CWPB & PFPB

SWPB

VSS

HSS

Figure 12 - Average anode effect duration of reporters benchmarked by technology type, 2013

International Aluminium Institute | www.world-aluminium.org

21 Anode Effect Minutes per Cell Day Anode Effect Minutes per Cell Day (AEM) are the product of anode effect frequency and duration and, for facilities employing the Slope Method. AEM relate directly to PFC emissions per tonne of aluminium produced through a slope factor that is either technology specific (IPCC Tier 2 methodology) or facility specific (Tier 3 methodology). 𝐸𝐹𝐶𝐹4 = 𝑆𝐶𝐹4 × 𝐴𝐸𝑀 Both PFPB and CWPB technologies have the same Tier 2 value for slope: 0.143 kg CF4/t Al

Cumulative Fraction of Reporting Entities

per AEM. However, the IPCC Tier 2 slope parameter for SWPB, VSS and HSS technologies are considerably different. The slope value is highest for the SWPB technology group, 0.272 kg CF4/t Al per AEM. The comparable slope values for VSS and HSS are 0.092 and 0.099, respectively. 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3

0.2 0.1 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 Anode Effect Minutes per Cell Day (minutes) CWPB & PFPB

SWPB

HSS

VSS

Figure 13 - Average anode effect minutes per cell day of reporters benchmarked by technology type, 2013

Anode Effect Overvoltage Figure 14 shows the benchmarking graph for anode effect overvoltage for PFPB cells operating with Rio Tinto Alcan AP technologies and which calculate PFC emissions from overvoltage process data. For these operators, the AEO parameter relates directly to anode effect related PFC emissions per tonne of aluminium produced.

International Aluminium Institute | www.world-aluminium.org

22

Cumulative Fraction of Reporting Entities

1.0 0.9 0.8 0.7 0.6 Positive

0.5

Algebraic

0.4 0.3 0.2 0.1 0.0 0

2

4 6 8 10 Anode Effect Overvoltage (mV)

12

14

Figure 14 - Average anode effect overvoltage of reporters benchmarked by technology type, 2013

Positive overvoltage reporting now predominates over algebraic overvoltage reporting. The positive overvoltage should give a better correlation with PFC emissions per tonne of aluminium than algebraic overvoltage since algebraic overvoltage recording can result in subtractions of voltage during the anode effect treatment period that do not relate to PFC emissions.

International Aluminium Institute | www.world-aluminium.org

23

Appendix A – Facility Emissions Calculation Methodologies Slope Method The basic equations for calculation of PFC emission rates from facilities reporting anode effect frequency and duration are: 𝐸𝐶𝐹4 = 𝑆𝐶𝐹4 × (𝐴𝐸𝐹 × 𝐴𝐸𝐷) × 𝑀𝑃 and 𝐸𝐶2 𝐹6 = 𝐸𝐶𝐹4 × 𝐹𝐶2 𝐹6 /𝐶𝐹4 where 𝐸𝐶𝐹4 = kilograms of 𝐶𝐹4 emitted 𝐸𝐶2 𝐹6 = kilograms of 𝐶2 𝐹6 emitted 𝑆𝐶𝐹4 = slope coefficient for 𝐶𝐹4 𝐹𝐶2 𝐹6 /𝐶𝐹4 = weight fraction of 𝐶2 𝐹6 to 𝐶𝐹4 While AEF and AED are reported data, the slope coefficient for CF4 can be either “facility specific” (IPCC Tier 3 methodology), or “technology specific” (IPCC Tier 2 methodology). The first of these options, Tier 3, is the more certain method for calculating emissions and involves use of a slope coefficient (and weight fraction) derived from direct measurement of PFC emissions at the facility. The Tier 2 method involves the use of slope coefficients that are an average of measurement data available in 2005 taken from facilities around the world within technology classes.

International Aluminium Institute | www.world-aluminium.org

24

Table 7 - Slope and overvoltage coefficients by technology, including uncertainty (Source: IPCC, 2006)

Participants in the Anode Effect Survey are asked to report if a facility-specific direct measurement of PFC emissions had been made and if a Tier 3 slope coefficient and weight fraction are available for calculating PFC emissions from the smelter. The remainder of the PFC emissions data are calculated using IPCC Tier 2 methodology with industry average coefficients.

Overvoltage Method For smelters that report overvoltage data, the following equations are employed: 𝐸𝐶𝐹4 = 𝑂𝑉𝐶 ×

𝐴𝐸𝑂 × 𝑀𝑃 𝐶𝐸⁄ 100

and 𝐸𝐶2 𝐹6 = 𝐸𝐶𝐹4 × 𝐹𝐶2 𝐹6 /𝐶𝐹4 where 𝐸𝐶𝐹4 = kilograms of 𝐶𝐹4 𝑒𝑚𝑖𝑡𝑡𝑒𝑑 𝐸𝐶2 𝐹6 = kilograms of 𝐶2 𝐹6 𝑒𝑚𝑖𝑡𝑡𝑒𝑑 𝑂𝑉𝐶 = overvoltage coefficient for 𝐶𝐹4 𝐶𝐸 = current efficiency, expressed as % 𝐹𝐶2 𝐹6 /𝐶𝐹4 = weight fraction of 𝐶2 𝐹6 to 𝐶𝐹4

International Aluminium Institute | www.world-aluminium.org

25 Again, a Tier 3 methodology applies a facility specific overvoltage coefficient and weight fraction, derived from on site PFC measurements and anode effect data and reported as part of the Survey return. Tier 2 calculations apply technology specific, average coefficients, which are outlined in the 2006 IPCC Guidelines for National Greenhouse Gas Inventories.

Global Warming Potentials Carbon dioxide equivalent (CO2e) emissions for survey participants are calculated by multiplying the total tonnes of each PFC component gas by the Global Warming Potential (GWP) values reported in the IPCC Fourth Assessment Report 3 F

:

𝐸𝐶𝑂2 𝑒 = (𝐸𝐶𝐹4 × 7390) + (𝐸𝐶2 𝐹6 × 12200) For benchmarking purposes (that is to say, comparing emissions performance between facilities of the same technology but with different levels of production), total (or “absolute”) CO2e emissions are divided by relevant aluminium production, to give an emission factor in tonnes of CO2e per tonne of aluminium produced: 𝐸𝐹𝐶𝑂2 𝑒 =

𝐸𝐶𝑂2 𝑒 𝑀𝑃

3

The latest data published by IPCC in the Fourth Assessment Report reports the CF 4 GWP as 7,390 and the C2F6 GWP as 12,200.

International Aluminium Institute | www.world-aluminium.org

Published by: INTERNATIONAL ALUMINIUM INSTITUTE 10 Charles II Street London SW1Y 4AA United Kingdom Tel: + 44 (0) 20 7930 0528 Fax: + 44 (0) 20 7321 0183 Email: [email protected]

U

© International Aluminium Institute A company limited by guarantee. Registered in London no. 1052007