View all publications

Appendix B: Inventory Data

B.1 Lamp Composition

Table B1a provides a breakdown of the compositions used for each lamp type assessed in the study.  These are based on data from the Stewardship Solution (2007) report for generic materials (glass, electronics, plastics, metals and other) and from material safety data sheets (MSDS) for hazardous materials (Philips 2001, 2002 and 2005 and Sylvania 2009).  Compositions are for those lamps which represent the mainstream types used in New Zealand for residential, commercial and industrial applications. 

Table B1a Lamp Composition Data


Component / Material

Unit

1. Linear fluorescent lamp (LFL)
35W, T8

2. Compact fluorescent lamp (CFL)
11W: external ballast

3. Compact fluorescent lamp (CFL)
20W: integral ballast

4. High pressure sodium (HPS)
150W

5. Metal Halide (MH)
400W

6. Mercury Vapour (MV)
250W

Total Mass g 120.0 55.0 120.0 150.0 240.0 166.5
Glass g 115.0 40.0 65.0 105.0 195.0 129.5
Electronics g - - 25.0 - - 1.1
Plastics g - 10.0 25.0 - - -
Metals Total g 3.0 3.0 4.0 44.5 42.0 8.5
*Other Total g 2.0 2.0 1.0 - 3.0 27.3
Nuisance Dust g 3.0 - - - - -
Cerium Terbium Magnesium Aluminate mg 300 - - - - -
Barium Magnesium Aluminate mg 600 - - - - -
Yttrium Oxide
(1314-36-9)
mg 600 - - - - -
Antimony
(7440-36-0)
mg 12 - - - - -
Manganese
(7439-96-5)
mg 24 - - - - -
Mercury
(7439-97-6)
mg 4 5 5 50 50 50
Iodine
(7553-56-2)
mg - - - - 48 -
Sodium
(7440-23-5)
mg - - - 15 48 -
Sodium Iodide
(7681-82-5)
mg - - - - 48 -
Yttrium Vanadate
(7440-65-5)
mg - - - - 1.4 4.2
Vanadium
(1314-62-1)
mg - - - - - 1.7
Yttrium
(7440-65-6)
mg - - - - - 1.2
Lead
(7439-92-1)
mg - - - - - 8.3
Phosphor powder mg - 1.1 2.4 - - -

Source: Philips (2001, 2002 and 2005), Stewardship Solutions (2008), Sylvania (2009).
*Note: ‘Materials Other’ includes all those materials listed below that line which are shown in the table.

B.2    Raw Materials Production

Table B2a, B2b and B2c show the inventory datasets and assumptions used to represent raw material production.  Datasets are primarily sourced from Ecoinvent 2.0 (2007), which represents European datasets.  In some cases, where representative datasets were not available, estimates have been made using surrogate datasets.  For some materials, the quantities of individual components have been estimated stoichiometrically based on the chemical formula.  All datasets have been adjusted to account for the electricity generation mix associated with their country of manufacture.  Additionally, for the main materials (glass, electronics, plastics and metals) the transport has been adjusted to reflect geographic conditions.
Table B2a         Raw Materials Inventory Data

Material

Assumptions

Inventory Dataset

Glass Glass used in lamp construction is assumed to be flat coated glass, represented by an Ecoinvent (2007) dataset. Flat glass, coated, at plant/RER SNI
Electronics See Table B2b for breakdown of electronics materials and inventory datasets used.  
Plastics All plastics used in lamp construction are assumed to be polyurethane, represented by an Ecoinvent (2007) dataset. Polyurethane, rigid foam, at plant/RER SNI
Metals See table B2c for metal composition of lamp end caps and inventory datasets used.  
Others All materials not identified in lamp construction are assumed to be HDPE, represented by an Ecoinvent (2007) dataset. Polyethylene, HDPE, granulate, at plant/RER SNI
Cerium Terbium Magnesium Aluminate Cerium Terbium Magnesium Aluminate, (Ce,Tb)MgAl11O19:Ce,Tb, has been estimated stoichiometrically on the following basis:
  • 3% Mg
  • 32% Al
  • 15% Ce
  • 33% O
  • 17% Tb
A data set for phosphate rock is assumed to represent cerium production, and uranium is assumed to represent terbium production.  Magnesium, aluminium and oxygen have used representative datasets from Ecoinvent 2.0 (2007). 
  • Magnesium, at plant/RER SNI
  • Aluminium  - Aluminium, primary, at plant/RER SNI
  • Cerium - Phosphate rock, as P2O5, beneficiated, dry, at plant/MA SNI
  • Oxygen  - Oxygen, liquid, at plant/RER SNI
  • Terbium - Uranium natural, in uranium hexafluoride, at conversion plant/CN SNI
Barium Magnesium Aluminate Barium Magnesium Aluminate, Al2Ba2Mg2O7, has been estimated stoichiometrically on the following basis:
56% Ba
10% Mg
11% Al
23% O
  • A dataset for magnesium has been assumed to represent barium production.  Magnesium, aluminium and oxygen have used representative datasets from Ecoinvent 2.0 (2007). 
Barium - Magnesium, at plant/RER SNI
Magnesium - Magnesium, at plant/RER SNI
Aluminium  - Aluminium, primary, at plant/RER SNI
Oxygen  - Oxygen, liquid, at plant/RER SNI
Cerium Magnesium Aluminate
(67542-72-7)
Cerium Magnesium Aluminate, (Ce)MgAl11O19:Ce), has been estimated stoichiometrically on the following basis:
  • 3% Mg
  • 39% Al
  • 18% Ce
  • 40% O
dataset for phosphate rock is assumed to represent cerium production.  Magnesium, aluminium and oxygen are represented by Ecoinvent (2007) datasets.
  • Magnesium - Magnesium, at plant/RER SNI
  • Aluminium  - Aluminium, primary, at plant/RER SNI
  • Cerium - Phosphate rock, as P2O5, beneficiated, dry, at plant/MA SNI
  • Oxygen  - Oxygen, liquid, at plant/RER SNI
Yttrium Oxide
(1314-36-9)
Yttrium oxide (Y2O3), has been estimated stoichiometrically on the following basis:
  • 79% Y
  • 21%  O
A dataset for phosphate rock is assumed to represent yttrium production.  Oxygen is represented by an Ecoinvent (2007) dataset.  Material weight represents up to 0.5% of total lamp.
  • Yttrium - Phosphate rock, as P2O5, beneficiated, dry, at plant/MA SNI
  • Oxygen  - Oxygen, liquid, at plant/RER SNI
Antimony
(7440-36-0)
No dataset available for antimony.  Material weight represents up to 0.01% of total lamp.  This quantity is considered to be negligible therefore raw material production data has been omitted. No dataset available. 
Manganese
(7439-96-5)
Manganese is represented by an Ecoinvent (2007) dataset. Manganese, at regional storage/RER SNI
Mercury
(7439-97-6)
Mercury is represented by the following datasets:
  • Primary mined uses an Ecoinvent (2007) dataset;
  • Recovered from vinyl chloride monomer facilities uses an Ecoinvent (2007) dataset;
  • Recovered from chlor-alkali facilities uses an Ecoinvent (2007) dataset;
  • Recycled via vacuum distillation process uses a datasets for battery recycling from ERM (2006); and
  • Recycled via pyrometallurgical process uses a datasets for battery recycling from ERM (2006).
Mercury, liquid, at plant/GLO SNI
  •  
Iodine
(7553-56-2)
A dataset for sodium sulphate is assumed to represent iodide production.  Material weight represents up to 0.02% of total lamp. Sodium sulphate, from natural sources, at plant/RER SNI
Sodium
(7440-23-5)
Sodium is assumed to be 100% sodium chloride represented by an Ecoinvent (2007) dataset. Sodium chloride, powder, at plant/RER SNI
Sodium Iodide
(7681-82-5)
Sodium iodide (NaI) has been estimated stoichiometrically on the following basis:
  • 15.3% Na
  • 84.7% I
A dataset for sodium sulphate is assumed to represent iodide production.    A dataset for sodium chloride has been assumed to represent sodium production, represented by an Ecoinvent (2007) dataset. 
  • Sodium - Sodium chloride, powder, at plant/RER SNI
  • Iodide – Sodium sulphate, from natural sources, at plant/RER SNI
Yttrium Vanadate
(7440-65-5)
Yttrium Vanadate (YVO4) has been estimated stoichiometrically on the following basis:
  • 44% Y
  • 25% V
  • 31% O
A dataset for phosphate rock is assumed to represent yttrium production.  Vanadium is represented by a dataset from IDEMAT (2001) and oxygen using a dataset from Ecoinvent 2.0 (2007). 
  • Yttrium - Phosphate rock, as P2O5, beneficiated, dry, at plant/MA SNI
  • Vanadium - Vanadium I (IDEMAT, 2001)
  • Oxygen - Oxygen, liquid, at plant/RER SNI
Vanadium
(1314-62-1)
Vanadium used in lamp construction is represented by a dataset from IDEMAT (2001).  Vanadium I (IDEMAT, 2001)
Yttrium
(7440-65-6)
A dataset for phosphate rock is assumed to represent yttrium production.  Material weight represents up to 0.7% of total lamp. Yttrium - Phosphate rock, as P2O5, beneficiated, dry, at plant/MA SNI
Lead
(7439-92-1)
Lead used in lamp construction is assumed to be primary lead, represented by an Ecoinvent (2007) dataset. Lead, primary, at plant/GLO SNI
Phosphor Powder Phosphor used in lamp construction is assumed to be white phosphor liquid, represented by an Ecoinvent (2007) dataset. Phosphorus, white, liquid, at plant/RER SNI
Hydrogen Hydrogen used in lamp construction is assumed to be liquid hydrogen, represented by an Ecoinvent (2007) dataset. Hydrogen, liquid, at plant/RER SNI

Table B2b Electronics Inventory Data

Material

Assumptions

Estimated percentage of composition (%)

Inventory Dataset

Inductor Inductors used in lamp construction assumed to be unspecified inductors, represented by an Ecoinvent (2007) dataset. 35 Inductor, unspecified, at plant/GLO SNI
PCB PCBs used in lamp construction assumed to be passive electronic component, represented by an Ecoinvent (2007) dataset. 12 Electronic component, passive, unspecified, at plant/GLO S
Electrolytic capacitor Electrolytic capacitors used in lamp construction are assumed to be electrolyte type capacitors, represented by an Ecoinvent (2007) dataset. 16 Capacitor, electrolyte type, < 2cm height, at plant/GLO SNI
Filter inductor Filter inductors used in lamp construction are assumed to be unspecified inductors, represented by an Ecoinvent (2007) dataset. 7 Inductor, unspecified, at plant/GLO SNI
Capacitor Capacitors used in lamp construction are assumed to be unspecified capacitors, represented by an Ecoinvent (2007) dataset. 12 Capacitor, unspecified, at plant/GLO SNI
Coil Coil used in lamp construction are assumed to be low-alloyed steel, represented by an Ecoinvent (2007) dataset. 5 Steel, low-alloyed, at plant/RER S
Transistor Transistors used in lamp construction are assumed to be unspecified transistors, represented by an Ecoinvent (2007) dataset. 5 Transistor, unspecified, at plant/GLO SNI
Diode Diodes used in lamp construction are assumed to be unspecified diodes, represented by an Ecoinvent (2007) dataset. 3 Diode, unspecified, at plant/GLO SNI
R0 (resistor) R0 resistors used in lamp construction are assumed to be unspecified resistors, represented by an Ecoinvent (2007) dataset. 1 Resistor, unspecified at plant/GLO SNI
Mainscord Mainscords used in lamp construction is assumed to be a ribbon cable, 20-pin, with plugs, represented by an Ecoinvent (2007) dataset. 3 Cable, ribbon cable, 20-pin, with plugs, at plant/GLO SNI
Resistor Resistors used in lamp construction are assumed to be unspecified resistors, represented by an Ecoinvent (2007) dataset. 2 Resistor, unspecified at plant/GLO SNI

Source: Philips (2009)

Table B2c  Metals Inventory Data

Material

Assumptions

Estimated percentage of composition (%)

Inventory Dataset

    1. LFL 2. CFLe 3. CFLi 4. HPS 5. MH 6. MV  
Nickel Nickel used in lamp composition assumed to be nickel 99.5%, represented by an Ecoinvent (2007) dataset.   94 94       Nickel, 99.5%, at plant/GLO SNI 
Aluminium Aluminium used in lamp composition assumed to be primary aluminium, represented by an Ecoinvent (2007) dataset.       94 94 94 Aluminium, primary, at plant/RER SNI
Brass Brass used in lamp composition assumed to be brass, represented by an Ecoinvent (2007) dataset. 94           Brass, at plant/CH SNI  
Solder Solder used in lamp composition assumed to be solder paste for electronics industry, represented by an Ecoinvent (2007) dataset. 6 6 6 6 6 6 Solder, paste, Sn63Pb37, for electronics industry, at plant/GLO SNI

Source: Sylvania (2009)

In cases where surrogate datasets were not available, the original ore containing the raw material in question has been identified.  An inventory dataset for another raw material mined from the same ore has been used, as the manufacturing process is assumed to be similar for both raw materials (Table B2d). 

Table B2d Ores Containing Raw Materials Used in Lamp Manufacture

Material

Assumptions

Inventory Dataset

Cerium Cerium is extracted from monazite ore – Ce (Ce, La, Pr, Nd, Th, Y)PO4.  This ore also contains the minerals Lanthanum , Praseodymium, Neodymium, thorium, yttrium and, phosphate.  A dataset for phosphate rock is assumed to represent cerium production.  This is  represented by an Ecoinvent (2007) dataset. Cerium - Phosphate rock, as P2O5, beneficiated, dry, at plant/MA SNI
Ytrrium Yttrium is extracted from monazite ore – Ce (Ce, La, Pr, Nd, Th, Y)PO4.  This ore also contains the minerals lanthanum , praseodymium, neodymium, thorium, cerium and, phosphate.  A dataset for phosphate rock is assumed to represent yttrium production.  This is represented by an Ecoinvent (2007) dataset. Yttrium - Phosphate rock, as P2O5, beneficiated, dry, at plant/MA SNI
Terbium Terbium is extracted from xenotime ore – YAsO4 or YPO4.  This ore also contains the minerals uranium, gadolinium, thorium, ytterbium, erbium, dysprosium, chernovite, and yttrium orthophosphate.  A dataset for natural uranium is assumed to represent terbium production.  This is represented by an Ecoinvent (2007) dataset.  Terbium - Uranium natural, in uranium hexafluoride, at conversion plant/CN SNI
Iodine Iodine is extracted from Caliche rock.  This rock also contains potassium nitrate, sodium borate, sodium nitrates, sodium sulphate.  A dataset for sodium sulphate is assumed to represent iodine production.  This is represented by an Ecoinvent (2007) dataset.  Iodine – Sodium sulphate, from natural sources, at plant/RER SNI

Source:  Wikipedia (2009), Mineral Information Institute (2009).

Manufacturing processes are described below for raw materials that did not have corresponding inventories.

Mercury

Based on latest available data representing 2005/06 (COWI, 2008, Concorde, 2008, Claushius, 2009) for mercury recycling, recovery and mining internationally, the following estimates have been calculated for mercury sources that are used in lamp production, as shown in Table B2e.  The countries listed below cover between 70% to 90% of sources of mercury for all lamp types, which is assumed to be sufficient coverage to represent an international average figure for this study.  The data shown in Table B2e have been combined with import data (as shown in Table B4a) to generate an average production mix per lamp type.

Table B2e  Sources of Mercury

Source 

China

Thailand

Europe

Primary mined 89% 0% 0%
Recovered from non-ferrous metal smelting and natural gas cleaning 0% 32% 0%
Recovered from vinyl chloride monomer facilities 11% 0% 0%
Recovered from chlor-alkali facilities 0% 62% 10%
Recycled via vacuum distillation process 0% 6% 62%
*Recycled via pyrometallurgical process 0% 0% 28%
Total 100% 100% 100%

Source: (COWI, 2008, Concorde, 2008, Claushius, 2009)

*Within the EU Claushuis Metaalmaatschappij B.V. (The Netherlands) is the main recycler using the pyrometallurgical process.  Other processes for hydrometallurgical and electrometallurgical are minimal in the EU and have been ignored. 

Inventory datasets to describe production mercury from these processes are listed in Table B2a above.  For allocation, of burdens to recovered mercury from chlor-alkali facilities, non-ferrous metal smelting and natural gas cleaning and vinyl chloride monomer, this is assumed to be burden free for mercury from these processes, with the exception of local transport to the point of manufacture which is estimated to be 500km by road.  Data for recycling has been used for batteries to represent the pyrometallurgical process (ERM, 2006) and primary data gathered in this study for mixed waste for Australia (EcoCycle, 2009) for the vacuum distillation process.  Input inventories for electricity have been adjusted accordingly to supply geography. 

Cerium Terbium Magnesium Aluminate

A method of manufacturing a terbium activated cerium magnesium aluminate phosphor has been described by Fan et. Al. (1989).  This is achieved by forming a uniform powder blend consisting essentially of aluminium oxide, cerium oxide, magnesium, fluoride, and terbium, then firing the blend in a non-oxidizing atmosphere at a temperature of approximately 1500 – 1800° C, for a sufficient time to produce the phosphor.  The phosphor is then cooled and deagglomerated.

Yttrium

Yttrium deposits can be found in xenotime, euxenite and fergusonite.  It can also be obtained as a by-product of some uranium ores processing. 

After Yttrium separation from the whole bulk of rare earth metals, it is subject to reduction. The metal is produced commercially by reduction, transferring yttrium oxide into yttrium halogenide.  During the process this compound is mixed with sublimated calcium and argon and heated in a furnace up to 1600°C.  The slag material is split off and hard oxygen and tantalum extracted, to produce yttrium ingots (Chemical-elements.info, 2007).

Iodine

Iodine is extracted from water and oil deposits and from mother waters of saltpeter production.  It may also be recovered from seaweeds ashes.  Mineralized water contains 0.001–0.01% iodine in iodides. During production, it is acidified and treated to extract elemental iodine, using activated coal or anionites.  Iodine is refined by sublimation or smelting (Chemical-elements.info, 2007).

Antimony

Antimony is sometimes found native, but more frequently it is found in the sulfide stibnite (Sb2S3) which is the predominant ore mineral (Wikipedia, 2009).

Antimony ores are concentrated by flotation and gravity methods. The element is extracted mostly by a pyrometallurgical method which is precipitation of the fusion with iron. Alternatively partially oxidized ores, or ores containing precious metals are subject of oxidizing roasting with sublimation.  It may be applied for Sb2O3, which is then processed by reduction smelting (Chemical-elements.info (2007).

Barium

The most common naturally occurring minerals containing barium are barium sulfate, BaSO4 (barite), and barium carbonate, BaCO3 (witherite). 

Barite concentrate is the main feedstock material for barium extraction. It is obtained by barite flotation using liquid glass as a dead rock depressor.  After BaSO4 reduction with lack coal, coke or natural gas, BaS is obtained which then is converted into other barium compounds. After roasting at 800, 1400 and 700°C BaO is produced which then is reduced with aluminium powder at 1100–1200°C (Chemical-elements.info, 2007).

Sodium

Naturally occurring sodium is bound to other elements in many minerals. The most common sodium-containing mineral is halite (or rock salt), chemically known as sodium chloride. Other minerals that contain sodium include cryolite (sodium aluminium fluoride), soda ash (sodium carbonate), and soda niter (or Chile saltpeter, sodium nitrate) (Wikipedia, 2009).  Sodium is now produced commercially through the electrolysis of fused (liquefied) sodium chloride. In this process, known as the Downs process, calcium chloride is mixed with the sodium chloride to lower the melting point below 600 °C. Sodium, but not calcium, is deposited on the cathode (New World Encyclopaedia, 2008).

The transport of raw materials from point of extraction to the locations of lamp manufacture has been estimated in tonne kilometres, based on shipping distances from the website ‘Sea Rates’ (2009) and locations for worldwide mineral production data from Index Mundi (2009). Tonne kilometres were estimated for the top three countries of lamp manufacture and the top three countries of raw material production.

B.3 Lamp Manufacture

The European Lamp Companies Federation (2008b) reports that during the production phase of an 11W CFL, energy use is 143 MJ.  Based on mass (kg), the electricity use for production of the five other lamp types was estimated as this information was not provided by lamp manufacturers, exact figures could not be obtained.  It is likely that this could be an overestimate for the higher mass lamps.  The study has not considered other material inputs from lamp production and has assumed a 1% material wastage in the production process. Table B3a shows the estimated electricity needed for production of each lamp type.

Table B3a Estimated Electricity Used to Manufacture Lamps

Lamp Type

Mass

Electricity Used

Electricity Used

Estimated Electricity

  g Mj  kWh kWh
1. Linear fluorescent lamp (LFL)
35W, T8
120     4.0
2. Compact fluorescent lamp (CFL)
11W: external ballast
55     1.8
3. Compact fluorescent lamp (CFL)
20W: integral ballast
120 143 4.0 -
4. High pressure sodium (HPS)
150W
150     5.0
5. Metal Halide (MH)
400W
240     7.9
6. Mercury Vapour (MV)
250W
166.5     5.5

Source: European Lamp Companies Federation (2008b)

B.4  Lamp Import into New Zealand

Table B4a shows a breakdown of the countries where lamps are manufactured, based on data provided by the Statistics New Zealand (2009) for the import of lamp products into New Zealand.  This data has been used to determine the electricity generation mix, by country of origin, for the production of lamps and transportation distances of lamps imported into New Zealand.

Table B4a  New Zealand Lamp Import Data

Lamp Type 1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
Country % % % % % %
Australia       1.4    
Belgium       15.4 21.4 3.4
Brazil       1.9    
China 60.3 60.3 60.3 51.1 38.4 60.3
Germany 9.1 9.1 9.1 3.9 12.9  
Hong Kong         1.0  
Hungary 2.4 2.4 2.4 10.5 8.3  
India         3.3  
Indonesia 1.4 1.4 1.4      
Italy 1.6 1.6 1.6      
Japan       2.8 3.1 28.1
Netherlands 1.1 1.1 1.1     3.6
Poland 1.1 1.1 1.1      
Slovakia       7.5    
Switzerland           1.1
Thailand 20.5 20.5 20.5      
United Kingdom       2.9    
United States of America       2.3 8.7  
Remainder 2.5 2.5 2.5 0.4 2.8 3.5
Total 100.0 100.0 100.0 100.0 100.0 100.0

Source: Statistics New Zealand (2008).

Table B4b shows the shipping distances from each country of lamp manufacture to Auckland, New Zealand.  These distances have been calculated using the website ‘Sea Rates’ (2009).  Road transport distances have been estimated from the place of manufacture to port in the country of origin as 400km and from Auckland port to the point of retail in New Zealand as 400km. 

Table B4b  Shipping Distances from Lamp Countries of Origin to New Zealand


Country

Shipping Distance

Shipping Distance

Estimated Trucking Distance –   Place of Manufacture to Port

Estimated Trucking  Distance –  Auckland Port to Retail Point in New Zealand

  Nautical miles km km km
Australia 1,638 3,034 400 400
Belgium 11,373 21,063 400 400
Brazil 6,780 12,557 400 400
China 5,197 9,625 400 400
Germany 11,627 21,533 400 400
Hong Kong 5,060 9,371 400 400
Hungary 10,658 19,739 400 400
India 7,169 13,277 400 400
Indonesia 4,549 8,425 400 400
Italy 10,353 19,174 400 400
Japan 4,883 9,043 400 400
Netherlands 11,379 21,074 400 400
Poland 11,968 22,165 400 400
Slovakia 10,658 19,739 400 400
Switzerland 10,785 19,974 400 400
Thailand 5,485 10,158 400 400
United Kingdom 11,327 20,978 400 400
United States of America 5,659 10,480 400 400

Source: Sea Rates (2009)

Table B4c shows the estimated tonne kilometres for each lamp type, based on the distances travelled above, and the proportion of lamps manufactured in each country.  Inventory data to represent shipping and trucking are based on Ecoinvent (2007) datasets. 

Table B4c  Estimated Tonne Kilometres for Importing and Distribution of Lamps to Point of Retail

 

1. Linear fluorescent lamp (LFL)
35W, T8

2. Compact fluorescent lamp (CFL)
11W: external ballast

3. Compact fluorescent lamp (CFL)
20W: integral ballast

4. High pressure sodium (HPS)
150W

5. Metal Halide (MH)
400W

6. Mercury Vapour (MV)
250W

  tkm tkm tkm tkm tkm tkm
Trucking Distance –   Place of Manufacture to Port 0.05 0.02 0.05 0.06 0.10 0.07
Shipping – Place of Manufacture to New Zealand 1.38 0.63 1.38 2.14 3.54 1.71
Trucking  Distance –  Auckland Port to Retail Point in New Zealand 0.05 0.02 0.05 0.06 0.10 0.07

Source: Statistics New Zealand (2008), Sea Rates (2009).

B.5  Retail and Use

Table B5a presents the electricity consumption for typical lamp usage in New Zealand, based on the lamps rated wattage and lifespan.  Typical lamp lifetimes have been sourced from Stewardship Solutions (2008). 
For the purposes of the study, the environmental impacts associated with retail are assumed to be negligible.

Table B5a  Typical Lamp Specifications

 

Unit

1. Linear fluorescent lamp (LFL)
35W, T8

2. Compact fluorescent lamp (CFL)
11W: external ballast

3. Compact fluorescent lamp (CFL)
20W: integral ballast

4. High pressure sodium (HPS)
150W

5. Metal Halide (MH)
400W

6. Mercury Vapour (MV)
250W

Lamp wattage W 35 11 20 150 400 250
Lifespan Hours 8,000 10,000 12,000 20,000 20,000 20,000
Energy use Mj 1,008 396 864 10,800 28,800 18,000
  kWh 280 110 240 3,000 8,000 5,000

Source: Calculated by ERM from data provided in Stewardship Solutions (2008)

B.6 Lamp Disposal

B.6.1   Waste Disposal in New Zealand

The New Zealand Environment Report (MfE, 2007b) provides an overview of waste and disposal routes in New Zealand.  This states that much of the solid waste generated in NZ is disposed of to landfills and cleanfills, although industrial waste such as that produced by agricultural, forestry, quarrying and mining activities is generally disposed of on site to dedicated industrial landfills. 

Based on information provided by the Ministry for the Environment (MfE, 2007), it is considered that all mercury-containing lamps used for domestic and industrial purposes in New Zealand are disposed of to municipal landfill or are collected for recycling.  A proportion may also be dumped but there is no data to clarify this and dumping has not been assessed as a disposal route.  Based on data provided by Stewardship Solutions (2008), an estimated level of 9% recycling has been estimated for HID lamp types.  In the absence of better data, the study assumes the same recycling rate for all other lamp types.  The proportion of mercury-containing lamps being disposed of through other routes is considered to be negligible.  Table B6a shows the current disposal routes for mercury-containing lamps in New Zealand for 2007 and for the product stewardship scenarios that have been assessed that represent increased recycling and recovery levels. 

Mercury is defined as a class 6.1b substance (HSNO, 1996) which is classified as hazardous.  According to the Hazardous Substances (Disposal) Regulations 2001 (HSNO 1996) in New Zealand the allowable treatment includes depositing the substance in a landfill, incinerator, or a sewage facility if this facility will treat the substance by changing its characteristics or composition so that the substance is no longer a hazardous.  Dilution of the substance with any other substance before discharge into the environment is not permitted. 

Under current NZ regulations, spent lamps are considered hazardous and are unacceptable for landfill when the concentration of mercury exceeds 0.2mg/L in a Toxicity Characteristic Leaching Procedure (TCLP) test (MfE, 2004; cited in Knapp B et al, unknown date).  The TCLP limit is based on US Environmental Protection Agency (USEPA) guidelines in accordance with testing of mercury leaching characteristics.  The regulations also prohibit the high-temperature incineration of hazardous waste, with the exception of some medical waste.  Consequently, no lamps are disposed of by incineration in New Zealand.

Table B6a End-of-life Management for Mercury-Containing Lamps in New Zealand

 

Unit Baseline Scenario 1: 50% recovery and recycling Scenario 2: 80% recovery and recycling Scenario 3: 100% landfill
Domestic waste          
Municipal landfill % 91% 50% 20% 100%
Cleanfill % 0% 0% 0% 0%
Hazardous landfill % 0% 0% 0% 0%
Industrial landfill % 0% 0% 0% 0%
Municipal incineration % 0% 0% 0% 0%
Hazardous incineration % 0% 0% 0% 0%
Recycling % 9% 50% 80% 0%
Industrial waste          
Municipal landfill % 91% 50% 20% 100%
Cleanfill % 0% 0% 0% 0%
Hazardous landfill % 0% 0% 0% 0%
Industrial landfill % 0% 0% 0% 0%
Municipal incineration % 0% 0% 0% 0%
Hazardous incineration % 0% 0% 0% 0%
Recycling % 9% 50% 80% 0%

Source: MfE (2007)

B.6.2   Waste Disposal in Victoria, Australia

Waste is categorised as three types in Victoria:

  • municipal solid waste, from households and council operations;
  • construction and demolition waste; and
  • commercial and industrial waste.

Of the total waste generated in 2006–2007, just under four million tonnes were disposed of to licensed landfills and over six million tonnes were recovered for reprocessing.  Almost half of the recovered material came from the construction and demolition sector (Victorian Government Department of Sustainability and Environment, 2009).

In the absence of better data, the study assumes that all waste from recycling of lamps is waste disposed to sanitary landfill and there is no waste disposed to incineration.  Refer to Section B.

B.6.3   Municipal Landfill Disposal in New Zealand and Australia

Sanitary landfill scenarios were constructed on the basis of the methods set out in Doka (2007) and Doka (2007a) from ecoinvent.

The sanitary landfill models for New Zealand and Australia account for short term emissions, which are those that are considered to occur in less than 100 years.  Further assumptions, inclusions and exclusions are noted below.

Rainfall

The sanitary landfill models account for the average rainfall in the two regions.  It is assumed that New Zealand (rainfall of 1102mm/year) has similar rainfall to Zurich (1089mm/year), which is the Ecoinvent default.  Australian rainfall is based on average data for Victoria (654mm/year).  The models account for the change in waste water treatment burdens (form reduced leachate) into account and change in release factors for each element. Revised element release factors were recalculated for Australia using the method presented in Doka (2007) associated with reduced values for rain infiltration rate and effective annual leachate.

Composition

Elemental lamp mass and composition is accounted for. This study has accounted for the release of metallic elements and compounds. All other materials e.g. glass and metal casing are assumed to be inert and have no impact.

Emissions of compounds are modelled as emissions of elements based on molecular mass.  Yttrium and Yttrium compounds have been excluded.  Cerium terbium magnesium aluminate and phosphor powder have been modelled as barium magnesium aluminate (BMA) due to lack of specific life cycle inventory data.  BMA (Al2Ba2Mg2O7) is not fully accounted for, as oxygen is excluded.   

Fate

Three pathways have been assumed for the fate of each metal in the model as follows:

  • gas;
  • effluent; and
  • sludge resulting from the treatment of landfill leachate. 

Gas is assumed to be flared with the emissions released to atmosphere, effluent is assumed to be released and sludge is assumed to be used in land applications.  Transfer coefficients and burdens associated with sludge digestion have been excluded due to a lack of available data.  Emissions of iodine to water and land modelled have been modelled as iodide.

Decomposition

A 100% decomposition rate is assumed for all trace metals.  This means that the waste has 100% degradability and is completely decomposed in the first 100 years after the waste placement.  According to Ecoinvent, this does not equate to total emission from the landfill, due to re-precipitation, storage deposits and delayed emissions. There is no uncertainty variation used in the model, as it is assumed that the uncertainty of the short-term transfer coefficients already covers possible variations in the decomposition process.

Release factors

Release factors are presented in Table B6b. It is possible for some elements to have a calculated release factor of greater than 100%; however, this is corrected to 100% to mass balance the equations.   

Table B6b   Calculated Release Factors for Municipal Landfill Disposal

Element  / Chemical Assumption Release Factor (%)
Aluminium (Al) ERM estimated. 5.00%
Antimony (Sb) - 10.50%
Barium (B) Assumed to be same as Mn 115.00%
Barium magnesium aluminate (BMA) ERM calculated as release of individual elements by molecular mass. 82.00%
Cerium terbium magnesium aluminate Assumed to be same as BMA 82.00%
Iodine (I) Assumed to be same as Cl 255.00%
Lead (Pb) - 0.59%
Manganese (Mn) - 115.00%
Mercury (Hg) - 9.59%
Phosphor (P) Assumed to be same as BMA 82.00%
Sodium (Na) - 414.00%
Vanadium (V) Assumed to be same as Sb (soluble oxianion) 10.50%

Source: Calculated by ERM from Doka (2007).

Fraction of Element Released in Gas or in Leachate

The fraction of each element that is released as a gas or in the leachate is based on Ecoinvent data presented in Table B6c

Table B6c Fraction of Element Released in Gas or in Leachate

Element  / Chemical Assumption % Released to Air % Released to Leachate
Aluminium (Al) - 0.025% 99.975%
Antimony (Sb) - 0.025% 99.975%
Barium magnesium aluminate (BMA) ERM calculated as release of individual elements by molecular mass. 52.000% 48.000%
Cerium terbium magnesium aluminate Assumed to be same as BMA 52.000% 48.000%
Iodine (I) - 1.380% 98.620%
Lead (Pb) - 0.033% 99.967%
Manganese (Mn) - 0.025% 99.975%
Mercury (Hg) - 28.600% 71.400%
Phosphor (P) Assumed to be same as BMA 52.000% 48.000%
Sodium (Na) - 0.025% 99.975%
Vanadium (V) - 0.025% 99.975%

Source: Calculated by ERM from Doka (2007).

Transfer coefficients

The transfer coefficient of an element to landfill gas is based on the product of the decomposition rate, release factor and fraction of the element released as a gas for a given element.  The transfer coefficient of an element to leachate is based on the product of the decomposition rate, release factor and remaining amount of the fraction left (subtracting the amount released as gas). 

It should be noted that parameters for compounds were modelled as a proportioned average of element parameters.

Metals in leachate are subject to waste water treatment and were apportioned between effluent and sludge using the Ecoinvent dataset presented in Table B6d

Figures are not available for some elements. For these elements estimates are made based on their aqueous chemistry.

Table B6d Transfer Coefficients for Metals in the Waste Water Treatment Process

Element  / Chemical

Assumption

Transfer coefficient to raw sludge (%) Transfer coefficient to effluent (%)
Aluminium (Al) - 95% 5%
Antimony (Sb) 50% to each compartment is assumed. 50% 50%
Barium magnesium aluminate (BMA) 50% to each compartment is assumed. 50% 50%
Cerium terbium magnesium aluminate 50% to each compartment is assumed. 50% 50%
Iodine (I) Iodine is assumed to be completely dissolved. 0% 100%
Lead (Pb) - 90% 10%
Manganese (Mn) - 50% 50%
Mercury (Hg) - 70% 30%
Phosphor (P) 50% to each compartment is assumed. 50% 50%
Sodium (Na) Sodium is assumed to be completely dissolved. 0% 100%
Vanadium (V) 50% to each compartment is assumed. 50% 50%

Source: Calculated by ERM from Doka (2007).

Landfill burdens

Process-specific burdens associated with the operation of a sanitary landfill, including infrastructure material, land transformation and occupation, are included in the model. 

Waste water treatment plant burdens were sourced from Doka (2007)b. These burdens are associated with energy used for pumping and heating, the sewer and plant infrastructure, transport and disposal of grit waste.  Only base plant burdens are accounted for; burdens for each individual element have not been included. Burdens associated with digestion have been excluded.

B.7  Lamp Collection for Recycling

Spent lamps are collected by Interwaste NZ Ltd (Interwaste) through the following five main collection routes:

  1. electricians;
  2. wheelie bins;
  3. retail drop-off;
  4. domestic post; and
  5. one-off collections. 

Lamps used by industry (HIDs, LFLs or CFLs) are either collected by electricians in recycling boxes (containing 50 or 100 lamps for LFLs) or delivered to Interwaste in wheelie bins.  Domestic lamps (CFLs) are sent to Interwaste through a free post recycling system, dropped off at retail outlets, or collected on a one-off basis. 
Interwaste has four collection depots in New Zealand, based in Auckland, Wellington, Christchurch and Dunedin, where waste lamps are deposited. 

Spent lamps collected in the South Island are crushed in a mobile crushing machine and packed into 200 litre steel drums, at either the Christchurch or Dunedin depot.  The 200 litre containers are then shipped to CMA Eco Cycle (Eco Cycle) in Melbourne for recycling and recovery of materials.   

Waste lamps that are collected in Wellington are consolidated and transported in bulk to Auckland, where they are crushed along with the Auckland waste lamp collection, and shipped in 200 litre steel drums to Eco Cycle in Melbourne, Australia. 

Primary data has been gathered from Interwaste to describe the number of lamps collected through each collection route to each depot, transport distances and vehicle types, process inputs and outputs for the lamp crushing machines, and specifications of packaging.  These data were collated and tonne kilometres calculated for each transport stage for lamp collection and for the transfer of lamps to Melbourne for recycling, as shown in Table B7d.  Where collection was part of another journey this has been allocated accordingly.  Vehicles are assumed to 100% full on outbound journey and 50% full on return. 

B.7.1   Material and Energy Inputs at Interwaste

Table B7a shows the quantity of electricity consumed for each lamp during the recycling process at Interwaste in New Zealand.  This was based on estimated data provided by Interwaste and lamp mass.

Table B7a Electricity Consumption for Lamp Crushing Process at Interwaste

 

1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
  kWh kWh kWh kWh kWh kWh
Electricity use per lamp 0.00069 0.00032 0.00069 0.00087 0.0014 0.00096

Source: Interwaste New Zealand Ltd 2009

Raw materials used in the crushers at Auckland and Christchurch/Dunedin are fuel for vehicles on site, plastic packaging, cleaning products, PPE, and replacement parts for equipment.  Tables B7b and B7c show the amount of materials consumed per lamp type, based on lamp mass.

Table B7b   Process Inputs at Interwaste’s Auckland Crushing Machine

 

1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
Propane (l) 2.6415E-05 1.21069E-05 2.6415E-05 3.30187E-05 5.283E-05 3.66398E-05
Packaging –Plastic wrap (kg) 1.98112E-05 9.08015E-06 1.98112E-05 2.47641E-05 3.96225E-05 2.74798E-05
Cleaning Product – Surfactant (l) 1.32075E-05 6.05344E-06 1.32075E-05 1.65094E-05 2.6415E-05 1.83199E-05

Source: Interwaste New Zealand Ltd 2009

Table B7c  Process Inputs at Interwaste’s Christchurch/Dunedin Crushing Machine

 

1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
Diesel (l) 0.000398 0.000182421 0.000398 0.000497512 0.000796 0.000552
PPE –Polycarbonate goggles (Kg) 0.000239 0.000109453 0.000239 0.000298507 0.000478 0.000331
Chutes and Chains – Steel (kg) 6.63E-06 3.04035E-06 6.63E-06 8.29187E-06 1.33E-05 9.2E-06

Source: Interwaste New Zealand Ltd 2009

Table B7d Estimated Transport for Spent Lamp Collection by Interwaste and Shipping to Eco Cycle (Australia)


Transport Stage

1. Linear fluorescent lamp (LFL)
35W, T8

2. Compact fluorescent lamp (CFL)
11W: external ballast

3. Compact fluorescent lamp (CFL)
20W: integral ballast

4. High pressure sodium (HPS)
150W

5. Metal Halide (MH)
400W

6. Mercury Vapour (MV)
250W

Assumptions

Inventory Dataset from Ecoinvent 2.0 (2007)

  tkm tkm tkm tkm tkm tkm    
Transport by all collection routes to Christchurch depot 0 0 0 0.0034 0.0054 0.0037 Collection – 10t truck Transport, lorry 7.5-16t, EURO4/RER SNI
Transport by all collection routes to Wellington depot 0.0002 0.001 0 0 0 0 Collection – 3.5t van Transport, van <3.5t/RER SNI
Transport by all collection routes to Auckland depot 0.0006 0 0 0 0 0 Collection – 6t truck Transport, lorry 3.5–7.5t, EURO4/RER SNI
Transport by all collection routes to Dunedin depot 0 0 0 0 0 0 Collection – 4.9t truck Transport, lorry 3.5–7.5t, EURO4/RER SNI
Transport of crushed lamps from Wellington to Auckland depot 0.0086 0.0101 0.0220 0 0 0 Transfer – 44t truck (Wgtn to Akl depot) Transport, lorry >32t, EURO4/RER SNI
Transport of drums from Dunedin to Christchurch depot 0.0056 0.0005 0.0011 0.0005 0.4796 0.0006 Transfer – rail (Dun to ChCh depot) Transport, freight, rail/RER SNI
Shipping of drums from Auckland or Christchurch depots to Melbourne 0.1490 0.1144 0.2497 0.2997 0.4796 0.3326 Transfer – container ship (Akl or Chch to Melbourne) Transport, transoceanic freight ship/OCE SNI
Transport of drums from Melbourne port to Eco Cycle plant 0.0022 0.0010 0.0023 0.0027 0.0043 0.0030 Transfer – 44t truck (port to Eco Cycle) Transport, lorry >32t, EURO4/RER SNI

Source: Interwaste (2009)

B.7.2     Process Wastes at Interwaste

It is assumed that all wastes produced at Interwaste collection depots are disposed of as industrial waste. 

Table B7e shows the types and weights of packaging associated with collection and transportation of waste lamps to Interwaste depots.  

Cardboard boxes and wheelie bins used for lamp collection are sent back to customers for reuse if in good condition, otherwise they are recycled.

Table B7e  Composition of Packaging used for Spent Lamp Collection

Material Unit LFL 100 LFL 50 CFL Globe Box Wheelie Bin Steel Drum
Cardboard kg 1.500 1.00 0.800 - -
Plastic Liner kg 0.002 0.002 0.002 - -
Plastic kg - - - 15.000 -
Steel kg - - - - 12.000
Total   1.502 1.002 0.802 15.000 12.000

Source: Interwaste New Zealand Ltd (2009)

B.8    Lamp Recycling Process

When the crushed lamps reach Eco Cycle in Melbourne, they are mixed with other mercury-containing and hazardous waste products, such as batteries, switches, thermometers, military wastes and catalysts.  Once the pure materials are extracted from these waste products, they are then distilled and recycled into marketable materials.  The recycling process for the lamps involves the following stages:

B.8.1   Crushing and Dry Separation

Used or obsolete mercury-containing lamps are processed in a machine that crushes and separates the lamps into three categories: glass, end-caps and a mercury/phosphor powder mixture (Basel Convention, 2007). 

This process involves crushing the spent tubes, and sieving and air stripping the large particles from the mercury-containing phosphor powder (The Solid Waste Association of North America, unknown date).  During crushing, mercury vapours are contained and filtered to eliminate airborne mercury emissions (Conrad & Associates Ltd, 2000).

B.8.2   Mercury Distillation

The phosphor powder is then retorted under vacuum and heat. The mercury is volatilized and then distilled to the required purity.  The recycling of mercury-containing lamps is a proven technology capable recovering greater than 99% of the mercury in the spent lamps (The Solid Waste Association of North America).

B.8.3   Production of Marketable By-products

At Eco Cycle, glass, aluminium, mercury and phosphor powder are produced as marketable by-products from the lamp recycling process.  These by-products are described in more detail in Section B10 below.  The estimated transportation (tonne kilometres) of these by-products from Eco Cycle to place of use is shown per lamp in Table B8a below. 

Table B8a   Transportation of Marketable By-Products from EcoCycle to Place of Use

  1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
  tkm tkm tkm tkm tkm tkm
Trucking from EcoCycle to place of use 5.53E-02 5.16E-02 5.29E-02 5.69E-02 6.13E-02 5.64E-02

Source: EcoCycle (2009b)

B.8.4   Material and Energy Inputs at EcoCycle

Table B8b shows the amount of energy consumed by each lamp type during the recycling process at Eco Cycle.  This was calculated using the mass of one lamp, divided by the total mass of lamps recycled per month, multiplied by the total energy consumption (kWh) used at the plant per month.

Table B8b  Electricity Consumption for Lamp Recycling Process at EcoCycle

  1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
  kWh kWh kWh kWh kWh kWh
Australia -Eco Cycle 0.039 0.018 0.039 0.049 0.079 0.055

Source: Eco Cycle (2009a)

The only additives used in the recycling process at Eco Cycle are nitrogen and oxygen gas.  The amount of each gas used per lamp type is shown in Table B8c.  This was calculated using the mass of one lamp, divided by the total mass of lamps recycled per month, multiplied by the total nitrogen and oxygen consumption (m3) used at the plant per month.

Table B8c  Gas Consumption for Lamp Recycling Process at Eco Cycle.

  1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
  kg kg kg kg kg kg
Nitrogen 0.0032 0.0014 0.0032 0.0039 0.0063 0.0044
Oxygen 0.0062 0.0028 0.0062 0.0077 0.0123 0.0086

Source: Eco Cycle (2009b), Engineering Toolbox (2009)

B.8.5   Process Wastes at EcoCycle

The lamp recycling processes at Eco Cycle does not generate any emissions to air or water.  The only solid wastes produced are packaging products, which are all recycled.  Steel drums for example are reused locally, or flattened and shipped to China for recycling.  Cardboard packaging is sent to the neighbouring paper recycling facility.  Copper contained in crushing machine filters is extracted, recycled, and sold to overseas markets.

B9 Electricity Generation

Section B9 presents the life cycle inventory modelling used for electricity generation for New Zealand and for countries where electricity is used for overseas manufacture of lamps, as listed in Section B4. 

B.9.1  Generation Mix

The electricity generation mix for each country that manufactures lamps were sourced from the International Energy Agency electricity statistics (IEA, 2007), which represents the generation mix for 2006. 

The electricity mix for New Zealand, as shown in Table B9a, was sourced from the Ministry of Economic Development (MED) Energy Data File June 2008 (MED, 2008), which represents the generation mix for 2007.

Table B9a  Electricity Generation Mix for New Zealand in 2007

Generation Fuel %
Hydro 54.9
Gas 26.4
Geothermal 7.7
Coal 6.9
Oil -
Wind 2.2
Biogas 0.4
Biomass 1.4
Total 100

Source: MED (2008)

Further descriptions of each generation fuel type are described below.

B.9.2   Transmission and Distribution losses

Power losses arise during the transmission and distribution of electricity, due to resistive and inductive losses in transformers and overhead lines. High voltage (HV) losses are referred to as transmission losses and medium-voltage (MV) and low-voltage (LV) losses are referred to as distribution losses.

Based on data provided by the MED (2008), the New Zealand national loss ratio (which represents low voltage losses) was 7.7% in 2007.  We have estimated HV and MV losses based on typical losses in other countries, where HV losses are often around 5% of total LV loss and MV losses around 10% of total LV loss.  For New Zealand this equates to an estimated loss of 0.4% for HV and 0.8% for MV.

B.9.3   Imports and Exports

The effect of imports or exports of electricity for overseas countries has not been included in the generation mix to describe each country’s profile.  Imports or exports of electricity are not relevant for New Zealand. 

B.9.4   Life Cycle Inventories – Overseas

Table B9b shows inventory datasets used for estimating electricity generation by fuel type for overseas countries where lamp manufacture occurs.  These datasets have been estimated based on the Ecoinvent 2.0 (2007).  Infrastructure has been excluded.  These datasets represent average European technology.

Table B9b Inventory Datasets Used for Electricity Generation

Generation Fuel Inventory Dataset from Ecoinvent 2.0 (2007)
Hydro Electricity, hydropower, at power plant/SK SNI
Gas Electricity, natural gas, at power plant/CENTREL SNI
Geothermal Electricity, hydropower, at power plant/SK SNI
Coal Electricity, hard coal, at power plant/SK SNI
Oil Electricity, oil, at power plant/SK SNI
Wind Electricity, at wind power plant/RER SNI
Biogas Electricity, at cogen ORC 1400kWth, wood, allocation exergy/CH SNI
Biomass Electricity, at cogen ORC 1400kWth, wood, allocation exergy/CH SNI

B.9.5   Life Cycle Inventories – New Zealand

Hydro Generation

Generation of electricity from hydro plants in New Zealand consists of around 50 generation plants, primarily from reservoir plant schemes and a small proportion of run-of-river and combined pumped storage (Wikipedia, 2009). 

Generation of electricity from hydro in New Zealand has been assumed to be all from reservoir plant schemes and modelled using Ecoinvent 2.0 (2007) Electricity, hydropower, at reservoir power plant/CH which represents European data.  The dataset represents a head height over 30m (which is analogous to New Zealand).  It includes the area occupied, an estimation of greenhouse gas emissions (based on Swiss conditions from the water reservoir), maintenance and infrastructure.

Gas Generation

Gas used for electricity generation is supplied from several gas fields both onshore and offshore in New Zealand.  Based on data from the Energy Data File June 2008 (MED, 2008) approximately 71% is onshore and is 29% is offshore, based on delivered energy. 

In terms of generators, Contact Energy Limited (operating Otahuhu B, Taranaki Combined Cycle and New Plymouth) and Genesis Power Limited (operating Huntly including the new e3p combined cycle plant) are the main electricity generators in New Zealand using natural gas.  Based on the data shown in Table B9c which shows a summary for New Zealand, a national average generation efficiency of 48% is estimated, based on the proportion of generation (by energy delivered) for the different gas turbine technologies. 

Table B9c  Electricity Generation via Gas in New Zealand

Generator Electricity Generated in 2007 Proportion of Electricity Generated in 2007 Generation Efficiency Technology
Genesis Power GWh % %  
Units 1–4 4239 33% 36% Conventional boiler and steam turbine technology.  Can burn coal, gas or mixture of both
Unit 5 3187 25% 57% Closed cycle gas turbine
Unit 6 149 1% 41% Open cycle gas turbine
Contact Energy        
Otahuhu B 2441* 19% 55.5% Combined cycle gas turbine
Taranaki 2300* 18% 55.5% Combined cycle gas turbine
New Plymouth (decommissioned part way through 2007) 610* 5% 35% Steam Turbine
Otahuhu A (reactive power – used to balance electricity grid) - - - Gas fired power station

Source: (Contact, 2008, 2009, 2009a and Genesis, 2008) 

*Note: estimated figures based on total thermal generation of 5351GWh and plant capacity (Contact, 2008). 

Generation of electricity from gas in New Zealand has been modelled using Ecoinvent 2.0 (2007) Electricity, natural gas, at power plant/UCTE and adjustments have been made to account for onshore or offshore gas source along with average generation efficiency.  No adjustments have been made for the emission of trace metals (including that for mercury) as data are not available for composition of gas in New Zealand.  This provides a reasonable estimate for trace metals emissions based on European data. 

Geothermal Generation

Generation of electricity from geothermal sources in New Zealand has been estimated based on life cycle inventory data for hydro generation, as described above.  No dataset for geothermal generation was available for the study from published and licensed sources.  The Ecoinvent dataset has been modified to include an estimate of fugitive methane emissions based on data for Contact Energy (2008), which reports 90tonnes of CO2e per GWh, which is principally methane.  The dataset assumes 90% methane and 10% carbon dioxide, based on global warming potential. 

Coal Generation

Generation of electricity from coal in New Zealand is provided by Huntly power station operated by Genesis Power.  This power station operates at a generation efficiency of 36% (Genesis, 2009).  In 2007, coal used in electricity generation was supplied form Huntley coalfield (approximately 37%) and imports from Indonesia (approximately 67%), based on data from the Energy Data File June 2008 (MED, 2008) and Genesis (2009).   Transport of coal from these locations has been estimated in the inventory dataset. 

Typical New Zealand coal composition has been accounted for in order to estimate trace metals emissions to air from coal generation.  Table B9d shows the estimated average trace metals composition of coal supplied to Huntly power station based on elemental analysis data (CRL, 2004) proportion of coal supplied from New Zealand and Indonesia.  

ERM has calculated air emissions of metals based on assumptions provided in Ecoinvent (Dones, 2007, Röder, 2007) which calculate emissions of trace metals to air from coal generation by accounting for elementary analysis of the coal, the element-specific emission factor to air and emission control equipment.  Table B9d shows the estimated emission factors to air for electricity generation. 

Table B9d  Coal Composition for Trace Metals and Power Station Emission Factor to Air

Element Estimated Concentration Estimated release factor to air
  ppm %
Aluminium 3293.6 10%*
Antimony 1.0 0%
Arsenic 5.2 1%
Barium 35.8 10%*
Bismuth 0.4 10%*
Boron 130.1 30%
Cadmium 0.2 0%
Caesium 0.4 10%*
Chromium 3.3 10%*
Cobalt 2.8 0%
Copper 3.9 10%*
Fluorine 22.9 90%
Iron 2329.0 10%*
Lanthanum 1.4 0%
Lead 3.2 10%*
Lithium 6.6 10%*
Manganese 24.1 10%*
Mercury 0.1 90%
Molybdenum 1.0 0%
Nickel 5.7 0%
Rubidium 2.3 10%*
Selenium 5.2 15%
Silver 0.4 10%*
Strontium 34.2 0%
Thallium 0.2 0%
Tin 2.2 10%*
Uranium 0.2 0%
Vanadium 5.9 0%
Zinc 7.8 0%

Source: (Röder, 2007, CRL, 2004)

* Note: These factors are estimated based on average of other metals shown. 

Generation of electricity from coal in New Zealand has been modelled using Ecoinvent 2.0 (2007) Electricity, hard coal, at power plant/UCTE and adjustments have been made to account for trace metals emissions to air and transport of raw coal, as described. 

Wind Generation

Based on data provided in the Energy Data File June 2008 (MED, 2008) and NZWEA (2009) New Zealand constitutes approximately the following wind generation profile by turbine capacity:

  • 53% are 3MW wind turbines;
  • 18% are 2MW wind turbines;
  • 28% are 1.65MW wind turbines; and
  • 1% are 500kW wind turbines.

Generation of electricity from wind has been modelled using Ecoinvent 2.0 (2007) Electricity, at wind power plant/RER, which represents 800kW European technology.  This provides a reasonable estimate based on plant capacity in New Zealand.  The dataset includes the operation of the wind power plant with the necessary change of gear oil.

Biogas Generation

Generation of electricity from biogas represents combustion of biogas and landfill gas sources.  The life cycle inventory is modelled using Ecoinvent 2.0 (2007) Electricity, at cogen with biogas engine, allocation exergy/CH, which represents European technology for use of biogas in a cogeneration unit.  Included are emissions to air, biogas consumption and use, and disposal of operational supplements.  This multi-output process delivers the co-products of heat and electricity.  The allocation is based on the exergy values which allocates electricity as 1 and heat as 0.17. 

Biomass Generation

Generation of electricity from biomass represents combustion of residential firewood wood and woody biomass from wood pellets (MED, 2008).  This life cycle inventory is modelled using Ecoinvent 2.0 (2007) Electricity, at cogen ORC 1400kWth, wood, allocation exergy/CH, which represents European technology for combustion of natural wood chips, including, the wood input, emissions to air, transport of the fuel, and disposal of the ashes.  This multi-output process delivers the co-products of heat and electricity.  Allocation is based on the exergy. 

B.10 Avoided Products

At EcoCycle the following by-products are produced from the lamp recycling process:

  • Nickel end caps are assumed to be recycled and stored onsite.
  • Brass end components are assumed to be recycled and stored onsite.
  • Aluminium end caps are sent to another CMA recycling facility in New South Wales where they are processed into ingots for foundry application.
  • Glass crushed into granules (2-3mm) and sent to a local business in Victoria where it is further processed into glass wool for insulation.
  • Mercury is processed into ingots and sent to a local dental industry where it is reused in the manufacture of dental amalgam.
  • Phosphorpowder is produced in small volumes and supplied to the agricultural industry to use as a soil enhancer.
  • Plastics are assumed to be recycled and stored onsite.

Table B10a shows the inventory datasets used for determining the environmental benefits associated with avoided materials from the cycling of lamps.  Based on data provided by Eco Cycle, the study assumes that 99% of materials from the lamps are recycled.  Adjustments are made for electricity generation and transportation, assuming all mercury is sourced from China and all other materials from within Australia.

Table B10a  Avoided Products Resulting from Lamp Recycling at Eco Cycle

Avoided Product Description  Inventory Dataset from Ecoinvent 2.0 (2007)
Nickel

Nickel is assumed to be 99.5% nickel, represented by an Ecoinvent (2007) dataset

Dataset originally represents 2006 data of a global production mix.  Some processes originate from European datasets.   Amendments made by ERM include the change of input electricity mix to match the country of recovery and recycling.  Processes includes: mining, beneficiation and disposal of overburden and tailings. Subsequently it includes the metallurgy step with the disposal of slag and the separation of the co-product copper, and the refining step yielding the desired Class I nickel. Production, application and emissions of most agents and additives used in beneficiation and metallurgy are also included.  Overburden and tailings are disposed and partly re-filled. Sulphur dioxide in the off-gas is recovered.

Nickel, 99.5%, at plant/GLO SNI 
Brass

Brass is represented by an Ecoinvent (2007) dataset.

Dataset originally represents 2003 technology of the EU average.   Amendments made by ERM include the change of input electricity mix to match the country of recovery and recycling.  Processes include copper and zinc, including their melting and casting of brass ingots. 

Brass, at plant/kg/CH
Mercury

Mercury is assumed to be liquid mercury, represented by an Ecoinvent (2007) dataset

Dataset originally represents technology from no specific geographic origin.  Amendments made by ERM include the change of input electricity mix to match the country of recovery and recycling.

Processes include raw materials and energy consumption for production, estimated emissions to air from production. No water emissions.  Technology uses data from lime mining, crushing and milling plus estimation of the additional furnace operation step.  Refer to Appendix B.2 for description on the mix of mercury production used to offset this inventory.

Mercury, liquid, at plant/GLO SNI
Glass

Glass is assumed to be flat glass, represented by an Ecoinvent (2007) dataset

Dataset originally represents 2000 technology in Germany (EU).   Amendments made by ERM include the change of input electricity mix to match the country of recovery and recycling.  Processes include all measured in-and output materials and energy carriers reported for a glass coating plant during operation (coating process, internal transports, packing and administration).  The flat glass coating technology includes the following stages: metal coating is applied to float glass by cathodic sputtering in vacuum.

Flat glass, uncoated, at plant/RER SNI
Phosphor powder

Phosphor powder is assumed to be liquid white phosphorous, represented by an Ecoinvent (2007) dataset

Dataset originally represents 2003 technology of no specific geographic origin.   Amendments made by ERM include the change of input electricity mix to match the country of recovery and recycling.  Processes include raw materials and chemicals used for production, by-products and waste produced, transport of materials to manufacturing plant, emissions to air and energy demand.  Technology uses production from phosphate rock with the aid of the Wöhler process, with a yield of 91%.

Phosphorus, white, liquid, at plant/RER SNI
Aluminium

Aluminium is assumed to be primary aluminium, represented by an Ecoinvent (2007) dataset

Dataset originally represents 2003 technology for EU processes.  Amendments made by ERM include the change of input electricity mix to match the country of recovery and recycling.  Processes include cast aluminium ingot production, transport of materials to the plant and disposal of wastes.   

Aluminium, primary, at plant/RER SNI
Plastics

Plastics are assumed to be polyurethane, represented by an Ecoinvent (2007) dataset.

Dataset originally represents 2003 technology for the EU.  Amendments made by ERM include the change of input electricity mix to match the country of recovery and recycling.  Processes include transport of the monomers as well as the production (energy, air emissions) of the PUR foam.

Polyurethane, rigid foam, at plant/RER SNI

The transportation of these avoided products from place of extraction to EcoCycle are estimated in tonne kilometres (Table B10b).  It is assumed that mercury is extracted from China and shipped to Melbourne.  All other avoided products are assumed to be sourced from Australia.

Table B10b  Transportation of Avoided Products from Place of Extraction to EcoCycle

  1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
  tkm tkm tkm tkm tkm tkm
Shipping from country of extraction to Melbourne port 3.73E-05 4.66E-05 4.66E-05 4.66E-04 4.66E-04 3.54E-04
Trucking from place of extraction (within Australia) and from port to EcoCycle 1.60E-02 1.28E-02 1.82E-02 1.58E-01 1.54E-01 3.59E-02

Source: EcoCycle (2009b)

B11    Sensitivity Analysis

Section B11 provides inventory data in relation to sensitivity analyses conducted for the LCA study. 

B11.1 Mercury Levels

Typical mercury levels have been assumed to increase or decease by 20 percent, depending on the design and quality of lamp, as shown in Table B11a.  It is considered that the range selected will take into account any mercury reductions in future lamp designs.  Specific data from manufacturers regarding lamp design and potential effects on lamp life and quality were not readily available for the study.  Although Philips (Ross, 2009) indicated the following when discussing reduced mercury levels:  

For fluorescent domestic lighting (LFL, CFLe and CFLi):

  • Light quality.  Light quality with regards to colour rendering (reproduction of colour) is unchanged.  The colour appearance can also change.  For example, low mercury levels may result in slight flickering of lamps and the tube shows a “pink” tinge.  This effect can be noted in other brands towards end of life if the phosphorous coating on the inside glass of the tube is not thick/even enough, and mercury is absorbed into the glass.
  • Lamp life.  The two main indicators of lifetime are emitter coating level and mercury level.  Excess levels of mercury allow longer life; however the production goal is to make market leading lifetime values for standard product with minimum mercury levels. 
  • Lamp outputs (W).  Output unchanged.  Although light output levels can decrease as electrons have less interaction with mercury molecules.
  • Other design changes.  Entry level lamps have all been optimised with lowest mercury levels, along with other material construction.  Under current processes, any further minimisation of lamp components will result in reduced performance.

For HID lamp (HPS, MH and MV):

  • Lamp life.  Higher mercury levels lead to longer lifetime.  It is not clear what relationship exists and if this relationship is linear. 

For this sensitivity analysis, we have assumed that in the future (egg in 5 years time) technology improvements will have been made to allow a further 20% reduction in mercury level which will achieve the same functional performance of lamp life, light quality and output for the purposes of this sensitivity analysis.  No changes have been made to this sensitivity beside mercury level.

Table B11a  Mercury Content in Lamps

Mercury Content Unit 1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
Minimum mg 3.2 4 4 40 40 40
Maximum mg 4.8 6 6 60 60 60

B11.2 Lamp Lifetimes

Maximum and minimum lamp lifetimes have been estimated at 50 percent above and below typical values (Table B11b).  The selected range in lifespan is assumed to cover variance between potential lamp design and quality. 

Table B11b  Range of Lamp Lifespans

  Unit 1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
Lamp wattage W 35 11 20 150 400 250
Lifespan Typical Hours 8,000 10,000 12,000 20,000 20,000 20,000
Lifespan Minimum Hours 4,000 5,000 6,000 10,000 10,000 10,000
 Lifespan Maximum Hours 12,000 15,000 18,000 30,000 30,000 30,000

Source: Calculated by ERM from data provided in Stewardship Solutions (2008) B.6  Lamp Disposal

B11.3 Light Fittings and Fixtures

The materials used in light fittings vary widely, depending on application and purpose.  Examples of typical light fittings are shown below (Figure B11a), and typical construction materials and weights are provided in Table B11c.  These materials do not include ballasts and additional lamp gear.

The sensitivity assumes one set of packaging per new lamp and one set of fittings per 100,000 hours of operation (or approximately 12 years for continuous operation).  For domestic lamp this may represent an underestimate for impacts of fittings, while industrial lamps this may represent and over estimate.  For CFL lamp two lamps are used in each fitting and for all other lamp types one lamp is sued per fitting.

Figure B11a.  Examples of typical light fittings: General Use Luminaire (left), Fluorescent Batten Fitting (centre) and Low Bay HID fitting (right)

    

Table B11c  Construction Materials and Estimated Weights of Typical Lamp Fittings

Fluorescent Batten Fitting
Linear fluorescent lamp (LFL)
Dimensions L 1225mm X W 58 mm X H 48mm  
Materials and Weights Nylon end cap
Cold rolled sheet
Epoxy powder
0.2 Kg
0.49 Kg
0.02 Kg
Total weight of Fitting   0.7 Kg
General Use Luminaire
Compact Fluorescent Lamps (CFLe and CFLi)
Dimensions W 297mm X L 297mm X H 83mm   
Materials and Weights Sheet steel
Enamel
Anodised aluminium
Cold rolled sheet
Epoxy powder
1.2 kg
0.2 kg
0.4 kg
0.99 kg
0.001 kg
Total Weight of Fitting   2.85 kg
Low Bay HID Fitting
High Pressure Sodium Lamps (HPS), Metal Halide Lamps (MH) and Mercury Vapour Lamps (MV)
Dimensions W 610mm x L 370mm x H 200mm  
Materials and Weights Pressed steel sheet
Epoxy resin
Integral control gear
Aluminium reflector
Ceramic lamp holder
Steel press frame
Shock resistant glass
3.5 kg
0.001 kg
0.059 kg
1.0 kg
0.14 kg
1.0 kg
0.5 kg
Total Weight of Fitting   6.2 kg
High Bay HID Fitting
High Pressure Sodium Lamps, Metal Halide Lamps (MH) and Mercury Vapour Lamps (MV)
Dimensions L 252mm X W 135mm X H 290mm
Bowl Diameter 475mm X H 270mm
 
Materials and Weights Cold rolled sheet steel
White powder finish
Aluminium spun reflector
Wire guard
Shock resistant glass
3.099 kg
0.001 kg
0.9 kg
0.09 kg
0.31 kg
Total Weight of Fitting   4.4 kg

Source: Estimated weights by ERM based on total weight data from All Products (2009)

B11.4 Packaging

Packaging used for lamp retail also varies widely depending on product brand.  Lamps may be purchased individually or in bulk.  Table B11d shows estimated typical packaging and material weights used for each lamp.

Table B11d Typical Packaging Types and Estimated Weights

  1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
Packaging Type Round cardboard tube Cardboard box with cardboard inserts Cardboard box with cardboard inserts Cardboard box
Corrugated cardboard sleeve
Cardboard box
Corrugated cardboard sleeve
Cardboard box
Corrugated cardboard sleeve
Dimensions of Packaging L 1215mm X W 30mm X  H 30mm L 110mm X W 60mm X  H 60mm L 115 mm X W 65mm X  H 65mm L 215mm X W 50mm X  H 50mm L 292mm X W 123mm X H 123 L 228mm X W 93mm X  H 93mm.
Cardboard weight 91 g 26 g 30 g 90 g 312 g 184 g

Source: Estimated packaging weights by ERM from based on data from Philips (2009), Eye Lighting (2009)

B11.5 Closed Loop Recycling

A closed loop recycling system assumes that all waste lamps are returned to their place of manufacture and recycled into new lamps.  Under this assumption, transport measured in tonne kilometres has been estimated for each lamp type (Table B11e).  These values are based on the transport for lamp import in Table B4c, with the addition of the weight of the steel drums used to contain crushed waste lamps.  Inventory data to represent shipping and trucking are based on Ecoinvent (2007) datasets.  Environmental benefit is attributed to 99% of the recovered materials for the avoided materials production that would have otherwise taken place.  1% represents a recycling process wastage.

Table B11e  Estimated Tonne Kilometres for Transporting Crushed Lamps back to Place of Manufacture

  1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
  tkm tkm tkm tkm tkm tkm
Trucking Distance – From Place of Use to Port in New Zealand 0.05 0.02 0.05 0.06 0.10 0.07
Shipping – From New Zealand Port to Port in Country of Manufacture 1.46 0.61 1.46 2.27 3.75 1.82
Trucking  Distance –  From Port to Place of Manufacture 0.05 0.02 0.05 0.06 0.10 0.07

Source: Statistics New Zealand (2008), Sea Rates (2009).

B11.6 Cut-offs for Material Production

Filling gases (argon and krypton) have been omitted from the study as the volume contained in lamps is considered negligible and due to a lack of data on the quantities contained in each lamp.  These noble gases are naturally occurring in the earth’s atmosphere.  They are chemically unreactive gases.  Environmental databases (Ecoinvent 2007) show low energy requirements for the production of argon and krypton.  Based on this information and the substantially low amounts contained in lamps, it is considered acceptable to not take their environmental impact into account for the LCA study.

B11.7 Warm-up Effects

Some research indicates (Parsons, 2006) that there is a warm-up effect for fluorescent lamps which increases power required in the initial period of operation compared to rated power.  

Based on data provided in (Parsons, 2006) for an 18W CFL lamp, ERM estimates that around 0.5% additional energy may be required for a CFL lamp operated over an a typical one hour period.  The effect of increased power at warm-up is dependent on the increased power required for warm-up (rated at 19W over 10minutes for an 18W lamp), as well as the time operated between warm-ups.  An increased time between warm-ups of more than one hour would reduce the additional energy requirements.   Based on one hour between warm-ups we estimate an increased energy consumption of 0.5% in the use phase.  We have assumed this is the case for all lamps, although this is an assumption for non-CFL lamp types.

Table B11f  Lamp Electricity Consumption

  Unit 1. Linear fluorescent lamp (LFL)
35W, T8
2. Compact fluorescent lamp (CFL)
11W: external ballast
3. Compact fluorescent lamp (CFL)
20W: integral ballast
4. High pressure sodium (HPS)
150W
5. Metal Halide (MH)
400W
6. Mercury Vapour (MV)
250W
Lamp wattage W 35 11 20 150 400 250
Lifespan Hours 8,000 10,000 12,000 20,000 20,000 20,000
Typical energy use kWh 280 110 240 3,000 8,000 5,000
Energy use with warm-up effects  kWh 281.4 110.6 241.2 3,015 8,040 5,025

Source: Calculated by ERM from data provided in Stewardship Solutions (2008)