In a cement kiln, calcium carbonate (CaCO₃) is broken down to CaO and carbon dioxide. The methodology used pertains to the IPCC Tier 2 approach (IPCC, 2000). The emission was estimated from the annual UK production of clinker (DETR, 2000d) and the IPCC default emission factor of 138.3 t C/ kt clinker produced (IPCC, 1997). So far it has not proved possible to obtain data on the CaO content of clinker either on a national basis or a site basis so a default is used. A revision included in the 1999 Inventory is to correct for the loss of cement kiln dust and here the default of 2% is used (IPCC, 2000). Hence the corrected emission factor is 141.1 t C/ kt clinker. The clinker produced is then ground up with gypsum to produce cement. Since clinker may be imported or exported the production of cement is not precisely related to the production of clinker. Hence it is preferable to use clinker data to estimate carbon dioxide emissions.

Section 2.6.1.1 describes the emission of CO₂ from the degradation of limestone used in cement manufacture. The NAEI category cement (fuel combustion) covers emissions of CO₂, methane and N₂O arising from the combustion of fuels in kilns. Emission factors are shown in Table A10. In DUKES (DTI, 2000), this fuel consumption is included under ‘industry’. Fuel estimates were derived from Blyth et al (1996) and British Cement Association (2000). The BCA data supplied an estimate of total kiln fuel in 1998 and the Blyth et al study gave a break down by fuel type used in cement and plaster production for 1993. Using these data and extrapolating to other years by using clinker production data (DETR, 2000), an approximate time series for cement coal, oil, petroleum coke and gas consumption was developed. Having estimated the consumption of coal, fuel oil and natural gas by the cement sector, the fuel consumption by other industry was adjusted so that the total consumption agreed with DUKES (DTI, 2000). Estimates were also made of emissions from the combustion of scrap tyres and waste oils in cement kilns. The scrap tyre data were taken from Collingwood (1997) and DTI (2000a). The consumption of waste oils is uncertain. Some estimates for 1994-95 are available (Collingwood, 1997) but for subsequent years we used the advice from BCA that the consumption of wastes as fuel was no more than 6%. (British Cement Association, 2000).

The non-CO₂ factors for scrap tyres are the same as for coal, because this fuel is normally burnt mixed with a larger proportion of coal. The carbon content of scrap tyres was estimated from data reported in Ogilvie, (1995).

For the pollutants NO_x, CO, NMVOC, PM₁₀ and SO₂, estimates were based on the site emissions for cement plant reported by the Environment Agency (2000). The aggregate factors were calculated based on estimated plant capacities supplied by British Cement Association (2000). They are expressed in terms of the emission per tonne of clinker produced. And are shown in Table A11.

a Emission as carbon

b British Coal (1989)

c UKPIA (1989)

d British Gas (1992)

e Based on Ogilvie (1995)

f As gas oil (UKPIA,1989)

g Brain et al (1994)

h EMEP/CORINAIR (1996)

i CORINAIR (1992)

j Fynes et al (1994)

Lime (CaO) is manufactured from limestone (CaCO₃ ) and dolomite (CaCO₃MgCO₃) by heating in a lime kiln resulting in the evolution of carbon dioxide. UK limestone use is given in BGS(2001) and an emission factor of 120 t carbon/kt limestone was used based on the stoichiometry of the chemical reaction assuming pure limestone. For dolomite an emission factor of 130 t carbon /kt dolomite would be used, however dolomite calcination data is not given explicitly but included in the limestone data. Hence emissions will be under estimated. Dolomite calcination is believed to be a small proportion of the total hence the under estimate is unlikely to be significant. The limestone calcination data used exclude limestone calcined in the chemical industry since a large proportion of this is used in the production of sodium carbonate by the Solvay process. The limestone is calcined to produce carbon dioxide but nearly all this is recovered and is sequestrated in the sodium carbonate produced. Also the calcination of limestone in the sugar industry is excluded because, the lime produced is used in sugar refining and the overall process does not result in a net emission of carbon dioxide. Some of the subsequent uses of sodium carbonate result in the emission of CO₂ (e.g. glass manufacture) but others do not (e.g. water softening). More data on sodium carbonate consumption by use is required before an estimate can be made. Hence the current estimate of lime calcination emissions are low.

Emissions from the use of limestone and dolomite were estimated from the following sources:

• iron and steel manufacture
• glass manufacture
• agricultural soils (liming of soils)

Usage data is available in BGS(2001) and ISSB(2000) for iron and steel use. The emission factors were 120 t carbon/kt limestone or chalk and 130 t carbon /kt dolomite and assume all the carbon is released to the atmosphere. Soil liming is reported under 10 Agriculture, Forestry and Land Use Change. The industrial emissions are reported under 4 Production Processes in Industry.

Some of the uses of sodium carbonate result in the emission of CO₂ (e.g. glass manufacture, food and drink, pharmaceuticals) but others do not (e.g. water softening). Only the emission from soda ash used in glass production is reported. This is based on estimates of the consumption of soda ash in the production of soda glass (British Glass, 2000). This is around 15% of the mass of glass produced. (British Glass, 1998). An emission factor of 113 kt C/ Mt soda ash is used. The estimates include both container and flat glass.

The inventory reports emissions of NMVOCs from asphalt paving and road construction. The emission estimates are based on consumption data of bitumen emulsions, cut-back bitumen and cut-back fluxes. The emission factors used are 7, 87.5 and 700 kg/t for each component respectively. Emissions from asphalt roofing are not reported due to lack of data.

The manufacture of nitric acid produces emissions of both NO_x and nitrous oxide. Up to 1988 estimates of NO_x are estimated from the annual production of nitric acid (CIS, 1991). The NO_x emission factor is 3.98 t/kt of 100% acid produced. This is an aggregate factor based on CORINAIR (1989) emission factors for the different types of processes ranging from 3-12 t/kt of 100% acid produced. The aggregate factor was based on data on UK Manufacturing plant provided by the Nitric Acid Association for the year 1985 (Munday, 1990)

UK production data have not been published since 1988. Since then a number of plants have either closed down, changed ownership, moved or have been fitted with abatement technology. Since 1994 estimates of NO_x were made based on returns from manufacturers. Emissions from 1989 to 1993 were estimated by linear interpolation.

All plant operators were able to supply data on nitrous oxide emissions for 1999 (Dupont, Kemira, Hydro Agri, Terra Nitrogen). For previous years, emissions were calculated for each plant by extrapolating the 1999 emission (or in some cases the 1998 emission) using production or plant capacity data. One plant currently has abatement measures so the uncontrolled default factor was used for the years prior to the retrofit. The default emission factor was 6kt/Mt 100% acid produced. This is an average factor based the range quoted in IPCC Guidelines (IPCC, 1997) for medium pressure plant. Up to 1988, the production data from (CIS, 1991) was used. Since 1994 production estimates based on returns from manufacturers were used. These are based mainly on plant capacity data which may over estimate true production levels. Production from 1989 to 1993 was estimated by linear interpolation. In the period 1990 to 1993, two other plants were known to operate however these have now closed down. Their emissions are estimated based on plant capacity and the default emission factor. Some nitric acid capacity is associated with the adipic acid plant. From 1990-1993, its emissions are reported combined with the adipic acid plant emissions.

Adipic acid is manufactured in a multi-stage process from cyclohexane via oxidation with nitric acid. Nitrous oxide is produced as a breakdown product from nitric acid. UK production figures and emission estimates have been provided by DuPont as the only company manufacturing adipic acid in the UK (DuPont, 2000). The estimates are based on an emission factor of 0.3 kt N₂O/Mt adipic acid produced. A small nitric acid plant is associated with the adipic acid plant which also emits nitrous oxide. From 1994 onwards this emission is reported as nitric acid but prior to 1994 it is included under adipic acid production. This allocation reflects the availability of data. In 1998 an N₂O abatement system was retrofitted to the plant. This has resulted in a 96% reduction in emissions in 1999.

Emissions of PM₁₀ from mining and quarrying were estimated using USEPA (1997) factors. This gives an average factor of 0.1g/kg of material throughput. Clearly this is a rather crude estimate and so no time series has been estimated.

Emissions of methane from the chemical industry are reported under 2B5. The emission is the sum of all emissions of methane from chemical processes reported in the Pollution Inventory (Environment Agency, 2000). The Pollution Industry does not fully identify the processes by type, hence it is not possible to disaggregate the emission by process. Also, activity data are not available for the sources listed in the IPCC Guidelines. Hence the total emission is reported for all years. It is expected that the detail of these estimates will be improved in future.

Emissions of NMVOC from chemical industry processes are based on data contained in the Pollution Inventory, as well as data received from the Scottish Environment Protection Agency and from individual process operators. The development of a time series is difficult because of the lack of activity data, and the interpretation of the available emissions data is complicated by the likelihood that earlier estimates contained in the Pollution Inventory did not generally include emissions from fugitive sources. A correction factor, based estimates of point and fugitive emissions for a limited set of processes, is applied for Pollution Inventory data before 1998. From the 1998 data set onwards, operators have been required to report fugitive emissions. This is not believed to have occurred fully during 1998, and so a correction factor with a value of half that used in 1997 is applied. The 1999 Pollution Inventory data are used without a correction factor.

Emissions of sulphur dioxide from sulphuric acid production are based on detailed manufacturers’ returns compiled by NSSA (2000). One other chemical process, involving the manufacture of pigments and dyes, has significant emissions of SO₂. Emission estimates, included in the inventory as the source ‘chemical industry (pigment manufacture)’, are based on data from the Pollution Inventory (Environment Agency, 2000). Emissions of chromium from a process manufacturing chromium based chemicals are included as the source ‘chemical industry (chromium chemicals). This is a major source of chromium and data are taken from the Pollution Inventory. Further data on historic emissions have been provided by the process operator which, because received after the finalisation of the 1999 inventory, are not taken into account in the current estimates. This will be rectified in the next inventory. Emission estimates for heavy metals and particulate matter from other chemical processes, based on Pollution Inventory data, are included in the source ‘chemical industry’

Although it is likely that almost all manufacturing processes give rise to some emissions of particulate matter, the inventory contains estimates for those where emissions are expected to be significant. Currently, estimates are made for the following:

• primary lead/zinc production
• secondary copper production
• secondary lead production
• other (Part A) non-ferrous metal processes
• chemicals manufacture
• glass production
• asphalt manufacture (roadstone coating)
• cement batching
• other Part B (LAPC) processes
• construction

Emission estimates for the first five sources are based on Pollution Inventory data, with the full time series being completed by extrapolation of the Pollution Inventory data based on metal production in the first three cases and based on DTI indices of manufacturing output in the other two cases.

Emission factors for glass production are based on information contained in the European IPPC Bureau’s BREF note for glass manufacture (EIPPCB, 2000), while the emission factors for asphalt manufacture is given in the US EPA compilation of emission factors (USEPA, 1997).

Emission estimates for other Part B processes are based on emission estimates for ‘typical’ processes, available from the UK Emission Factor Database at http://rsk.co.uk/ukefd/ multiplied by estimates of the number of processes in the UK. These estimates are subject to a very high level of uncertainty.

The emission estimates for construction are based on an emission estimate for 1990 presented in the third QUARG report (QUARG, 1996). Estimates for other years are extrapolated from this figure on the basis of the value of construction work done.

Natural gas is used as a feedstock for the manufacture of ammonia (for fertilizer), methanol and acetic acid. The largest use is for ammonia manufacture by the steam reforming of natural gas to make hydrogen: The simplified reactions are:

CH₄ + H₂O Û CO + 3H₂

CO + H₂O Û CO₂ + H₂

The hydrogen is then reacted with nitrogen from air to form ammonia

N₂ + 3H₂ Û 2NH₃

If there is no use of the by-products CO and CO₂ formed then these are emitted to atmosphere. The CO is oxidised to CO₂ prior to emission. Hence the CO₂ emission can be estimated from the natural gas usage or the amount of ammonia produced. In principle the emission is 0.97 t CO₂ /t NH₃ produced based on the reaction stoichiometry.

In the UK some ammonia plants are integrated with methanol and acetic acid manufacture for greater efficiency. Thus hydrogen formed as a by-product from acetic acid manufacture is used as the feedstock for ammonia manufacture. Some carbon monoxide and carbon dioxide from the reforming process is used to manufacture methanol. This carbon is sequestrated as methanol and is not emitted to atmosphere.

Methanol is manufactured from natural gas using a process similar to the steam reforming process:

CO + 2H₂ Û CH₃OH

so that all the carbon content of the natural gas is sequestrated as methanol.

Acetic acid is manufactured from methanol and natural gas and again the carbon content of the natural gas is sequestrated.

Two estimates were made:

• The amount of CO₂ emitted from ammonia manufacture
• The amount of natural gas used in the manufacture of products. This can then be deducted from the total combustion of gas by industry in order to calculate the combustion emissions of the non- CO₂ pollutants.

The procedure adopted for the emission of CO₂ from ammonia manufacture was:

1. Data on the plant capacity, natural gas consumption or CO₂ emission from

ammonia plant

acetic acid plant

methanol plant

were collected from manufacturers. This included a breakdown between natural gas used as a feedstock and natural gas used as a fuel.

2. The ammonia capacity of the plants using hydrogen by-product from acetic acid manufacture was excluded.
3. Corrections were made based on manufacturers advice on the ‘recovery ‘ of carbon in methanol manufacture.

The procedure used to estimate the natural gas use as a feedstock was to perform a carbon balance over the three processes:

1. Methanol plant capacity data was used to estimate its natural gas use
2. The natural gas usage of the acetic acid plant was available
3. The natural gas use equivalent to the CO₂ emission from ammonia manufacture was calculated
4. The total feedstock use of natural gas was estimated as the sum of items 1-3

It is known that some of the CO₂ produced is recovered and sold for use in the food industry and nuclear industry. Because all this carbon dioxide is eventually emitted, the estimates are compiled including sales. This procedure will be more reliable than trying to identify the end use of the carbon dioxide particularly as some carbon dioxide used in the food industry arises from fermentation.

The Inventory includes an estimate of the NO_x emission from the ammonia reformer reported under ammonia combustion. This arises from the combustion of natural gas to produce the high temperatures required by the process. The estimate was based on data provided by the manufacturers.

The necessary data were supplied by Terra Nitrogen, Kemira, ICI and BP Chemicals.

Aluminium is produced by the electrolytic reduction of alumina in large pots. During the reduction, the carbon anode is consumed resulting in the emission of CO₂ , SO₂ and other pollutants. In the UK most aluminium is produced by the prebaked anode cell process, though one plant operates the older Soderberg Cell process. Emissions were estimated based on the production of aluminium for each type of process and the carbon emission factors shown in Table A13. The carbon emission factors reflect current practice, and higher emission factors were used for earlier years. For the other pollutants, emission factors are available for anode baking as well as production.

a CO₂ as carbon, Alcan (1997).

b Alcan (2000)

Emissions from limestone use in blast furnaces are discussed in 2.6.3. The following emissions are also estimated and reported under production processes.

• Iron and Steel Blast Furnaces: CO₂ and other process emissions (see Section 2.5.5)
• Flaring of blast furnace gas
• Electric arc furnaces
• Basic Oxygen Furnaces.

Electric arc furnaces are used in the production of stainless and mild steel and also for recycling scrap. Emissions are based mainly on default emission factors taken from the EMEP /CORINAIR (1999) Chapter on electric arc furnaces. The CO₂ emission arises from the consumption of a graphite anode and is based on manufacturer’s data.

Environment Agency (2000) has reported large emissions of carbon monoxide from iron and steel processes. Their precise sources are not identified but we were advised that the main sources were sinter plant and basic oxygen furnaces (BOF). The reported emissions were allocated to basic oxygen furnaces and sinter plant in proportion to the USEPA (1997) emission factors for uncontrolled plant. Sinter plant emissions are reported under combustion in industry and emission factors are given Table A5.

Table A14: Emission Factors for Electric Arc and Basic Oxygen Furnaces

a BISPA(1997)

b EMEP/CORINAIR (1999)

c Emission factor for 1999 based on Environment Agency (2000)

e USEPA (1997)

Emissions of NMVOC are estimated from the hot rolling and cold rolling of steel using emission factors 1 g/tonne product and 25 g/tonne product respectively (EMEP/CORINAIR, 1996). Activity data is taken from ISSB(2000).

There is insufficient activity or emission factor data to make an estimate for emissions from ferroalloys. Emissions of CO₂ will be included in 2C1, since the fuels used as reducing agents are included in the energy statistics.

Activity data for food and drink is taken from ONS (2001), and supplementary data are obtained from the Scotch Whisky Association (2000) and the Gin and Vodka Producers Association (2001).

Emissions of VOC from refinery processes are based on data provided by the United Kingdom Petroleum Industry Association (UKPIA, 2000), who have provided data on an annual basis since 1994. The UKPIA data refer to the following installations:

• Texaco, Milford Haven

• Elf, Milford Haven

• BP, Coryton

• Shell, Shell Haven (closed during 1999)

• Conoco, South Killingholme

• Lindsey, Killingholme

• Shell, Stanlow

• PIP, North Tees

• Esso, Fawley

• BP, Grangemouth

• Gulf, Milford Haven (closed during 1997)

The UKPIA data for refinery processes are split into four categories – tankage, flares, drainage systems and process emissions (vents and fugitive emissions). UKPIA also supply estimates for loading of petrol into road and rail tankers at refineries – see section 2.7.2.7.

Prior to 1994, process emissions are estimated by extrapolation from the 1994 figure on the basis of refinery throughput, whereas emissions from tankage, flares and drainage systems are assumed to be constant.

Emissions from specialist refineries producing bitumens, lubricants, and solvents are included in petroleum processes (see section 2.6.16).

This sector covers VOC emissions from specialist refineries (Llandarcy, Eastham, Dundee, & Harwich), onshore oil production facilities, and miscellaneous petroleum processes not covered elsewhere in the inventory (most significant of which are the Tetney Lock and Tranmere oil terminals, and the production of aromatic hydrocarbons from petroleum feedstocks on Teeside). Emissions are taken from the Pollution Inventory ( Environment Agency, 2000). No emissions data have been found for the Dundee refinery.

This sector covers VOC emissions from onshore gas production facilities, refining and odourisation of natural gas, natural gas storage facilities, and processes involving reforming of natural gas and other feedstocks to produce carbon monoxide and hydrogen gases. Emissions are taken from the Pollution Inventory ( Environment Agency, 2000)..

An estimate of 1.9 ktonnes of VOC emitted annually is included in the inventory for ship purging, based on the work of Rudd & Mikkelson (1996), who concluded that. Ship purging refers to the occasional practice of vapours from ships’ tanks being vented to atmosphere following delivery of a volatile cargo, in order to prevent contamination of the following cargo. In previous versions of the NAEI, ship purging has been included in the estimate for the chemical industry but has now been separated in order to improve the level of detail in the inventory.