Emissions of carbon dioxide, sulphur dioxide and black smoke from road tranport are calculated from the consumption of petrol and diesel fuels and the sulphur content of the fuels consumed. Data on petrol and diesel fuels consumed by road transport in the UK are taken from the Digest of UK Energy Statistics (DTI, 1998) and corrected for consumption by off-road vehicles.

Emissions of CO₂, expressed as kg carbon per tonne of fuel, are based on the H/C ratio of the fuel; emissions of SO₂ are based on the sulphur content of the fuel. Values of the fuel-based emission factors for CO₂, SO₂ and black smoke from consumption of petrol and diesel fuels are shown in Table A20. Values for SO₂ vary annually as the sulphur-content of fuels change and are shown in Table A20 for 1997 based on data from UKPIA (1998).

a Emission factor in kg carbon/tonne, based on UKPIA (1989)

b 1997 emission factor calculated from UKPIA (1998) figures on the weighted average sulphur-content of fuels delivered in the UK in 1997

Emissions of CO₂, SO₂ and black smoke can be broken down by vehicle type based on estimated fuel consumption factors and traffic data in a manner similar to the traffic-based emissions described below for other pollutants. The 1997 inventory used fuel consumption factors expressed as g fuel per kilometre calculated for each vehicle type and average speed from the emission functions and speed-coefficients provided by COPERT II (European Environment Agency, 1997). Average fuel consumption factors calculated from these functions are shown in Table A21 for each vehicle type, emission regulation and road type in the UK. Combining these fuel consumption factors with the traffic and fleet data discussed below enabled the total petrol and diesel consumed by road transport to be calculated. Compared with the actual Dti total fuel consumption statistics for the UK, the calculations underestimated the amount of petrol consumed by 10% and the amount of diesel consumed by 7% in 1997. These differences probably reflect the range of average speeds and drive cycles of vehicles on UK roads as well as variations in fuel efficiencies among vehicles of similar types.

Emissions of the pollutants NMVOCs, NOx, CO, benzene, 1,3-butadiene, CH₄ and N₂O are calculated from measured emission factors expressed in grammes per kilometre and road traffic statistics from the Department of Environment, Transport and the Regions (DETR, 1998b). The emission factors are based on experimental measurements of emissions from in-service vehicles of different types driven under test cycles of different average speeds. The road traffic data used are national vehicle kilometre estimates for the different vehicle types and different road classifications in the UK road network. These data have to be further broken down by composition of each vehicle fleet in terms of a) the fraction of diesel- and petrol-fuelled vehicles on the road and b) the fraction of vehicles on the road made to the different emission regulations which applied when the vehicle was first registered. These are related to the age profile of the vehicle parc.

For a particular vehicle, the drive cycle over a journey is the key factor which determines the amount of pollutant emitted. Key parameters affecting emissions are the acceleration, deceleration, steady speed and idling characteristics of the journey, as well as other factors affecting load on the engine such as road gradient and vehicle weight. However, work has shown that for modelling vehicle emissions for an inventory covering a road network on a national scale, it is sufficient to calculate emissions from emission factors in g/km related to the average speed of the vehicle in the drive cycle (Zachariadis and Samaras, 1997). Emission factors for average speeds on the road network are then combined with the national road traffic data.

DETR estimate annual vehicle kilometres for the road network in Great Britain by vehicle type on roads classified as trunk, principal and minor roads in built-up areas (urban) and non-built-up areas (rural) and motorways (DETR, 1998b). These estimates are based on traffic counts from the rotating census and core census surveys. Traffic data for Northern Ireland has been made available from a travel survey produced by Oscar Faber for the Transportation Unit of the Department of the Environment for Northern Ireland (DoE, N. Ireland, 1997). These data have been combined with the DETR data for Great Britain to produce a time-series of total UK vehicle kilometres by vehicle and road type from 1970 to 1997.

Table A22 shows the regulations which have come into force up to 1997 for each vehicle type. The year 1997 saw the introduction of the Euro II standard apply for new registrations of cars and LGVs in the UK.

The average age profile and the fraction of petrol and diesel cars and LGVs in the traffic flow each year are based on the composition of the UK vehicle fleet using DETR Vehicle Licensing Statistics. The Transport Statistics Report "Vehicle Licensing Statistics: 1997" (DETR, 1998c) either gives historic trends in the composition of the UK fleet directly or provides sufficient information for this to be calculated from new vehicle registrations and average vehicle survival rates. The vehicle licensing data are combined with data on the change in annual vehicle mileage with age to take account of the fact that newer vehicles on average travel a greater number of kilometres in a year than older vehicles. For cars and LGVs, such mileage data are from the National Travel Survey (DETR, 1998e); data for HGVs of different weights are taken from the Continuous Survey of Road Goods Transport (DETR, 1996a).

The fraction of diesel cars and LGVs in the fleet was taken from data in "Vehicle Licensing Statistics: 1997" (DETR, 1998c). For 1997, the fraction of diesel cars in the fleet was 10.7% while the fraction of diesel LGVs was estimated to be 66%.

The emission factors for NOx, CO and NMVOCs (and PM₁₀ for diesel vehicles) used for the 1997 inventory are based on data from TRL (Hickman, 1998) and COPERT II, "Computer Programme to Calculate Emissions from Road Transport" produced by the European Topic Centre on Air Emissions for the European Environment Agency (1997). Both these sources provide emission functions and coefficients relating emission factor (in g/km) to average speed for each vehicle type and Euro emission standard derived by fitting experimental measurements to some polynomial functional form. These functions were then used to calculate single emission factor values for each vehicle type and Euro emission standard at average speeds on the four specified types of roads.

Whenever possible, the emission factors were calculated from the equations provided by TRL (Hickman, 1998) from analysis of data gathered from emission measurements of on-road vehicles of different ages tested on rolling roads or engine test beds. The measurements were made by TRL on UK vehicles or were drawn from the Workbook of Emission Factors for Road Transport produced at UBA Berlin (Infras, 1995). Where data were unavailable from TRL for particular Euro standards, the data were taken from the equations recommended by COPERT II. For the same pollutant, vehicle type and emission standard, the TRL and COPERT II data gave broadly similar emission factors.

For the more recent Euro II standard (and Euro I for some vehicle types), there have not yet been any measurements of emissions from in-service vehicles. Therefore, emission factors were calculated using scaling factors relative to Euro I values taken from COPERT II or Gover et al. (1994).

For each type of vehicle, both TRL and COPERT II provide equations for different ranges of vehicle engine capacity or vehicle weight. Emission factors calculated from these equations were therefore averaged, weighted according to the proportion of the different vehicle sizes in the UK fleet, to produce a single average emission factor for each vehicle type and Euro class at each average speed. These average emission factors are given in Tables A23 to A25 for each of the different vehicle types and Euro emission regulations.

The older in-service vehicles in the test surveys that were manufactured to a particular emission standard would have covered a range of different ages. Therefore, an emission factor calculated for a particular emission standard (e.g. ECE 15.04) from the emission functions and coefficients from TRL and COPERT II is effectively an average value for vehicles of different ages which inherently takes account of possible degradation in emissions with vehicle age. However, for the more recent emission standards (Euro I and II), the vehicles would have been fairly new when the emissions were measured. Therefore, based on data from the European Auto-Oil study, the deterioration in emissions with age or mileage was taken into account for catalyst cars. It was assumed that emissions of CO and NOx increase by 60% over 80,000 km, while emissions of NMVOCs increase by 30% over the same mileage (DETR, 1996c). Based on the average annual mileage of cars, 80,000 km corresponds to a time period of 6.15 years.

Emission factors for motorcycles were taken from TRL (Hickman, 1998) at urban speeds for three different classes of motorcycles: mopeds (<50cc), >50cc, 2st and >50cc, 4st. Measures to reduce emissions from motorcycles are not taken into account in the 1997 inventory. Although the EU has adopted emission legislation for mopeds implemented in 1997, data are not currently available on the number of these that were on the road in the UK in 1997.

Emission factors for benzene were taken from two sources. For non-catalyst cars (i.e. ECE 15.00 to 15.04) and pre-Stage I diesel cars, emission factors were taken from Eggleston and Irwin (1993); for other vehicle types and emission standards (including catalyst cars), the emission factors are based on those for NMVOCs and percentage by weight of NMVOC speciation figures from COPERT II. These give figures of 3.5% of NMVOCs as benzene from petrol vehicles with three-way catalysts and 2% of NMVOCs as benzene from all diesel exhausts. Emission factors for benzene for the four road types are shown in Table A26 in units of milligrammes/km.

For 1,3-butadiene, the emission factors were taken from actual measurements of this pollutant emitted from light duty diesel vehicles over different drive cycles (urban, rural and motorway). The measurements were made at the Warren Spring Laboratory (Bailey and Schmidl, 1989) and the emission factors scaled up to determine estimated emission factors for heavy duty vehicles according to the relative fuel consumption and NMVOC emission trends of these vehicles. The 1,3-butadiene emission factors for Euro I and Euro II vehicles were derived by scaling the pre-Euro I factors according to the reductions in the NMVOC emission factors. Emission factors for 1,3-butadiene for the four road types are shown in Table A27 in units of milligrammes/km.

Hot emission factors for methane and nitrous oxide were taken from estimates for European vehicles and emission control technologies published in the IPCC Revised 1996 Guidelines for National Greenhouse Gas Inventories and are shown in Tables A28 and A29 (IPCC, 1997). Fewer measurements have been made on emissions of these pollutants from vehicles. Therefore, only average emission factors are used, covering all vehicle speed or road types. Their uncertainties can be expected to be quite large. However, the emission factors used reflect the fact that three-way catalysts are less efficient in removing methane from the exhausts than other hydrocarbons and also lead to higher N₂O emissions than non-catalyst vehicles.

For petrol cars and LGVs, the emission factors used depend on whether the vehicles are run on leaded or unleaded fuel and whether fitted with a three-way catalyst. The petrol emission factors are shown in Table A30. These are given by Gover et al (1994) and are taken from a survey of the literature by the Transport Research Laboratory. Insufficient information was available to distinguish emission rates under different driving modes and average speeds.

Data from the Digest of UK Energy Statistics on the percentage of sales of leaded and unleaded petrol and the fraction of cars on the road with catalysts calculated in the fleet model are used to derive overall weighted average emission factors for cars and LGVs in each year. The fraction of petrol cars running on leaded, unleaded without catalyst and unleaded with catalyst estimated for 1997 in the UK is also shown in Table A30.

For diesel vehicles, emission factors for PM₁₀ were calculated for the four specified types of roads from speed-emission functions taken from TRL (Hickman, 1998) and COPERT II in the same format as the pollutants NOx, CO and NMVOCs. These are shown in Table A31.

PM₁₀ emissions have been found to be dependent on the sulphur content of diesel fuel. The emission factors for pre-Euro II diesel vehicles were based on measurements of emissions from vehicles running on diesel fuel of up to 2000 ppm sulphur content, the limit for diesel sold in Europe up to 1996. In October 1996, legislation came into effect which limits the sulphur content of diesel fuel to 500 ppm. It has been estimated that this reduces primary PM₁₀ emissions from pre-Euro II diesel cars and LGVs by 2.4% and from HGVs and buses by 13%. The emission factors for these vehicles shown in Table A31 are applicable to the standard 500ppm S diesel fuel marketed in 1997.

Research from the European Auto-Oil study has suggested a deterioration in PM₁₀ emissions from diesel cars and LGVs with age or mileage and this has now been taken into account in the same way as emissions of other pollutants from catalyst cars. Based on recommendations by DETR (1996c), it is assumed that PM₁₀ emissions increase by 100% over 80,000 km for diesel cars and LGVs.

The 1997 inventory has used a new procedure for estimating cold-start emissions taken from COPERT II (European Environment Agency, 1997), taking account of the effects of ambient temperature on emission factors for different vehicle technologies and its effect on the distance travelled with the engine cold. A factor, the ratio of cold to hot emissions, is used and applied to the fraction of kilometres driven with cold engines to estimate the cold start emissions from a particular vehicle type using the following formula:

An average trip length for the UK of 8.4 km was used, taken from Andre et al (1993). This gives a value for b of 0.23.

This methodology was used to estimate annual UK cold start emissions of NOx, CO, NMVOCs, PM₁₀, benzene and 1,3-butadiene from petrol and diesel cars and LGVs. Emissions were calculated separately for catalyst and non-catalyst petrol vehicles. Cold start emissions data are not available for heavy duty vehicles, but these are thought to be negligible (Boulter, 1996). For benzene and 1,3-butadiene, the same cold start parameters were used in these equations as for total NMVOCs.

Evaporative emissions are dependent on ambient temperature and the volatility of the fuel and, in the case of diurnal losses, on the daily rise in ambient temperaure. Fuel volatility is usually expressed by the empirical fuel parameter known as Reid vapour pressure (RVP). For each of these mechanisms, equations relating evaporative emissions to ambient temperature and RVP were developed by analysis of empirically-based formulae derived in a series of CONCAWE research studies in combination with UK measurements data reported by TRL. Separate equations were developed for vehicles with and without evaporative control systems fitted such as carbon canister devices. The overall methodology is similar to that reported by COPERT II (European Environment Agency, 1997), but the data are considered to be more UK-biased.

All the equations for diurnal, hot soak and running loss evaporative emissions from vehicles with and without control systems fitted developed for the inventory are shown in Table A32.

For diurnal losses, the equations were developed from data and formulae reported by CONCAWE (1987), TRL (1993) and ACEA (1995). Based on historic trends in Meteorological Office temperature data for the UK, an average maximum daily temperature of 15^oC and an average daily diurnal rise in temperature of 9^oC was used. Current market fuel has an RVP of around 74kPa in summer blends (June - August) and 94 kPa in winter blends (September - May) (Watson, 1999). An annual average value of 90 kPa was used.

The equations specified in Table A32 give diurnal loss emissions in g/vehicle.day for uncontrolled (DL_uncontrolled) and canister controlled (DL_controlled) vehicles. Total annual diurnal losses were calculated from the equation:

For hot soak losses, the equations were developed from data and formulae reported by CONCAWE (1990), TRL (1993) and COPERT II. Based on historic trends in Meteorological Office temperature data for the UK, an annual mean temperature for the UK of 11^oC was used.

The equations specified in Table A32 give hot soak loss emissions in g/vehicle.trip for uncontrolled (HS_uncontrolled) and canister controlled (HS_controlled) vehicles. Total annual hot soak losses were calculated from the equation:

For running losses, the equations were developed from data and formulae reported by CONCAWE (1990) and COPERT II.

The equations specified in Table A32 give running loss emissions in g/vehicle.km for uncontrolled (RL_uncontrolled) and canister controlled (RL_controlled) vehicles. Total annual running losses were calculated from the equation:

Particulates are emitted by the wear of vehicle tyres and brake linings. Emission factors are based on USEPA data for emissions from a 4-wheel light duty vehicle (USEPA 1985), but have been scaled up for heavy duty vehicles according to the average number of wheels on the different vehicle classes relative to those on a light duty vehicle. The base emission factor for PM₁₀ emissions from passenger cars is 0.0012 g/km for tyre wear and 0.00795 g/km for brake wear (USEPA, 1985). Emissions can be expected to be dependent on vehicle speed and load per wheel, but this information is not known. Therefore, only speed-independent figures can be provided.