'HWHULRUDWLRQIDFWRUV - Danish emission inventories for road transport and other mobile sources

For three-way catalyst cars the emissions of NO

, NMVOC and CO gradually increase due to catalyst wear and are, therefore, modified as a function of total mileage by the so-called deterioration factors. Even though the emission curves may be serrated for the individual vehicles, on average, the emissions from catalyst cars stabilise after a given cut-off mileage is reached due to OBD (On Board Diagnostics) and the Danish inspection and maintenance programme.

For each forecast year, the deterioration factors are calculated per first registration year by using deterioration coefficients and cut-off mileages, as given in EMEP/CORINAIR (2007), for the corresponding layer. The deterioration coefficients are given for the two driving cycles: ”Urban Driving Cycle” (UDF) and ”Extra Urban Driving Cycle” (EUDF: urban and rural), with trip speeds of 19 and 63 km/h, respectively.

Firstly, the deterioration factors are calculated for the corresponding trip speeds of 19 and 63 km/h in each case determined by the total cumu-lated mileage less than or exceeding the cut-off mileage. The Formulas 4.3 and 4.4 show the calculations for the ”Urban Driving Cycle”:

$ 07& 8 8

8')= ⋅ +

, MTC < U

MAX

(4.3)

% 0$;

$ 8 8

8')= ⋅ +

, MTC >= U

MAX

(4.4)

where UDF is the urban deterioration factor, U

and U

the urban dete-rioration coefficients, MTC = total cumulated mileage and U

MAX

urban cut-off mileage.

In the case of trip speeds below 19 km/h the deterioration factor, DF, equals UDF, whereas for trip speeds exceeding 63 km/h, DF=EUDF. For trip speeds between 19 and 63 km/h the deterioration factor, DF, is found as an interpolation between UDF and EUDF. Secondly, the dete-rioration factors, one for each of the three road types, are aggregated into layers by taking into account vehicle numbers and annual mileage levels per first registration year:

∑

⋅

=

₍ ₎

) (

, ,

, )

(

) (

, , , /<HDU M

M )<HDU

L L\ L\

\ L M

/<HDU

M )<HDU

L L\ L\

') 1

0 1 ')

') ^(4.5)

where DF is the deterioration factor.

For N

O and NH

, COPERT IV takes into account deterioration as a lin-ear function of mileage for gasoline fuelled EURO 1-4 passenger cars and light duty vehicles. The level of emission deterioration also relies on the content of sulphur in the fuel. The deterioration coefficients are given in EMEP/CORINAIR (2007), for the corresponding layer. A cut-off mileage of 120.000 km (pers. comm. Ntziachristos, 2007) is behind the calculation of the modified emission factors, and for the Danish situation the low sulphur level interval is assumed to be most representative.

&DOFXODWLRQPHWKRG

(PLVVLRQVDQGIXHOFRQVXPSWLRQIRUKRWHQJLQHV

Emissions and fuel-use results for operationally hot engines are calcu-lated for each year and for layer and road type. The procedure is to com-bine fuel consumption and emission factors (and deterioration factors for catalyst vehicles), number of vehicles, annual mileage levels and the relevant road-type shares given in Table 3.1. For non-catalyst vehicles this yields:

\ M

\ M N

\ N M

\ N

() 6 1 0

(

, ,

=

, ,

⋅ ⋅

⋅

(4.6)

Here E = fuel consumption/emission, EF = fuel consumption/emission factor, S = road type share and k = road type.

For catalyst vehicles the calculation becomes:

\ M

\ M N

\ N M

\ N

') () 6 1 0

(

, ,

=

, ,

⋅

, ,

⋅ ⋅

⋅

^(4.7)

([WUDHPLVVLRQVDQGIXHOFRQVXPSWLRQIRUFROGHQJLQHV

Extra emissions of NO

, VOC, CH

, CO, PM, N

O, NH

and fuel con-sumption from cold start are simulated separately. For CO

and SO

, the extra emissions are derived from the cold start fuel consumption results.

In terms of cold start data for NO

, VOC, CO, PM and fuel consumption no updates are made to the COPERT IV methodology, and the calcula-tion approach is the same as in COPERT III. Each trip is associated with a certain cold-start emission level and is assumed to take place under ur-ban driving conditions. The number of trips is distributed evenly across the months. First, cold emission factors are calculated as the hot emission factor times the cold:hot emission ratio. Secondly, the extra emission fac-tor during cold start is found by subtracting the hot emission facfac-tor from the cold emission factor. Finally, this extra factor is applied on the frac-tion of the total mileage driven with a cold engine (the β -factor) for all vehicles in the specific layer.

The cold:hot ratios depend on the average trip length and the monthly ambient temperature distribution. The Danish temperatures for 2006, 2005 and 2004 are given in Cappelen et al. (2007, 2006, 2005). For 2000-2003, 1990-1999 and 1980-1989 the temperature data are from Cappelen (2004, 2000 and 2003). The cold:hot ratios are equivalent for gasoline fu-elled conventional passenger cars and vans and for diesel passenger cars and vans, respectively, see Ntziachristos et al. (2000). For conventional gasoline and all diesel vehicles the extra emissions become:

) 1

(

, , ,

= ⋅ 1 ⋅ 0 ⋅ () ⋅ &(U −

&(

β

M\ M\ 8 M\

^(4.8)

Where CE is the cold extra emissions, β = cold driven fraction, CEr = Cold:Hot ratio.

For catalyst cars, the cold:hot ratio is also trip speed dependent. The ratio is, however, unaffected by catalyst wear. The Euro I cold:hot ratio is used for all future catalyst technologies. However, in order to comply with gradually stricter emission standards, the catalyst light-off temperature must be reached in even shorter periods of time for future EURO stan-dards. Correspondingly, the β-factor for gasoline vehicles is reduced step-wise for Euro II vehicles and their successors.

For catalyst vehicles the cold extra emissions are found from:

) 1

(

, ,

=

UHG

⋅

(852,

⋅

8 M\

⋅

(852,

−

1 0 () &(U

&( β β ^(4.9)

where β

red

= the β reduction factor.

For CH

, specific emission factors for cold driven vehicles are included in COPERT IV. The β and β

red

factors for VOC is used to calculate the cold driven fraction for each relevant vehicle layer. The NMVOC emissions during cold start are found as the difference between the calculated re-sults for VOC and CH

.

For N

O and NH

, specific cold start emission factors are also proposed

by COPERT IV. For catalyst vehicles, however, just like in the case of hot

emission factors, the emission factors for cold start are functions of

cu-mulated mileage (emission deterioration). The level of emission deterio-ration also relies on the content of sulphur in the fuel. The deteriodeterio-ration coefficients are given in EMEP/CORINAIR (2007), for the corresponding layer. For cold start, the cut-off mileage and sulphur level interval for hot engines are used, as described in the deterioration factors paragraph.

(YDSRUDWLYHHPLVVLRQVIURPJDVROLQHYHKLFOHV

For each year, evaporative emissions of hydrocarbons are simulated in the forecast model as hot and warm running losses, hot and warm soak loss and diurnal emissions. For evaporation, no updates are made to the COPERT IV methodology, and the calculation approach is the same as in COPERT III. All emission types depend on RVP (Reid Vapour Pressure) and ambient temperature. The emission factors are shown in Ntziachris-tos et al. (2000).

Running loss emissions originate from vapour generated in the fuel tank while the vehicle is running. The distinction between hot and warm run-ning loss emissions depends on engine temperature. In the model, hot and warm running losses occur for hot and cold engines, respectively.

The emissions are calculated as annual mileage (broken down into cold and hot mileage totals using the β-factor) times the respective emission factors. For vehicles equipped with evaporation control (catalyst cars), the emission factors are only one tenth of the uncontrolled factors used for conventional gasoline vehicles.

) )

1 ((

1 0 +5 :5

5

_M_\

=

_M_\

⋅

_M_\

⋅ − β ⋅ + β ⋅ ^(4.10)

where R is running loss emissions and HR and WR are the hot and warm running loss emission factors, respectively.

In the model, hot and warm soak emissions for carburettor vehicles also occur for hot and cold engines, respectively. These emissions are calcu-lated as number of trips (broken down into cold and hot trip numbers using the β-factor) times respective emission factors:

) )

1 ((

+6 :6

O 1 0 6

WULS

\ M

= ⋅ ⋅ − β ⋅ + β ⋅ ^(4.11)

where S

is the soak emission, l

trip

= the average trip length, and HS and WS are the hot and warm soak emission factors, respectively. Since all catalyst vehicles are assumed to be carbon canister controlled, no soak emissions are estimated for this vehicle type. Average maximum and minimum temperatures per month are used in combination with diurnal emission factors to estimate the diurnal emissions from uncontrolled ve-hicles E

(U):

) ( 365

)

(

8 1 H 8

(

_M_\^G

= ⋅

_M_\

⋅

(4.12)

Each year’s total is the sum of each layer’s running loss, soak loss and

diurnal emissions.

)XHOFRQVXPSWLRQEDODQFH

The calculated fuel consumption in COPERT III must equal the statistical fuel sale totals according to the UNFCCC and UNECE guidelines for emissions reporting. The statistical fuel sales for road transport are de-rived from the Danish Energy Authority data (see DEA, 2007). The DEA data are further processed for gasoline in order to account for e.g. non road and recreational craft fuel consumption, which are not directly stated in the statistics, please refer to paragraph 5.5.5 for further informa-tion regarding the transformainforma-tion of DEA fuel data.

A balance between estimated fuel consumption and fuel sold is obtained by means of the so called fuel scale factors, derived separately for gaso-line and diesel. The latter factors go into the equations 4.6-4.11 by multi-plication with the annual mileage, since this parameter is regarded as the most uncertain variable of the calculation expressions as such.

For gasoline vehicles all mileage numbers are equally scaled in order to obtain gasoline fuel equilibrium, and hence the gasoline mileage factor used is the reciprocal value of the COPERT IV:DEA gasoline fuel use ra-tio (Figure 4.3).

For diesel the fuel balance is made by adjusting the mileage for light and heavy-duty vehicles and buses, given that the mileage figures for these vehicles are regarded as the most uncertain parameters in the diesel en-gine emission simulations. Consequently, the diesel mileage factor used is slightly higher than the reciprocal value of the COPERT IV:DEA diesel fuel use ratio (Figure 4.3).

In the figures 4.3 and 4.4 the COPERT IV:DEA gasoline and diesel fuel use ratios are shown for fuel sales and fuel consumption from 1985-2006.

The data behind the figures are also listed in Annex 8. The fuel consump-tion figures are related to the traffic on Danish roads.

)XHOVFDOHIDFWRUVEDVHGRQIXHOVDOHV

0.00 0.50 1.00 1.50 2.00

1985 198

7 198

9 1991

1993 1995

1997 1999

2001 2003

2005

Gasoline Diesel (fuel ratio) Diesel (mileage factor)

)LJXUH DEA:NERI Fuel ratios and diesel mileage adjustment factor based on DEA fuel sales data and NERI fuel consumption estimates

From the Figures 4.3 and 4.4 it appears that the inventory fuel balances for gasoline and diesel would be improved, if the DEA statistical figures for fuel consumption were used instead of fuel sale numbers. The fuel difference for diesel is, however, still significant. The reasons for this in-accuracy are a combination of the uncertainties related to COPERT IV fuel use factors, allocation of vehicle numbers in sub-categories, annual mileage, trip speeds and mileage splits for urban, rural and highway driving conditions.

For future inventories it is intended to use improved fleet and mileage data and improved data for trip speed and mileage split for urban, rural and highway driving. The update of road traffic fleet and mileage data will be made as soon as this information is provided from the Danish Ministry of Transport and Energy in a COPERT IV model input format.

The final fuel use and emission factors are shown in Annex 6 for 1990-2006. The total fuel use and emissions are shown in Annex 7, per vehicle category and as grand totals, for 1990-2006 (and NFR format in Annex 15). In Annex 14, fuel-use and emission factors as well as total emissions are given in CollectER format for 2006.

In the Figures 4.5 and 4.6, the fuel related emission factors for gasoline and diesel, respectively, are shown for CH

and N

O (from 1990-2006), and SO

, NO

, NMVOC, CO, NH

and TSP (from 1985-2006), per vehicle type for the Danish road transport.

7DEOH Fuel-specific emission factors for CO2 for road transport in Denmark Fuel type CO2 [kg/GJ]

Gasoline 73 (72.8 in 2006)

Diesel 74 LPG 65

The emission factors for CO

(Table 4.3) are country specific values, and come from the DEA. From 2006, bio ethanol has become available from a limited number of gas filling stations in Denmark. Bio ethanol is re-garded as CO

neutral and has a sulphur content of zero, and hence, the aggregated CO

(and SO

) factors for gasoline have been adjusted, on the basis of the energy content of pure gasoline and bio ethanol.

)XHOVFDOHIDFWRUVEDVHGRQIXHOFRQVXPSWLRQ

0.00 0.50 1.00 1.50 2.00

1985 1987

1989 1991

1993 1995

1997 1999

2001 2003

2005 Gasoline Diesel (fuel ratio) Diesel (mileage factor)

)LJXUH DEA:NERI Fuel ratios and diesel mileage adjustment factor based on DEA fuel consumption data and NERI fuel consumption estimates

&+IDFWRUVJDVROLQHYHKLFOHV

-2-wheelers Heavy duty vehicles Light duty vehicles Passenger cars

12IDFWRUVJDVROLQHYHKLFOHV

-2-wheelers Heavy duty vehicles Light duty vehicles Passenger cars

62IDFWRUVJDVROLQHYHKLFOHV

-12[IDFWRUVJDVROLQHYHKLFOHV

-2-wheelers Heavy duty vehicles Light duty vehicles Passenger cars

1092&IDFWRUVGLHVHOYHKLFOHV

-2-wheelers Heavy duty vehicles Light duty vehicles Passenger cars

&2IDFWRUVJDVROLQHYHKLFOHV

-2-wheelers Heavy duty vehicles Light duty vehicles Passenger cars

1+IDFWRUVJDVROLQHYHKLFOHV

-Heavy duty vehicles Light duty vehicles Passenger cars 2-wheelers

763IDFWRUVJDVROLQHYHKLFOHV

-Heavy duty vehicles Light duty vehicles Passenger cars 2-wheelers

)LJXUH Fuel related emission factors of CH4, N2O (1990-2006), SO2, NOX, NMVOC, CO, NH3 and TSP (1985-2006) for gasoline per vehicle type for Danish road transport

1RQH[KDXVWSDUWLFXODWHHPLVVLRQVIURPURDGWUDQVSRUW

The TSP, PM

and PM

2.5

emissions arising from tyre and brake wear (SNAP 0707) and road abrasion (SNAP 0708) are estimated for the years 2000-2006 as prescribed by the UNECE convention reporting format. The emissions are calculated by multiplying the total annual mileage per ve-hicle category with the correspondent average emission factors for each source type. The calculation procedure is consistent with the COPERT III model approach used to estimate the Danish national emissions coming from exhaust. A more thorough explanation of the calculations is given by Winther (2004), and emission factors are taken from EMEP/-CORINAIR (2007). The emission factors and total emissions for 2006 are shown in Annex 14.

&+IDFWRUVGLHVHOYHKLFOHV