Data sources

Source fuel consumption and emission factors

The EMPA emissions testing programme (Soltic, 2001; Stettler et al., 2004; and Weilenmann, 2005) provides a large set of driving patterns from high-speed freeway driving to urban stop-and-go driving, along with their fuel consumption and emission factors for a number of vehicles. Vehicle makes and models tested in the programme cover many of the vehicle models prevalent in the South African vehicle fleet.

Data from the EMPA emissions testing programme were used to develop the Swiss-German-Austrian Handbook of Emission Factors (SAEFL, 2004) and the PHEM instantaneous emissions model. The HBEFA has been used to develop emissions inventories in Germany, Switzerland and Austria, and PHEM has been used in road network emissions models (Linßen et al., 2005; Zallinger et al., 2005).

Driving cycles

Driving cycles used in this study include the NEDC (New European driving cycle), the German autobahn cycle and four EMPA real-world driving cycles called R1, R2, R3 and R4. (The speed-time series were obtained from the EMPA in electronic format.) Each cycle is represented in a speed-time relationship and is split into three phases. Each phase in a cycle can be referred to as a driving pattern and has a fuel consumption and emission factors for HC, NOx, CO and CO2 associated with it for each vehicle model considered.

The speed-time series for the driving cycles are included in APPENDIX A: Driving Cycles from the EMPA Testing Programme.

EMPA test vehicles

The individual makes and models of vehicles from the EMPA emissions testing programme used to develop the emissions model are listed in APPENDIX B: Vehicles from the EMPA Emissions Testing Programme. The list of vehicles is summarised in Table 3.1 terms of emissions regulations and fuel types.

Table 3.1: Number of vehicles tested in the EMPA emissions testing programme and used to develop the fuel consumption and emissions model.

Fuel

Technical details of vehicles were obtained from motoring journals (Car, various years), vehicle specification databases (Carfolio.com, Carinfo.autold.com, Globalcar.com and Vehix.com), Mead and McGrouther 2003 Vehicle Digest (2003), and the Bosch Automotive Handbook (Robert Bosch GmbH, 2000). These details include vehicle properties and performance characteristics (maximum power, maximum torque and corresponding engine speeds, engine capacity, fuel type, bore and stroke), aerodynamic properties (frontal area and drag coefficient), gear ratios and wheel sizes.

Data pre-processing

Data used during the analysis was manually copied from the EMPA reports (originals are in Adobe Acrobat PDF format), motoring journals and other sources and inserted into a Microsoft Access^® database for further analysis.

3.3.4. Analysis

Calculation of engine-operating patterns

Engine speed and engine load were calculated for every second of each combination of driving pattern and vehicle model from the EMPA emissions measurement programme.

Conventional units used to compare engine performance maps are mean piston speed in m s^-1 (which is directly proportional to engine speed in rpm and engine stroke) and bmep (brake mean effective pressure) in kPa, were used to indicate engine speed and engine load respectively.

Mean piston speed was calculated from the product of vehicle speed, gear ratios, final drive, wheel sizes, wheel slip, engine stroke and the NEDC gear changing scheme (used as the gear changing schemes in all EMPA emissions test), using Equation 4 derived from Heywood (1988:44) and Wong (2001:240):

) speed and the NEDC gear changing schema: if 0≤S<15, then i=1; 15≤S<35 i=2; 35≤S<50 i=3; 50≤S≤70 i=4; and S≥70 i=5; λ is percentage slip assumed to be 3.5%

(Wong, 2001:240); Dr is the wheel rim diameter in m; Wt is the tire width in m; and Pt is the tire profile in percentage.

Engine load was calculated using equations modified from Heywood (1988:49), Robert Bosch GmbH (2000:337) and Gillespie (1992:119):

⎟⎟

where bmepinst is instantaneous load at the fly wheel in kPa; S is speed of the vehicle in km h^-1; η is the combined efficiency of gearbox and final drive taken from Wong (2001:238) as a first approximation: 94% 1st gear, 95% 2nd gear, 96% 3rd gear, 97% 4th gear, 98% 5th gear and 95% final drive; m is mass of the vehicle in kg; a is acceleration of the vehicle in m s^-2; Rr is the rolling resistance coefficient of the vehicle (taken as 0.015); g is gravitational acceleration 9.81 m s^-2; α is inclination of the road in degrees (assumed 0 for all the cycles used by the EMPA for emissions measurements); ρ is air density taken as 1.2 kg m^-3; Cd is drag coefficient of the vehicle; Af is frontal area of the vehicle in m²; N is engine speed in rev s^-1; and Vd is engine displaced volume in litres.

Engine-operating patterns were produced by grouping calculated data points for each combination of vehicle and driving pattern into intervals of engine speed and engine load.

The intervals used for engine speed and engine load are defined in Equations 6 and 7 respectively. The number of data points in each engine speed and engine load interval was then divided by the total number of data points for the driving pattern to normalise the engine-operating patterns i.e. the sum of all the values in each interval in the patterns is equal to 1.

(j - 0.5) and (j + 0.5) in m s^-1 where j is an integer and (1≤j≤17) (6) (100(k - 1)) and (100k) in kPa where k is an integer and (1≤k≤17) (7)

Processing of the EMPA source data was performed using a set of macros and SQL (Structured Query Language) queries, which implemented the calculations described in Equations 4 to 7 within the Microsoft Access^® environment. The graphical outputs were produced using Microsoft Excel^® linked to the Access^® database. All processing was done on X86 processor based computers running the Windows XP^® operating system.

Six hundred and seventy five engine-operating patterns were produced from the combination of driving cycles and vehicles from the EMPA emissions measurement program (15 driving patterns × 45 vehicle models). These patterns are referred to here as the original engine-operating patterns and are uniquely numbered using a randomly assigned number between 0 and 676. An example of an engine-operating pattern is presented in Figure 3.2.

Figure 3.2: An example engine-operating pattern (the sizes of all the circles add up to unity and indicate the fraction of time spent in each engine speed and engine load interval).

Reducing the number of engine-operating patterns

The original engine-operating patterns calculated using the procedure above represent the demand placed on the engines of individual vehicles. Emissions inventories, however, need average fuel consumption and emission factors for groups of vehicles to be able to estimate fuel consumption and emissions for vehicle fleets.