Data used in the analysis

Data used in the analysis

The required inputs for the transport module can be broadly categorized as follows: (i) Baseline vehicle stock data; (ii) Household survey data; (iii) Baseline vehicle use data; (iv) Data on operating conditions; and (v) Data on vehicle capital and operating costs.

In the first case, the data should be procured for at least 15 consecutive years. The mortality rates determined by Winfrey curve are then applied to the sales data to estimate the fraction of vehicles retiring each year over the analyzed period. The surviving vehicle stock is calibrated by adjusting the average life of each vehicle type to yield the active vehicle population numbers consistent with the exogenous data for the baseline year (e.g. from vehicle registration database).

When historic sales data for such an extended period is not available, the model is populated with in-use survey or vehicle registration data for the baseline year. In this case, the vehicles of each type are grouped by year of manufacturing, and zero mortality rate is applied (no vehicles are retired, since the input data reflect the actual fleet composition in the baseline year; the vehicles are grouped by the year of manufacturing to ensure accurate calculation of the average age).

The vehicle stock data requirements vary for each transport mode (road transport, metro and railway); however, the input data for all modes should be disaggregated by type (car class for private motor vehicles and rolling stock categories for railway transport), age, and technology (emissions standards, fuel type and consumption intensity).

The baseline vehicle stock data used in the analysis were provided by the Ministry of Internal Affairs of Georgia. The dataset comprising 749, 232 vehicles registered in Georgia as of 2011 specifies the year of manufacturing and the year of registration, vehicle model, make, type and engine displacement of each vehicle. The data were further refined and grouped to match the input requirements of the EFFECT model.^{CITATION Gla08 \n \t \l 1033} EFFECT distinguishes 13 vehicle types and 9 subtypes, classified according to vehicle size and engine displacement. The vehicle classification used in EFFECT as well as the grouping criteria applied to the vehicles listed in the registry are summarized in Table 12.

Table 12: Vehicle classification used in EFFECT

For road vehicles, the annual age-sensitive vehicle use data by vehicle type, subtype and usage pattern applied in simulations for Georgia are based on expert-determined estimates used in previous ESMAP studies. For railway and metro transport, the model allows entering the actual country-specific number (annual distance traveled, number of passengers, and loading factor) to compute the aggregate vehicle use data endogenously.

The explanatory variables employed in projecting vehicle use are closely linked to the operating conditions dataset. Thus, the average trip length determines the number of starts per day, which, in its turn, causes aggregate cold start over-emissions.

For heavy-duty vehicles, both passenger and freight, vehicle loading has a significant impact on fuel consumption and emissions. Usually, a 50% load condition is used to determine average fuel consumption to account for deadheading (the truck running empty on the return journey or to pick up a load) and to cover the frequent occurrence for lighter weight loads where the volumetric capacity of the truck is reached before its maximum weight limitation. These assumptions have been applied for the scenarios modeled for this study.

• 
  Ambient temperature. Low ambient temperatures may cause over-emissions due to cold starts and evaporative emissions. While determining emission-factors, it is necessary to include monthly mean high and low temperatures based on a vehicle-population-weighted average rather than on a national geographical average. In scenarios analyzed under the present study, the monthly mean high and low temperatures for Georgia are taken from historical airport records.
  
• 
  Driving conditions. Different urban, rural and highway road speed conditions result in varying fuel consumption and emissions levels. Whenever possible, the percentage of vehicle-kilometers travelled for each type of vehicle under different road conditions should be determined by using a traditional four-step transport model for each year of the modeling period. Given the data availability limitations, the travel mixes for the present analysis were obtained from previous ESMAP research and are based on expert judgment.
  
• 
  Bio-fuel mix. The incorporation of bio-fuels into the petrol and diesel fuel mix changes the emissions characteristics of each vehicle. Additionally, under UNFCCC rules, the emissions from the combustion of biofuels are counted as agricultural emissions, not transport emissions, to avoid double-counting. Given the scarcity of data on vehicle fleet composition in Georgia by type of fuel used, bio-diesel blend share is considered in the current scenarios.
  
• 
  Inspection and maintenance. Effective vehicle inspection programs reduce fuel consumption and emissions, particularly from older vehicles, because of the improved maintenance. Therefore, it is necessary to take into account the average accumulated mileage of the vehicles in the population. For Georgia, the effects of improved inspection and maintenance programs have not been analyzed.
  

The data on new and used vehicle fixed costs (purchase prices) are based on publicly available market reference prices (www.autopapa.ge). The information on vehicle fuel prices (petrol and diesel) was provided by the Ministry of Energy of Georgia. The other operating costs are estimated using the empirical data from previous research conducted by ESMAP.

Figure 31: Winfrey S3 Survival Curves

Source: ESMAP