M-DEV0448: Determination of parameters post project start instead of prior start

AM0031: Methodology for Bus Rapid Transit Projects

According to the methodology AM00031, version 3.1, several parameters shall be determinate ex-ante before the project start.
The start of operation of the project activity “MEGABUS, Pereira, Colombia” was on 21 August 2006.
The following parameters, which according to AM0031, version 3.1 shall be determined ex-ante before the project start, have been determined based on a survey conducted in March 2007 (Megabus, 2007) and thus only within 8 months of start of operation of the project in August 2006.
1) SEC(T,G/T,D) - Specific gasoline / diesel consumption of taxis
2) TDi - Trip distance of people which would have used cars, taxis or motorcycles
3) OCi - Average occupancy rate baseline of vehicle categories cars, taxies and motorcycles
4) SRS - Share of road space
5) VBL/VP - Vehicle speed baseline and project

Moreover, the following parameters, which according to AM0031, version 3.1 shall be determined ex-ante before the project start, were determined based on data for 2006 published by Instituto de transito y transporte pereira, AMCO, in January 2007. Data for 2006 is thus partially prior project start and partially post project start.
6) VDT - Annual actual distance driven taxi
7) TRC - Daily trips made by passenger cars
8) RSB - Road space available baseline

The data for above listed parameters was thus not determined prior (or fully prior) to the project start as required by AM0031. The project participants were not able to perform the studies required prior to the project start due to the short time between the approval of version 1 of AM0031 (27 July 2006) and the project start date of the project activity (21 August 2006).

Data for determining parameters 1-5) could in contrast to all other data not be derived from feasibility reports or other studies made for the project, but had to be collected specifically for the CDM project.
The deviation is in DNV’s opinion justified as discussed below for each of the parameters 1) to 8) listed above.

1) SEC(T,G/T,D) - Specific gasoline / diesel consumption of taxis
Taxis specific gasoline / diesel consumption values used are from early 2007 and based on a vehicle mix circulating. No significant amount of vehicles was replaced within 8 months of the project start, so that the survey in March 2007 can be considered representative for the situation prior to the project start. The technical improvement factor is applied per year as prescribed by the methodology. The specific gasoline / diesel consumption of taxis and other vehicles in general is also not related in any manner with the existence or not of a Bus Rapid Transit (BRT) system (note that only the specific gasoline / diesel consumption for taxis is derived based on date after the project date while all other specific fuel consumption values had been established prior project start, especially the most important value of buses as this parameter has a major impact on emission reductions).

2) TDi - Trip distance of people which would have used cars, taxis or motorcycles
This data is monitored and therefore usage of the data ex-ante is only for projections. The project can thus not gain additional emission reductions through using a value after project start. As the TD refers in fact to the trip distance realized using the Megabus BRT the most precise projection can be made if the survey idem to the monitoring survey is conducted on the running BRT. Therefore, it is adequate and considered even better to use this data than any other projection based on trip distance which is not necessarily identical to the ones users of the BRT would make. The same approach was also used by other registered CDM projects, also using AM0031: BRT Zhegzhou (project 4744) and BRT Chongqing (project 3760).

3) OCi - Average occupancy rate baseline of vehicle categories cars, taxies and motorcycles
Occupancy rate changes of vehicles do not happen quickly and hence data collected within 8 months of project start is considered representative for the situation prior to the project start. Also as shown in Attachment 2 the limited outreach of the project is not expected to have a measurable impact on occupation rates due to the limited amount of mode switch versus the large number of trips made per mode i.e. the impact of the project disappears in the statistical noise. Bus occupation rate which is the most important parameter is established prior to project start based on available information.

4) SRS - Share of road space
The roads on which measurements were made in the survey are high volume traffic roads where the BRT does not operate. Thus no change of the SRS is expected less even in the short run. Moreover, taking the alternative formulae (equation 17 of AM0031, version 3.1) SRS would be lower and thus leakage congestion would also be lower. It can thus be concluded that the data used is conservative for calculation of leakage emissions.

5) VBL/VP - Vehicle speed baseline and project
The roads selected in the survey are not in the area of the trunk roads and thus are not roads affected by operations of the Megabus BRT. The speed monitoring made is to establish a correlation between speed and vehicle units. Thereafter based on this correlation the calculation for the impact of the Megabus BRT is made by projecting the number of reduced vehicles and therefore the expected speed change. As the roads selected are not affected directly by Megabus BRT (no trunk roads in that vicinity) the speed-number of units correlation is not affected by the operation of the Megabus BRT. Therefore, it is indifferent if this data is prior or post project operations.

6) VDT - Annual actual distance driven taxi
The parameter VDT is only required for leakage calculation of occupation rate taxis, i.e. if a change of the occupation rate of taxis occurs then VDT is used to calculate the leakage emissions. Even in case VDT would be affected by the BRT the leakage calculation would in fact be correct using the latest available data as actual leakage occurs only for actual distance driven (no emissions and thus leakage can occur for non-driven distances). For simplification purposes AM0031 does not require this parameter to be monitored but to be determined ex-ante. This is however only a simplification whilst technically correct would be to use actual distance driven of that year. This simplification was based on the fact that the BRT has no incidence on annual taxi distance driven.
Compared to other Colombian cities the value of 90 000 km is also high. Bogota has 57 000 km (see registered PDD 0672), Barranquilla 72 000 (PDD in registration), Medellin 63 000 (PDD in registration) and Cali 64 000 (PDD in validation). A higher value is more conservative (more leakage).

7) TRC - Daily trips made by passenger cars
The data for TRC is used to calculate rebound leakage (equation 18 of AM0031, version 3.1). The higher the value of TRC the higher the leakage. As TRC is based on the number of registered cars and as the car population is growing the calculated leakage emissions will be higher if a later data year is taken, which is conservative.

8) RSB - Road space available baseline
The road space will not change significantly in five months (i.e. August to December 2006). Hence, to apply data for the year 2006 is considered representative for the situation prior to the project start even though the BRT was in operation during the last five months of 2006, the year the selected data is based on.

The deviation is due to the fact that the project participants were not able to determine several parameters prior to the project start date, as required by AM0031. This deviation is project-specific and does not deviate from the methodology to the extent that a revision of the methodology would be required.

1) SEC(T,G/T,D) - Specific gasoline / diesel consumption of taxis
The usage of a value from the year 2007 instead of the year 2006 is justified and conservative due to the following:
a) The specific fuel consumption of vehicles does not change within months due to fleet renewal. If 10% of the fleet are renewed per annum then the impact of taking data one year later is only for 10% of the vehicles and assuming a 1% improvement rate per annum (as assumed per default by the methodology) the actual difference between data values will be 1%.
b) The methodology states an annual default improvement factor of 1% annually. Taking thus later data thus means that vehicles are newer and more efficient and thus the value will be lower and more conservative.
c) The project itself has no relation to fuel consumption values of vehicles. At best, but even this effect is only marginal, it can slightly reduce the fuel consumption through less congestion as projected by the project (see positive speed impact) and therefore SEC might drop marginally.
b) and c) indicate that using data of 2007 is more conservative than using data prior project start in the year 2006. Slightly lower SFC reduces baseline emissions (see equation 1 and 4 of AM0031, version 3.1).
A recent publication of the GEF (Manual For Calculating Greenhouse Gas Benefits For Global Environment Facility Transportation Projects, 2010, table 3) also has a default SFC for diesel taxis of 9.1 l/100km (project 8.0) and of 12.5 l/100km for gasoline taxis (project 8.5). Both values are higher than those measured by the project showing the conservative nature of the data used.
SEC is also included in determining leakage due to change of occupation rate of taxis. This leakage is monitored through the occupation rate and refers to the leakage of the monitoring year y. Therefore the transport emissions factor per distance EF(KM) for taxis also needs to be of this year. As the EF is fixed prior project start the EF is updated based on the technology improvement factor of 1% per annum. It is thus more exact to use a more recent than an elder data as this reflects better the actual EF in that year. Therefore, there is no underestimation of leakage through taking a later data.

2) TDi - Trip distance of people which would have used cars, taxis or motorcycles
This data is monitored and therefore usage of the data ex-ante is only for projections. If during the monitoring period the actual value is lower than the baseline value, the monitored value and not the baseline value is taken. Therefore the baseline value has only an influence on projections not however on actual emission reductions.

3) OCi - Average occupancy rate baseline of vehicle categories cars, taxies and motorcycles
The test of the potential impact of the project performed by the project participant (refer to Attachment 2) demonstrates that theoretical possible variations are inside the normal fluctuation rate or can statistically not be declared as a change outside random variations at a 95% confidence level. Hence, the usage of data from 2007 is idem to usage of data from 2006 and thus does have any significant impact on the project’s emission reductions.

4) SRS - Share of road space
The roads on which measurements were made in the survey are high volume traffic roads where the BRT does not operate. Thus no change of the SRS is expected less even in the short run. Moreover, taking the alternative equation for determining leakage due to changes in SRS (equation 17 of AM0031, version 3.1) SRS would be lower and thus leakage congestion would also be lower (refer to Attachment 3). It can thus be concluded that the data used is conservative for calculation of leakage emissions.

5) VBL/VP - Vehicle speed baseline and project
The data is used to establish a speed correlation and calculate the speed effect. The data was taken from roads not affected by the project and therefore the correlation would not be affected by operations. The correlation is thereafter used together with the pre-project data to establish the speed effect of the project. Equation 19 of AM0031, version 3.1, is based on speed differentials and thus absolute speeds of specific roads are irrelevant.

6) VDT - Annual actual distance driven taxi
The parameter VDT is only required for leakage calculation of occupation rate taxis, i.e. if a change of the occupation rate of taxis occurs then VDT is used to calculate the leakage emissions. Even in case VDT would be affected by the BRT the leakage calculation would in fact be correct using the latest available data as actual leakage occurs only for actual distance driven (no emissions and thus leakage can occur for non-driven distances). For simplification purposes AM0031 does not require this parameter to be monitored but to be determined ex-ante. This is however only a simplification whilst technically correct would be to use actual distance driven of that year. This simplification was based on the fact that the BRT has no incidence on annual taxi distance driven.
Compared to other Colombian cities the value of 90 000 km is also high. Bogota has 57 000 km (see registered PDD 0672), Barranquilla 72 000 (PDD in registration), Medellin 63 000 (PDD in registration) and Cali 64 000 (PDD in validation). A higher value is more conservative (more leakage).

7) TRC - Daily trips made by passenger cars
The data for TRC is used to calculate rebound leakage (equation 18 of AM0031). The higher the value of TRC the higher the leakage. As TRC is based on the number of registered cars and as the car population is growing the calculated leakage emissions will be higher if a later data year is taken, which is conservative.

8) RSB - Road space available baseline
The road space will not change significantly in five months (i.e. August to December 2006). Hence, to apply data for the year 2006 is considered representative for the situation prior to the project start even though the BRT was in operation during the last five months of the months of 2006 where the BRT was in operation.