Quantifying emissions from Railways (including LRT/MRT) is really tricky. Some of the variables which often trouble analysts are – use of construction, technology and occupancy factors which can make or break an analysis. Last year, Mikhail Chester brought forward a very interesting analysis on complete carbon footprint of transport modes and this study was critically accepted. The study provided a comprehensive environmental life-cycle assessment of not only vehicle and fuel components but also infrastructure components for automobiles, buses, commuter rail systems and aircraft. Many processes were included for vehicles (manufacturing, active operation, inactive operation, maintenance, insurance), infrastructure (construction, operation, maintenance, parking, insurance), and fuels (production, distribution). The vehicles inventoried in the study were sedans, pickups, SUVs, urban diesel buses, light rail, heavy rail and aircraft.
The important argument made by Chester was that one needs to closely look at the occupancy of Rails and its built infrastructure which often tip the scales. But, more often researchers think that one can always borrow the emission factors from different sources and this would provide some estimates of reasonable accuracy. This is a myth.
In order to demolish the argument of usage of common emission factor, we summarize many of the emission factors available online. The data collation was further helped by inputs from ADB-TA - Reducing Carbon Emissions from Transport Projects.
It is to be noted that emission factors have been quantified using different methodologies with different boundaries. What sets them apart is the huge variation. The variation is between 16 to 1200 g/pkm. The Asian MRT’s which have very high occupancy ratios have values between 20 to 110 g/pkm. Segregating heavy rails, MRT and LRT may help in refining this further. However, it is to be noted that emission factors cannot be constant but dynamic with time in order to reflect changes in design, occupancy and other factors.
Thus, one cannot borrow the emission factor straight away. What one should do is to measure the fuel/electricity consumption to derive emissions. There is no easy way out.
 Please send us a request in case you would like to access the sources.