A bullet train arriving at Tokyo Station
Technical Summary

High-Speed Rail

Project Drawdown defines high-speed rail as: track construction for increased use of high-speed rail for intercity travel. This solution replaces other forms of travel, such as by airplane and by internal combustion engine (ICE) car.

High-speed rail (HSR) has experienced significant growth in recent decades, especially in China, whose HSR network grew to almost 30,000 kilometers in a few years, representing around two thirds of the global total. Growth is expected to continue as China and the European Union add to their extensive networks and new countries build their first links. This report seeks to quantify the greenhouse gas emission reductions and costs associated with increased HSR travel.

Project Drawdown takes the definition of the International Railway Union (UIC)[1] for HSR as a railway system where tracks are “…new lines designed for speeds above 250 km/hr and in some cases, upgraded existing lines for speeds up to 220km/hr.”[2] HSR’s most noted competitor—airplanes—emit more greenhouse gases per passenger-kilometer than HSR.[3] Thus, while HSR is a solution to multiple high-emitting modes of transportation, this study focuses on the replacement of air travel specifically.


Total Addressable Market[4]

The total addressable market for HSR is defined as the total non-urban travel projected by key sources like the International Energy Agency (IEA) and the International Council on Clean Transportation (ICCT). The current market share[5] of HSR is estimated from UIC and Amtrak data. The data shows that just over 3.2 percent of the global market is provided by HSR on the world’s existing HSR tracks.

Adoption Scenarios[6]

Projections of HSR adoption are based on announced plans for HSR track construction, which is chiefly done by governments. Using UIC data and other web searches, estimates were collected of total track length under construction and planned, along with estimated completion dates. Total track length estimates amount to 49,000 kilometers.

As HSR has very different travel densities, or usage rates, in different parts of the world, the travel densities for countries with existing HSR were collected, and these values were used to predict the passenger-kilometers for new tracks.[7] For countries without any existing tracks, travel densities of the most similar country or region with HSR tracks were used.

Impacts of increased adoption of high-speed rail from 2020-2050 were generated based on two growth scenarios, which were assessed in comparison to a Reference Scenario where total HSR travel remains fixed at the current levels (3.2 percent of the total market).

  • Scenario 1: The potential adoption was determined based on the total announced track construction with the expected completion dates, and the travel densities of each country with an additional 1 percent annual growth rate applied to HSR travel on existing tracks.
  • Scenario 2: The same inputs were used as in the Scenario 1 above, but a 2.5 percent annual growth in the travel density of each country was applied.

Emissions Model

For operating emissions, increased grid electricity use from HSR adoption was included, along with jet fuel emissions for the conventional alternative. Indirect emissions were also included for infrastructure operations.

Financial Model

First costs for HSR were considered as track construction costs. A value of US$27.2 million per track-kilometer was used,[8] corresponding to the global average.[9] For the conventional alternative, aviation, no first costs were entered since the marginal costs for new flight routes is not only very low (compared to that for HSR routes), but they are also quite hidden since it involves an airline doing business analysis on its existing services, and possibly negotiating with local airports, governments and providers. It could also mean, setting up a new office in a new airport destination, but if the new link is between two airports where the airline already has operations, then this cost would be zero. As a result of all these unknowns, the first costs for new airline links has been set to zero.

Operating costs for HSR and the conventional alternative include operations and maintenance costs of the infrastructure and energy for train motion (grid electricity), airports (ground operations) and aircraft (jet A/A1 fuel). For the train, these are normalized on a per passenger-km or track-km basis. For aviation, this was normalized on a per passenger-km or available seat-km basis.


The Scenario 1 estimates an emissions reduction of 1.3 gigatons of carbon dioxide-equivalent greenhouse gases over the 2020-2050 period at the end of which would exist 81,000 km of track or about 30,000 more than today. Most emissions reductions will come from Asia due to the extensive HSR growth planned in China. Globally, this is expected to result in an additional US$612 billion in construction costs, and should incur an additional US$849 billion in operating costs over the lifetime. In the Scenario 2, the avoided emissions could be as high as 3.8 gigatons.


Other benefits of HSR not calculated in this model include: reduced road congestion, fewer road accidents, lower local pollution and noise, less susceptibility to weather than aviation, easier and more comfortable travel. As indicated in our results, however, HSR is expensive to build, with higher expenses for higher speeds. Some research points to positive economic effects of HSR, but it is unclear whether these are actually generated by HSR or merely re-distributed to other regions.

A policy of encouraging more usage of existing HSR tracks does make sense, as that improves the solution’s climate and financial impacts, and as the marginal cost of adding more trains is very low. The EU has this policy, and we see many more countries following suit. We also expect increased travel on Chinese lines which already have excellent coverage of the country.

Often transport impact analyses use some sort of choice model, but this is not possible at the global level because of sparse data. Our approach is reasonable since it is grounded in firm construction announcements and existing travel. Due to the high quality of HSR travel, it tends to attract many air travelers on reasonable-length routes,[10] but it also attracts people who would have used cars or conventional rail, or who would not have traveled at all before HSR. We did not model this diversity of alternatives due to the inherent complexity in identifying and combining worldwide data.

The positive effects of HSR can be maximized by building HSR lines in travel corridors where the train will compete for travelers with higher-polluting travel modes such as airlines and automobiles, and where travel demand  is already high such as between larger cities in close proximity.


[1] UIC means “International Railway Union” in French (Union Internationale de Chemins–de-fer).

[2] 250 km/h = 155 mph, 220 km/h = 136 mph

[3] This is generally the case even when infrastructure emissions are included, but depends heavily on the number of people in the HSR train and in the aircraft. Emissions per passenger-kilometer for a full aircraft can be lower than that of an almost empty train. Therefore, increasing travel on existing train services can reduce unit emissions.

[4] For more on the Total Addressable Market for the Transport Sector, click the Sector Summary: Transport link below.

[5] Current adoption is defined as the amount of functional demand supplied by the solution in 2018. This study uses 2014 as the base year.

[6] For more on Project Drawdown’s growth scenarios, click the Scenarios link below. For information on Transport Sector-specific scenarios, click the Sector Summary: Transport link.

[7] This omits numerous factors that affect demand for new HSR lines, but as a first approximation at the global level, it is very useful.

[8] All monetary values are presented in US2014$.

[9] This number includes tracks in China, the USA, Japan, and Europe.

[10] The typical distance where HSR can capture a lot of the traveling market is where it can complete the journey in around 3 hours (at an average speed of 220 kilometers per hour, this is just under 660 kilometers, or 410 miles).