– Last Update August 1, 2023 –

The main obstacle on the road to sustainable aviation is the framing: aviation policy appears to be driven more by perceptions than by analysis. Which is a pity because a factual analysis is not very complicated. It requires no more than a few straightforward calculations. For starters, these quickly identify the options.


Making aviation sustainable is certainly a challenge. Trains can be electrically powered via overhead lines and cars via batteries, so both can be made sustainable easily. But batteries are not an option for airplanes. For the same amount of energy, they weigh forty times as much as kerosene and for longer flights the kerosene already takes up a quarter to sometimes even a third of the maximum take-off weight. You cannot increase that maximum weight either because that kills the high energy efficiency of flying.

In terms of weight, hydrogen is an option. It is one third the weight of kerosene for the same amount of energy. But volume is a problem: kerosene provides 36,000 kJ per liter, hydrogen only 10. Pressurizing helps, but even at 700 bar it’s still only 5,000 kJ per liter. Liquefaction is also possible, but then it must be cooled to below minus 253 oC and you still get only 8,500 kJ per liter, so you still need about four times the volume you need with kerosine.

Short distance

Anyway, for smaller planes and short distances, say up to 500 km, it is an option. Much development is still needed however and hydrogen-powered aircraft will not be available in meaningful numbers before 2040. That is not a problem, because the hydrogen must of course be made with emission-free electricity and that too will not be available in sufficient quantities before 2040 at the earliest.

To give an idea of the effort required, let’s look at the total energy consumption in Europe. In 2019 that was 41,000 PJ, of which 10,000 PJ was electricity, of which 13% from hydro power, 16% solar and wind, plus 26% from nuclear power plants. All in all, 55% emission-free. Which, incidentally, means that for the time being the electric car and the train still run on fossil fuel for 45%, but let’s disregard that. Electric cars are probably still a good investment for an electricity-centered future.

This means that, if we assume that most possibilities for hydro power are in use and that the German policy of closing down nuclear power plants will become the European policy, the amount of energy from solar and wind must increase by a factor of five to six to make all electricity emission-free, and then by a factor of five to six again to be able to satisfy the remaining need for the energy with electricity too, and, when that is not possible for certain purposes, with hydrogen. In total, therefore, at least 25 to 35times as much capacity is needed as is now installed.

Four times 10,000 equals 40,000 of course, but quite a lot of hydrogen will be needed, and the efficiency of hydrogen production is about 70 to 80%, therefor the factor five at least. Nuclear energy will probably be unavoidable to meet this requirement and building nuclear power stations can easily take fifteen to twenty years. All in all, 2040 might even be a bit optimistic for all that to happen, so no need to worry about the time needed for building hydrogen-powered aircraft.

Long distance

For longer distances, some form of kerosene remains the only option, but that can also be electricity-based: synthetic kerosene, also known as e-kerosene. This can be produced with hydrogen and CO2 captured from the atmosphere. But that process too needs to use emission-free electricity and that too will therefore only be possible in meaningful amounts after 2040 at the earliest.

The other option is bio-kerosene. This is already possible now and of course biomass must be grown without displacing food production or forests. For example, by using algae or elephant grass (Miscanthus). The problem here is the cost. Bio-kerosene is still too expensive, partly due to its small scale. A chicken or egg problem. It is not bought because it is too expensive and because it is not bought due to lack of scale production remains too expensive. Policy could break that vicious circle. But a policy to replace aircraft with high-speed rail on all flights shorter than 500 km will certainly not do that. It could even backfire when it comes to climate change.

Current policy

The second step in the analysis, to see whether current policy promotes sustainability, requires some more calculations. Much of this has already been done by the European Court of Auditors (ECA). They say:

“High-speed rail has only a limited competitive edge. (…) [H]igh-speed rail in Europe is generally competitive over travel distances of between 200 and 500 km, with journeys lasting up to four hours.”

Page 49

This is in line with the policy of replacing all flights within 500 km with HSR. This changes with the second finding, because that one calls into question the feasibility of the ambition:

“The Commission’s target of tripling the length of the high-speed rail network (reaching more than 30,000 km in 2030) is not supported by credible analysis. We consider it unlikely that this target will be reached, because it takes around 16 years for high-speed rail infrastructure to be planned, built, and to begin operations.”

Page 57

But the ECA continues and questions cost efficiency too:

“Judging by a benchmark, a high-speed line should ideally have nine million passengers per year to be successful. (…) High-speed rail infrastructure is expensive, and is becoming more so: on average, the lines we audited cost 25 million euros per km (not taking into account the more expensive tunneling projects).”

Page 8

It is not clear why more expensive tunnels are not taken into account, but this does make the average costs appear to be a bit on the low side. The cost of HSR South in the Netherlands was 64 million euros per kilometer. Nevertheless, even with 25 million euros per kilometer, the costs for the 20,000 km of HSR track to be built still amount to a respectable 500 billion euros. With depreciation over 40 years, that is 12.5 billion per year.

As a result of this policy, up to a quarter of the European flights could be replaced and that sounds impressive. But because these are short flights only, that saves at most 7.5 Mt CO2. That is less than 4% of the total emissions from European aviation. And less than 0.13% of total emissions in the EU. Moreover, construction and maintenance of the track also causes CO2-emissions: more than 2.5 Mt per year.

On balance, the theoretical maximum annual savings would be less than 5 Mt. This makes the costs per avoided ton to over 5000 euros. Heat pumps cost 400 euros per ton avoided, offshore wind costs 100 euros per ton avoided and nuclear energy 20 euros. The ECA noticed this difference too. In the summary of the report, they put it mildly:

“The decision to build high-speed lines is often based on political considerations, and cost-benefit analyzes are not used generally as a tool to support cost-efficient decision-making.”

Page 8

Effective policy

The fact that politicians in Europe are willing to pay so much for a ton of aviation CO2-avoided is great though. Because that means there is more than enough money available for an effective policy.

To start with, we need to get rid of the frame ‘airplanes are bad/trains are good’. Both modes of transport are necessary, and they have their own place in the transport mix. And that is where they are most effective, also regarding emissions. You can easily calculate what that place is and most of the calculations have already been done, by technical consultancy Royal HaskoningDHV. Although their report concludes that the train is almost always a good alternative for short flights, this is mainly due to two interesting omissions. The researchers themselves identified these and they can be easily repaired.

Traffic volume

The first omission is that DHV does calculate how many travelers will use a connection, but does not elaborate on what that means:

“Berlin case: Not part of this case is whether the proposed improvement of the timetable is justified by sufficient transport demand and transport volume. This should be the subject of further research.”

Page 32, translation by me

This is somewhat strange, because the traffic volume has a decisive influence on costs as well as on emissions per passenger. An explanation could be that the organizations that commissioned the report preferred not to hear some of the results. It is perhaps understandable that as a consultant you then still mention these omissions, even if only to cover yourself.


That would also explain the second omission. DHV does not consider emissions from construction and maintenance of the railway infrastructure. It is only mentioned once in a side note:

“In addition, the construction of infrastructure is quite CO2-intensive. Based on figures from the International Union of Railways, (UIC, 2016), an average CO2-emission of 50 tons CO2/km/year has been assumed, including operations. A study into the construction of HighSpeed2 between London and Birmingham (Greengauge21, 2010), shows an average emission of 5,200 tons of CO2/km.”

Page 27, Translation by me

If you check this at the source, UIC, you will see that the 50 tons apply to normal track. For HSR, the UIC gives over 250 tons/km/year. And if you check the DHV example of Birmingham, you will arrive at 130 tons/km/year with a depreciation of 40 years. And if you consult the original source, and not the action group Greengauge21, you will see that 1.2 Mt total emissions are most likely, which for that stretch of 176 km* comes down to 6,818 tons per km and therefore 170 tons/km/year. But that is a mere detail. We calculated with the lowest value, 130 tons/km/year, for two of the routes DHV recommended.

Case Dusseldorf

Distance 240 km, so costs 6 billion, is 150 million/year. With four return flights per day and eighty passengers per flight, this amounts to just under 240,000 passengers per year. This brings the costs to 625 euros per passenger per one-way trip and the emissions due to the infrastructure to 540 gr/pkm (gr/pkm = grams per passenger per kilometer). Incidentally, DHV estimates the passenger potential to be 800,000:

“This is the full number of potential passengers for the Schiphol market, including air passengers and rail passengers.”

Page 4, translation by me

If we accept this, the costs will still come to 185 euros each way and the emissions to 160 gr/pkm. Just for the infrastructure. Note: the emissions for airplane passengers for that distance are 118 gr/pkm.

Case Berlin

This case is particularly interesting because Hanover and Berlin are already connected by HSR track. So, only the stretch from Amsterdam to Hanover, about 400 km, needs to be added. DHV puts the potential for Berlin at 700,000 passengers, but you can of course add the potential for Hanover, 210,000. Let’s round it up to 1 million passengers per year. Then the calculation looks like this:

Route 400 km, costs 10 billion, makes 250 million per year. Costs 250 euros each way, emissions 130 gr/pkm. You then must add the costs and emissions of the Hanover – Berlin route for passengers going to Berlin, but the picture is sufficiently clear. For many routes, HSR will at high cost perform worse than an aircraft due to emissions from the infrastructure. In other words: as a general policy, always opting for HSR rail is not a very good idea. Of course, when HSR is the better option you should build it and then it is also a wonderful way of transportation.

Policy option

When it comes to sustainability, stimulating the development of the production of bio-kerosene is much more promising and much cheaper, and it can also go much faster than constructing HSR track. The focus must be on the network companies that form the European core of the three major alliances. So British Airways for Oneworld, Lufthansa for Star Alliance and KLM/Air France for Skyteam. After all, by far the most emissions come from flights to destinations outside Europe, which are mainly served by these networks. It therefore seems sensible to make them responsible for the development of bio-kerosene, including building the stable long-term relationship that offers producers of biomass and bio-refineries the necessary security.


The implementation could look like this:

  1. Levy all intra-European airline tickets with a sustainability surcharge of 20 euros.
  2. Divide the proceeds among the network companies mentioned, with the condition that a maximum of half of that money may be used for research and development. The remainder must be used for the purchase of bio-kerosene that does not displace food production or forests.
  3. This bio-kerosene can be burned in any aircraft but is allocated to the shortest flight route without an HSR connection. For KLM that would be Amsterdam-Düsseldorf.
  4. Give a bonus of 1 euro per kg from EU funds on the total annual consumption on that route as soon as that route runs entirely on the allocated bio-fuel. This bonus is initially awarded every year and may only be used for the purchase of bio-kerosene.
  5. The surplus of bio-fuel is allocated to the second shortest route and the process repeats.
  6. When all flights in Europe shorter than 500 km run on bio-fuel, whether allocated or not, assess whether the policy is sufficiently stimulating or whether it should be expanded to flights up to 750 km. Or maybe it could be abolished.

This approach would significantly promote the development of bio-kerosene due to the emerging market, which can therefore grow rapidly. In the Netherlands, for example, a collaboration could arise between Wageningen University for the development and improvement of feed stocks and production methods, and farmers all over Europe for the actual growth.

And that at a relatively low cost. Achieving the saving of 7.5 Mt CO2 requires a bit less than 2.5 billion kg of bio-kerosene. The cost for the EU would then be 2.5 billion euros per year, at 316 euros per ton of emissions avoided. That is less than the cost per ton avoided of the much-praised heat pumps.

Of the apparently available 12.5 billion per year, 10 billion would remain. You could invest that in improvements to the normal track, which the ECA also advocates, and/or in the construction of nuclear power stations. With a cost of 20 euros per ton of CO2-emissions avoided, that might not be a bad idea either.


This analysis shows once again that framing aviation as the Big Bad Wolf can have major harmful consequences, as it can easily lead to policies based on that framing. To reiterate the statement of the ECA, this time as it reads in the main body of the report:

“The quality of the assessment of needs in the Member States is low. Alternative solutions, such as upgrading existing conventional lines instead of building new high-speed lines, are only considered systematically in Italy and Germany; this is a good practice which should be followed universally. Decisions to build are national and political; they are rarely based on proper cost-benefit analyses.”

Page 58

* September 2, 2021 – A 2020 update of the project gives 225 km for the length of the route, which brings emissions to 5,333 tons/km. So about 130 tons/km/year after all. However, the costs appear to have more than doubled: from £17 billion to £40 billion. Mainly because more tunnels were required than assumed in the first planning. Emissions/km will therefore be approximately double as well, to about the 250 tons/km/year already mentioned by the UIC.