German Rail Traffic Surges

DB announced today that it had 500,000 riders across the two days of last weekend. This is a record weekend traffic; May is so far 5% above 2019 levels, representing full recovery from corona. This is especially notable because of Germany’s upcoming 9-euro ticket: as a measure to curb high fuel price from the Russian war in Ukraine, during the months of June, July, and August, Germany is both slashing fuel taxes by 0.30€/liter and instituting a national 9€/month public transport ticket valid not just in one’s city of domicile but everywhere. In practice, rail riders respond by planning domestic rail trips for the upcoming three months; intercity trains are not covered by the 9€ monthly pass, but city transit in destination cities is, so Berliners I know are planning to travel to other parts of Germany during the window when local and regional transit is free, displacing trips that might be undertaken in May.

This is excellent news, with just one problem: Germany has not invested in its rail network enough to deal with the surge in traffic. Current traffic is already reaching projections made in the 2010s for 2030, when most of the Deutschlandtakt is supposed to go into effect, with higher speed and higher capacity than the network has today. Travel websites are already warning of capacity crunches in the upcoming three months of effectively free regional travel (chaining regional trains between cities is possible and those are covered by the 9€ monthly pass). Investment in capacity is urgent.

Sadly, such investment is still lagging. Germany’s intercity rail network rarely builds complete high-speed lines between major cities. The longest all-high-speed connection is between Cologne and Frankfurt, 180 km apart. Longer connections always have significant slow sections: Hamburg-Hanover remains slow due to local NIMBY opposition to a high-speed line, Munich’s lines to both Ingolstadt and Augsburg are slow, Berlin’s line toward Leipzig is upgraded to 200 km/h but not to full high-speed standards.

Moreover, plans to build high-speed rail in Germany remain compromised in two ways. First, they still avoid building completely high-speed lines between major cities. For example, the line from Hanover to the Rhine-Ruhr is slow, leading to plans for a high-speed line between Hanover and Bielefeld, and potentially also from Bielefeld to Hamm; but Hamm is a city of 180,000 people at the eastern margin of the Ruhr, 30 km from Dortmund and 60 from Essen. And second, the design standards are often too slow as well – Hanover-Bielefeld, a distance that the newest Velaro Novo trains could cover in about 28 minutes, is planned to be 31, compromising the half-hourly and hourly connections in the D-Takt. Both of these compromises create a network that 15 years from now is planned to have substantially lower average speeds than those achieved by France 20 years ago and by Spain 10 years ago.

But this isn’t just speed, but also capacity. An incomplete high-speed rail network overloads the remaining shared sections. A complete one removes fast trains from the legacy network except in legacy rail terminals where there are many tracks and average speeds are never high anyway; Berlin, for example, has four north-south tracks feeding Hauptbahnhof with just six trains per hour per direction. In China, very high throughput of both passenger rail (more p-km per route-km than anywhere in Europe) and freight rail (more ton-km per route-km than the United States) through the removal of intercity trains from the legacy network to the high-speed one, whose lines are called passenger-dedicated lines.

So to deal with the traffic surge, Germany needs to make sure it invests in intercity rail capacity immediately. This means all of the following items:

  • Building all the currently discussed high-speed lines, like Frankfurt-Mannheim, Ulm-Augsburg (Stuttgart-Ulm is already under construction), and Hanover-Bielefeld.
  • Completing the network by building high-speed lines even where average speeds today are respectable, like Berlin-Halle/Leipzig and Munich-Ingolstadt, and making sure they are built as close to city center as possible, that is to Dortmund and not just Hamm, to Frankfurt and not just Hanau, etc.
  • Purchasing 300 km/h trains and not just 250 km/h ones; the trains cost more but the travel time reduction is noticeable and certain key connections work out for a higher-speed D-Takt only at 300, not 250.
  • Designing high-speed lines for the exclusive use of passenger trains, rather than mixed lines with gentler freight-friendly grades and more tunnels. Germany has far more high-speed tunneling than France, not because its geography is more rugged, but because it builds mixed lines.
  • Accelerating construction and reducing costs through removal of NIMBY veto points. Groups should have only two months to object, as in Spain; current practice is that groups have two months to say that they will object but do not need to say what the grounds for those objections are, and subsequently they have all the time they need to come up with excuses.


      • Project Do Better: tools for a kinder world

        Fair enough. So, my argument to my fellow Americans, that if we were to build sufficient community and empathy/cultural change, that the infrastructure upgrade is possible, does hold water. Thank you for this hope, even if it is a slim one.

        Belated Chag Sameach l’Shavuot, if you observe it,

  1. Matthew Hutton

    I’m not sure the NIMBY veto points are the be-all and end-all. I’m pretty sure to some degree that they’ve always existed – just that before the late 20th century they were un-transparent and not public. Oxford for example blocked the railway for longer than Cambridge.

    The difference however between modern projects and 19th and mid 20th century projects is that the older projects bought advantages to the people most affected by them. When the Great Western railway was built the rural towns and villages benefited with much closer connections to each other and London – even if they only got 1 slow train a day. When the motorways were built after world war 2 the rural towns and villages close to those roads gained a much faster connection to other towns and villages nearby and further afield.

    Even with the Tokaido Shinkansen the smaller places along the line benefit from the twice hourly all stations Kodoma service, and a fair few of the stops get the middle speed Hikari trains too. This means pretty much everyone gains at least something from the project.

    The difference with modern projects (such as HS2) is that the only clear winners at the big urban centres and everywhere else else appears lose. The thing is I think you can gain a lot by merely explaining how that isn’t true or adjusting the project so it isn’t true.

    For example I think with HS2 you could do a hell of a lot by merely explaining how the people affected by the construction benefit – which they do to at least a comparable extent to the bigger cities and making the (small) adjustments necessary for that to be true – such as having a good connection without long waits at Birmingham international for the Cross Country and classic west coast mainline trains and making sure Aylesbury is connected to East West rail as originally planned.

    Now sure that doesn’t get rid of NIMBY’s altogether – but it should make the community meetings less one-sided.

    • Eric2

      If you build an express bypass to an existing line, the small towns on the old line benefit by getting more frequent service, as the fast trains will switch to the new line.

      Though maybe the old and new lines are not in the same place, and maybe the small towns on the old line are car-dependent so they don’t care much about train frequency…

      • Matthew Hutton

        To give an example, the time from Birmingham international to Glasgow by road now at 9:30pm with almost no traffic is ~90 minutes slower than the indicative HS2 time.

        Even with slow public transport from one’s house to Birmingham international if an effort is made to ensure good public transport connections where you change it’d be tough for driving to beat that time from Aylesbury/Oxford/Banbury/Milton Keynes/Coventry to Glasgow.

        However you’ve still really got to show your workings, because if the connections were super poor and you were going to end up waiting 45 minutes for a train then it could still be slower than driving, even with the inevitable break at a service station.

    • Matthew Hutton

      Got an even older example of NIMBYism. Amsterdam wanted to build a Paris boulevard in the 17th century and the local residents rejected it so a wider canal was built instead so they could unload their good directly into their warehouses.

  2. Peter L

    I stopped reading after mention of the fuel tax reduction (elimination?). Germany and the rest of the EU wouldn’t be in such a quandary if they weren’t so dependent of Putin’s Petroleum. Reducing/eliminating the tax is not going to fix that … and robs the state of some of the money it needs to upgrade the rail network.

    Is the Bundestag now run by brainless American politicians? (I was going to say “Republicans” but the other party is nearly as stupid lately).

    • Eric2

      Yes obviously it’s not an ideal policy, but maybe it’s necessary politically, as the drivers who are paying for higher (I presume) fuel costs need to feel they are getting something out of the law too.

    • Alon Levy

      The fuel tax reduction is thanks to FDP, a party whose main program is that FDP should have power and be listened to. They thought just subsidizing public transport with the 9€ ticket was too hippie, so they tacked on the fuel tax reduction (but bear in mind, fuel prices here are on the order of 2€/liter right now, the tax reduction is from high levels). If you think they are stupid, you are correct and judging by recent election results, quite a lot of the German public agrees with this assessment.

      • Krist

        To be honest, I think the 9 euro ticket is a bad idea, that will backfire badly.
        If they really wanted to provide some temporary relief for the citizenry they should just have lowered the VAT. Public transport is infrastructure, not a social program.

        There are several reasons why this is a bad idea. Firstly if you really need to give away your product in order to find a buyer then your product is not very good. Improve your product. Do not dump it. Switzerland has high public transit prices and high public transit usage. The high prices make the high quality possible.

        Secondly a lot of people who never uses public transit before will now do so, will have an absolutely awful experience and will be turned off it forever.

        This is not how to promote public transit.

      • Oreg

        The FDP is frequently called the party of the wealthy. As the fuel tax cut is regressive (the rich drive bigger cars than the poor), it is a case in point. IIRC the FDP first pushed for the tax reduction and then the Greens wanted a transit subsidy in return, but I guess that doesn’t make a big difference. As a rule, most stupid policies of this coalition are indeed thanks to the FDP.

        • adirondacker12800

          Poor people don’t own cars. Lower middle class people, perhaps, but poor people don’t own cars.

          • Alon Levy

            Yes, FDP is aware. The three guiding principles of the party are ego, rich-people populism, and looking at where political winds blow. In six months they went from deficit hawks to borrow-and-spend Republican budgeting, when circumstances have changed in the exact opposite direction.

          • Oreg

            Half of those under the poverty line in Germany still own a car. In some countries, many homeless live in their cars.

            Those who can’t even afford a car bolster my point: They don’t get anything out of the fuel tax cut whereas the rich profit the most, making the cut regressive.

    • plaws0

      Let me also add, weeks later, that it’s just as stupid when Sleepy Joe Biden tries to get Congress to have a “fuel tax holiday”. The Federal Motor Fuel Excise tax has been US$0.184/US gal since :checks notes: 1993. Locally, gas prices can swing by more than that all by themselves in a week’s time.

      This will not be noticed by most gasoline consumers and Biden will get no credit (which, really, he shouldn’t).

      Stupid stupid stupid.

  3. adirondacker12800

    So to deal with the traffic surge, Germany needs to make sure it invests in intercity rail capacity immediately.

    Should they have Amazon deliver the new trains, overnight or surf AliExpress to see what’s available? Even if the nine euro deal becomes permanent people won’t be catching up on things they haven’t done in the past two years. They enact all of your plans when does something actually go into service? 2024? 25? They can’t do it all at once because there isn’t enough capacity to make all the stuff all at once or the labor to construct it.

    • Eric2

      You make the decisions immediately, so that there’s no unnecessary delay in when the tracks get built and the trains get delivered.

    • Alon Levy

      They should procure 300 km/h trains sometime this decade, whereas the coalition agreement specifically says no new >250 km/h rolling stock until the 2030s. It’s fortunate that Germany does have capacity right now for present traffic levels, just not for the traffic levels it will have in 10 years if traffic growth continues.

      • Max Wyss

        The decision to buy 330 km/h enabled trains right here right now, would get a fleet operational some time in 2028 (even if Bombardier is out of the game now).

    • Henry Miller

      You need to start now. Better would be 20 years ago, but now will do. Start building tracks, there are plenty of places to build, so keep building them, getting better all the time. The big train manufactures can make more trains – if they believe you will keep ordering enough that will make a new factory that can build them for less than current prices, but it takes a large investment to make a factory so it needs to pay off. The cost a train could be cut by 2/3rds if orders for trains go up 100x, (this won’t happen, but the principal sound: the more trains you buy the cheaper they are for everyone).

      You have to start now to make a better world in 20 years.

      • Phake Nick

        Taiwan is now in a situation where they are doing limited expansion of service to their system and need a grand total of four extra trains for the dedicated spec of Taiwan’s system – Got told by suppliers that the price tag will be an order of magnitude higher than what Japanese companies usually get their Shinkansen due to all the unoque design and construction of supply line – And get taken aback and try to look for alternative supplier that can give them what they want at cheaper price.

        • Henry Miller

          I realized after hitting post that I should have mentioned build something standard. Any train manufacturer can make you a standard gauge train, but if you want a different wheel gauge it might or might not be possible to do that. There are advantages to non-standard wheel gauges, but at least pick a gauge that someone else already has so that is is worth someone making a train to fit. Likewise there are a few common widths, pick one that is already common don’t make your own. Choose a door spacing that is common: there are a lot of choices, don’t make a new one. Don’t build a new system with anything other than 25kv overhead wire, though if you have a legacy system maybe we can work with what you have. You can get any color paint you want on your train. You can choose a few different styles of seats. Don’t customize anything else if you can help it, customization costs everyone money – not just you, but other systems that could benefit from you ordering the same train as them making the costs to build more expensive jigs worth it.

          Monorail has some interesting advantages on paper. However because you cannot buy a monorail train from anyone that will fit the costs are high. If you really want to use a monorail, then at least get a dozen other cities to commit to buying the same train as you: this will make it worthwhile to keep a standard monorail train in production. Same goes for maglev, and anything else that isn’t already a common standard: don’t build it alone, get a dozen others to build the same thing.

          • Phake Nick

            Taiwan use Japanese standard and European components because of both political and diplomatic issues, that is not something the construction of transportation itself can change. Japan’s side do intend to mitigate this issue in part by having same type of train exporting to Taiwan and Texas Central, but with the current progress of Texas Central it seems unlikely to match what Taiwan need.

          • Max Wyss

            The points shown are pretty much irrelevant. Rolling stock manufacturers have platforms which they adjust on demand. Gauge is not an issue at all (Velaros exist for standard and broad gauge, for example). Line voltage is absolutely no problem; you may almost have to pay extra if you want a single voltage train instead of a multisystem one…

            FWIW, the reason why the FLIRT is such a successful train is that it is a platform, and “customisation” is quite formalised (and the customer’s requirement do vary in any case).

          • adirondacker12800

            The polyphase motors are on the far side of inverter that takes DC and makes variable frequency, variable voltage AC. The software changes what is coming out it depending on demand. Where the DC comes from is likely something they have in their catalog because there are only a few standards. Grid frequency is good because then there isn’t a complicated conversion step from grid to whatever the train runs on.

          • Max Wyss

            A considerable amount of non-general grid frequency power actually comes from plants owned by the railways, thus directly creating 16.7 Hz, for example.

          • adirondacker12800

            The conversion from grid frequency to a different one isn’t 100 percent efficient. If you lose a few percent in that step it’s a significant if you are talking about megawatts. The equipment to do that isn’t cheap either.

    • adirondacker12800

      The deep discount on local travel lasts until September. What kind of trains are running on what kind of tracks in 2030 doesn’t make travel any better in August of this year.

        • adirondacker12800

          Silly me, they have a deep discount for the next three months. I thought “immediately” was something that could actually be done in the next three months. What are they going to do in the next three months?

      • Alon Levy

        Right, which is why I’m not complaining about a crunch now; I’m leading with record traffic before the deep discounts even started and talking about medium- and long-term planning, something sorely missing in the German government on most matters.

        • adirondacker12800

          “So to deal with the traffic surge, Germany needs to make sure it invests in intercity rail capacity immediately. ”
          When is the traffic surge happening? 2028? Are we going to have another global pandemic in 2026 and 2027 that makes everybody put off travel until 2028? If they make decisions in August of this year or February of next year no one is going to notice that it get completed in early 2033 instead of mid 2032. And the travel restrictions of 2027 won’t be affecting travel as much. And the government can change it’s mind in 2025 and nothing gets done.

          • Alon Levy

            The traffic surge is happening now (well, not literally now, but last month). And this isn’t a pandemic-induced surge, it’s just long-term growth in intercity rail traffic blowing past projections.

          • adirondacker12800

            I’m confused. People aren’t taking trips they couldn’t take last year, this year? And they aren’t taking advantage of the 9 Euro ticket? They are going to take the trip they didn’t take in 2021, again, in 2023? Plan everything you in your omniscient wisdom want to do, how much of it is done by August?
            Rational people don’t build capacity for once a century anomalies. They don’t build capacity, to and from resorts, for summer weekends. They don’t build capacity for Thanksgiving weekend in the U.S. or New Year in China. 2030 isn’t “immediately”. Whatever they are up to in 2030 doesn’t have to be decided next week. They can’t do it all by 2025 either because there isn’t enough capacity. And no matter how hard the click the heels of your ruby slippers, no matter what they do or when they do it, it’s not going to make travel better last month. I’m confused.

  4. danohu

    In the medium term, it feels Germany could make a lot of progress just by buying more rolling stock. I’ve rarely been anywhere in Germany that feels like it is squeezing the maximum out of track and stations.
    And that’s fortunate, because lack of trains can be fixed in years rather than decades. I assume this is part of the plan behind the 9-euro tickets: get non-habitual rail users onto the trains in their overcrowded state, to build political momentum behind making things better.


    Dear Alon,

    I have read your “Pedestrian Observations” blog for several years and am very impressed with the depth and insight of your comments found nowhere else.

    I assume you know this but if not, your orkext was quoted and discussed by noted NY Times columnist Ezra Klein:

    He cites the high subway transit and railroad capital costs in the US versus other developed countries due to US government obsession with processes and hearing every voice instead of outcomes. It came from the Transit Costs project.

    I am a NY MTA corporate management retiree of three years. I get to Berlin from time to time and live in NYC. Hope to meet you in person one of these days.

    Regards Bob

    Robert Newhouser Retired Deputy Director, Strategic Initiatives NYC Transit Also President, Electric Railroaders Association (

    Sent from my iPhone


    • Benjamin Turon

      That was a very good opinion piece, and its not just infrastructure and social programs, the Pentagon and NASA struggle to get things done, like building ships or moon rockets.

  6. Phake Nick

    I feel that, with full high speed rail, key corridors in and around Germany, like Amsterdam-RhineRuhr-Frankfurt-Munich-Vienna, Brussels-RhineRuhr-Hamburg-Berlin, Paris-Frankfurt-Leipzig-Berlin-Warsaw, Hamburg-Hanover-Frankfurt-Zurich-Milan, should be able to generate enough traffic to support Shinkansen level frequency where passengers arriving from anywhere can just hop on next train with minimal wait, and thus better speed them up as much as possible instead of rely on takt for connection.

    • Alon Levy

      Yes. However, unlike in Japan, this system would still branch a lot. So if you trust my model, there should be extremely high ridership between Cologne and Frankfurt, enough to fill 12+ tph, but they’d heavily branch at both ends, and probably many wouldn’t be serving both Cologne Hbf and Frankfurt Hbf but rather use Deutz or Frankfurt Airport to continue north-south.

      • Phake Nick

        I think 4 trains per hour is already enough to justify ignoring takt and pursue faster operation, as that mean 15 minutes headway and average wait time of ~7.5 minutes if passengers arrive at the platform randomly, and for long distance trip saving 7.5 minutes would be much easier than trying to keep a takt, especially with how most ICE service in Germany still need to share track with other trains and thus on time performance isn’t that great.

        Google search turned up that last year only 75% IC/ICE trains were on time, with this level of punctuality itvis effectively meaningless to time connection with accuracy down to minutes level. 75% is like, if a train arrive at a station and provide connection opportunity to each of every cardinal directions, then on average there would be one of the train being late for every arrived train. And apparently the a train is only late if it is late by more than 5 minutes. This make the matter even worse as takt being presented usually won’t have more than 5 minutes grace period for passengers to make connections.

        • df1982

          Yeah, Deutsche Bahn punctuality is trash. It doesn’t help that they often send their ICEs out on long circuitous routes (like Berlin to Munich via Frankfurt) that end up cascading delays.

          But you raise an important question about the value of a Taktfahrplan at different frequencies. It seems that 30-minute headways are the sweet spot for a Takt to be effective. Longer than that (hourly or two-hourly) and it’s more passenger unfriendly and wasteful of resources (since you have to hold trains longer at termini to make them fit into the Takt). And once you get to 10-minute headways there’s no point doing a Swiss-style Taktfahrplan since you’ll have an average 5-minute wait to interchange anyway. 15-20-minute headway seem to be the grey area. Is it still worth having timed interchanges at these frequencies? Or should you just run as fast as possible and hope the connections aren’t too bad?

          • Alon Levy

            At very high frequency, an internal Takt is valuable as a planning tool – capacity bottlenecks are unavoidable like difficult junctions or terminals and so it’s best to have a consistent timetable for them. The midday Shinkansen are fairly Takt-y, if not as regular as Swiss trains.

            What loses value is the Swiss run-as-fast-as-necessary maxim. Under this maxim, if you can’t get Stuttgart-Munich to around 55 minutes, it’s not worthwhile to get it below around 1:25, so current D-Takt plans are for 30 minutes minus turnaround for each of Stuttgart-Ulm, Ulm-Augsburg, Augsburg-Munich. But if frequencies are higher, then running at 300 km/h to do Stuttgart-Munich in around 1:05 is valuable, and this argues in favor of a Munich-Augsburg NBS. (But most NBSes that Germany isn’t building but should be do fit in the maxim anyway – Hanover-Dortmund is doable in just less than an hour at 300 km/h, and so is Erfurt-Frankfurt).

          • df1982

            Yes, I should have been more specific: having regularly scheduled trains is still beneficial (i.e. the same stopping pattern every 10 minutes throughout the day), but requiring trains to meet at nodes becomes less so, and if it evens slowing trains down from what they could be (either to meet the node, or through longer dwells to allow interchanging across multiple platforms to take place) it can even be counterproductive.

            Erfurt-Frankfurt has to be the most infuriating part of the present Deutsche Bahn network. The Berlin-Frankfurt travel market is huge, but trains end up meandering through a curvy, indirect route at sub-150km/h speeds for much of the journey. They need to just punch a 300km/h line straight through from Erfurt to Fulda, even if the terrain is difficult, but to the extent that there are plans for improvements on the line, they recreate the indirectness of the present route.

          • Phake Nick

            Midday Shinkansen timetable, especially at smaller stations, are just regular clockface schedule. They might have some feeder transit timed to match the Shinkansen arrival/departure time, but those are usually individually planned by local transit agency instead of any conscious efforts to provide timed connection, let alone designing the entire system based on such schedule.

    • Matthew Hutton

      The Tokaido Shinkansen goes between a Region of 20 million and a region of 35 million. I’m not sure many of those other places you name are quite that big 😀.

      • Luke

        There’s no Tokyo-size anchor, but then, unlike in Japan–whose only growing places are Tokyo and 5-hours-away-by-train Fukuoka, Europe’s population isn’t in absolute decline, and my general impression is that there’s still a certain amount of urbanization going on in most European cities. In any case, there are enough places well-enough connected to internal rail services that external (i.e., intercity) rail services will generate decent ridership for an appropriate level of service. Maybe more Kodama and Hikari than Nozomi, but still plenty enough to make the full investment worthwhile.

        • Matthew Hutton

          Yeah but for London for example I think it’s going to be more like a handful of trains a day like London-Lyon-Montpellier-Barcelona-Madrid or London-Brussels-Cologne-Frankfurt-Munich-Vienna or London-Lyon-Turin-Milan-Rome or London-Brussels-Dortmund-Hannover-Berlin trains rather than a train every 10 minutes that you get from Tokyo to Kansai.

          • Matthew Hutton

            Plus you have (for example) the channel tunnel where the high speed trains go through as a flighted pair. So perhaps you could run ~3 trains together through the tunnel every 30 minutes – assuming there is capacity at passport control for that.

  7. Pingback: Les bonnes intentions de l'Allemagne d'un laissez-passer de transport à 9 euros pourraient se retourner contre lui - Blog Voyage

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.