Why You Should Complete High-Speed Lines

Some countries build complete high-speed rail networks, on which one can travel between cities almost entirely at high speed, such as France, Japan, and China. Others build partial networks, mixing low- and high-speed travel, such as Germany. The planning lingo in the latter is “strategic bypass” or “strategic connection.” And yet, there is nothing strategic about most mixed lines. If a line between two cities is partly high-speed and partly low-speed, it is usually strategic to complete the high-speed line and provide fast travel – the benefits will exceed those of having built the original high-speed partial segment. Since Germany’s rail network largely consists of such mixed lines, the benefits of transitioning to full high-speed rail here are large.

The arguments I’m about to present are not entirely new. To some extent, I discussed an analog years ago when arguing that in the presence of a complete high-speed line, the benefits of building further extensions are large; this post is a generalization of what I wrote in 2013. Then, a few months ago, I blogged about positive and negative interactions. I didn’t discuss high-speed rail, but the effect of travel time on ridership is such that different segments of the same line positively interact.

The upshot is that once the basics of a high-speed rail networks are in place, the benefit-cost ratio of further extensions is high. In a country with no such network, the first line or segments may look daunting, such as India or the UK, but once it’s there, the economics of the rest tend to fall into place. It takes a while for returns to diminish below the point of economic viability.

A toy model

Take a low-speed rail line:

Now build a high-speed line parallel to half of it and connect it with the remaining half:

You will have reduced trip time from 4 hours to 3 hours. This has substantial benefits in ridership and convenience. But then you can go all the way and make the entire line fast:

Are there diminishing returns?


The benefits of reducing travel time per unit of absolute amount of time saved always increase in speed; they never decrease. The gravity model holds that ridership follows an inverse square law in total cost, including ticket fare and the passengers’ value of time, which time includes access and egress time. Reducing in-vehicle travel time by a fixed amount, say an hour, increases ridership more if the initial travel time is already lower.

This is on top of reductions in operating costs coming from higher speed. Trains on high-speed track consume less electricity than on legacy track, because they cruise at a constant speed, and because head-end power demand scales with time rather than distance traveled. Crew wages per kilometer are lower on faster trains. And the cost of rolling stock procurement and maintenance is spread across a longer distance if the same train is run more kilometers per year. In the toy model, there are actually increasing returns coming from rolling stock costs: upgrading half the line to high speed requires running an expensive high-speed train on the entire line, whereas completing the high-speed line does not require increasing the cost per unit of rolling stock.

Diminishing returns do occur, but only in the context of an increase in top speed, not in that of speeding up slow segments to match the top speed of faster segments. In that context, benefits do diminish and costs do rise, but that is not the same as completing high-speed lines.

As the maximum speed is increased from 160 to 200 km/h, the train speeds up from 22.5 seconds per kilometer to 18. To provide the same increase further, that is to reduce the time taken to traverse a kilometer by a further 4.5 seconds to 13.5, the speed must increase to 266.67 km/h. To provide the same 4.5-second increase once more, the speed must increase to 400. Curve radius is proportional to the square of speed, so these increases in speed must be accompanied by much more exacting track geometry. Tunnels may well be unavoidable at the higher speeds in topography that could accommodate 200-250 entirely at-grade.

What’s more, operating costs rise too as top speed increases. The electricity consumption on a 300 km/h cruise is lower than on a legacy line on which trains transition back and forth between 200 and 100 and all speeds in between, but the electricity consumption on a 350 km/h cruise is definitely higher than on a 250 km/h cruise.

However, what is relevant to the decision of what standards to build a line to is not relevant to the decision of how far to extend this standard. Once a 300 km/h segment has been built, with a dedicated fleet of trains that cost 30 million per 200-meter set, the returns to upgrading the entire segment the train runs on are higher than those of just building the initial segment.

Can some strategic segments be easier to build than others?

Yes, but only in one specific situation: that of an urban area. The toy model says nothing of construction costs – in effect, it assumes the cost of making the first 200 km fast is the same as that of making the next 200 km fast. In reality, different areas may have different construction challenges, making some parts easier to build than others.

However, if the construction challenge is mountainous topography, then the higher cost of mountain tunnels balance out the greater benefit of fast trains across mountains. The reason is that in practice, legacy rail lines are faster in flat terrain than in the mountains, where past construction compromises led to sharp curves.

This situation is different in urban areas. In urban areas as in the mountains, costs are higher – land acquisition is difficult, and tunnels may be required in areas where the alternative is buying out entire city blocks. But unlike in the mountains, the existing rail line may well be reasonably straight, permitting average speeds in the 120 km/h area rather than the 70 km/h area. In that case, it may be advisable to postpone construction until later, or even keep the legacy alignment.

One example is the Ruhr area. The tracks between Dortmund and Duisburg are not high-speed rail – the fastest trains do the trip in about 34 minutes, an average speed of about 95 km/h. Speeding them up by a few minutes is feasible, but going much below 30 minutes is not. Thus, even if there is a 300 km/h line from Dortmund to points east, the returns to the same speedup between Dortmund and Duisburg are low. (Besides which, Dortmund is the largest city in the Ruhr, and the second largest, Essen, in the middle between Dortmund and Duisburg.)

Another is Connecticut. East of New Haven, there is relatively little urban development, and constructing a 300-360 km/h line roughly along the right-of-way of I-95 poses few challenges. West of New Haven, such construction would require extensive tunneling and elevated construction – and the legacy line is actually somewhat less curvy, it’s just slower because of poor timetable coordination between Amtrak’s intercity trains and Metro-North’s regional trains. While the returns to building 250-300 km/h bypasses around the line’s slowest points in southwestern Connecticut remain high enough to justify the project, they’re lower than those in southeastern Connecticut.

The situation in Germany

On the following map, black denotes legacy lines and red denotes purpose-built 300 km/h high-speed lines:

The longer red segment, through Erfurt, is the more challenging one, including long tunnels through the mountains between Thuringia and Bavaria. The complexity and cost of construction led to extensive media controversy. In particular, the choice of the route through Erfurt came about due to Thuringia’s demands that it serve its capital rather than smaller cities; DB’s preference would have been to build a more direct Leipzig-Nuremberg route, which would have had shorter tunnels as the mountains in eastern Thuringia are lower and thinner.

Since then, a lot of water has passed under the bridge. The route opened at the end of 2017 and cut travel time from 6 hours to 4, bypassing the slowest mountain segment, and is considered a success now. In the North German Plain, the trains mostly cruise at 200 km/h, and trains traverse the 163.6 km between Berlin and Halle in 1:09-1:11, an average speed of 140 km/h.

Nonetheless, the benefits of painting the entire map red, roughly from the city limits of Berlin to those of Munich, are considerable. The North German Plain’s flat topography enables trains to average 140 km/h, but also means that building a high-speed line would be cheap – around 137 km of new-build line would be needed, all at-grade, at a cost of about €2.5 billion, which would cut about half an hour from the trip time. In Bavaria, the topography is rougher and consequently the legacy trains’ average speed is lower, but nonetheless, high-speed rail can be built with cut-and-fill, using 4% grades as on the Cologne-Frankfurt line.

I’m uncertain about the exact travel time benefits of such a high-speed line. I put a route through my train performance calculator and got about 2.5 hours with intermediate stops at Südkreuz, Erfurt, Nuremberg, and possibly Ingolstadt (skipping Ingolstadt saves 3 minutes plus the dwell time), using the performance characteristics of the next-generation Velaro. But I’m worried that my speed zones are too aggressive and that the schedule should perhaps accommodate TGVs coming from Paris via Frankfurt, so I won’t commit to 2:30; however, 2:45-2:50 should be doable, even with some unforeseen political compromises.

But even with less optimistic assumptions about trip times, Germany should do it. If it was justifiable to spend €10 billion on reducing trip times from 6 hours to just under 4, it should be justifiable to spend around half that amount on reducing trip times by another hour and change.


  1. Alon Levy

    I’m deliberately focusing on a single line in this post and not on a network. In the context of a branched network, there can arise situations in which a mixed line is best, same as with subway-surface rail lines within cities. For example, South Korea today has fully high-speed lines connecting all its major cities, but its first line opened from Seoul to Daegu and had a branch going from Daejeon southwest to Jeolla in addition to the mainline going southeast to Busan.

    However, in the specific case of Berlin-Munich, this isn’t how the line has been prioritized. The line plausibly has branches from both ends to Frankfurt, and Ingolstadt-Nuremberg does speed up Munich-Frankfurt trains, but the expensive segment across the mountains serves Berlin-Munich and nothing else. In contrast, Berlin-Halle would speed up trains to Frankfurt as well as to Leipzig

  2. Martin

    The travel time math is of course a strong motivation for dedicated high speed rail. But the real reason it is not built I assume is lacking demand for traveling. Very few intercity connections in the world have enough demand to run even 8 car trains at 15/trains an hour or so. German cities I assume reach these proportions very rarely. High speed rail, without stops, or capacity for slower trains I think would just be to underutilized. All train economics in developed countries is based on regional travel, and train investments that exclude regional travel will only in extreme high demand corridor make sense.

    • Benjamin Turon

      The other issue is construction cost. Its better to build something then nothing, letting the “perfect” being the enemy of the “good enough”. Ideally you can stage it — doing segments as funding becomes available — as has been done in France and South Korea.

      • Benjamin Turon

        I actually would be interested in Mr. Levy’s opinions on rail projects in Malaysia concerning both the electrification of the existing Meter Gauge network and the two Standard Gauge projects.

        • yuuka

          There isn’t even any half done HSR around here to finish in the first place.

          Maybe a hypothetical extension to Penang, like what the old man hinted a while back?

    • df1982

      Germany has an insanely large number of potential HSR city-pairs within a relatively small area, however, so this makes up for the fact that each individual one is not as compelling as, say, Tokyo-Osaka. In Alon’s example, a high-speed line for Berlin-Halle would be used by services from Berlin to Leipzig, Erfurt, Frankfurt, Nuremberg and Munich, and probably also Mannheim, Stuttgart and other south-western cities (including, looking further afield, Strasbourg and Basel beyond Germany). Taken together these markets would likely justify the investment, but it requires HSR trains to through-run to conventional tracks.

      The only issue is if the existing line is already reasonably fast, then what are the relative benefits of just making improvements to the line (straightening alignment, track amplification, enhanced signalling) versus building a totally new line. The latter option might only be justified if congestion on the existing line is such that two parallel services are required (as is the case in England with HS2).

      • Martin

        In order for high-speed rail to truly makes sense, you need at least 10 tph or so. Otherwise, most places would probably be better off dedicating the track and investment to regional trains, for people living within 150 km, making 200 trips a year instead of 1-3. That is where the large demands for actual rail capacity is.

        Germany may have many cities at a convenient distance for HSR, but I doubt the demand is there. And there is always an alternative cost in regional train investment instead.

        For trips under 2 hours, fare economics are also very important. A very large share of the demand very happy trades an extra 20 minutes for half the fare price. In general, politicians overinvest in HSR, and underinvest in regional rail.

        • Alon Levy

          …why do you need 10 tph? Japan only has that on Tokaido and Tohoku, Europe only on the LGV Sud-Est. The Sanyo Shinkansen prints money for JR West and runs 7-8 tph peak, 4-5 off-peak.

          The problem with investing money into Regionalbahn is that Regionalbahn isn’t serving the strongest markets, esp. in the future as Germany’s big cities grow and the rest of the country doesn’t. Munich keeps posting fast population growth numbers and so do its suburbs, but places like Lower Bavaria don’t. Same thing with Berlin and Brandenburg.

          • Martin

            Well, I guess you pretty much made a complete list of financially successful HSR routes in the world (except China). Though, I agree that if there is truly a demand for 7-8 trains that is perhaps okay, but if we are talking 4-5 I don’t think dedicated inter-city rail investments are motivated compared investments in regional/urban transit.

            The Sanyo Shinkansen I assume also use 16 car trains (with 5/row seating), which means that even 7-8 tph actually translates to massive demand/capacity (and in extension profitability), and that the low frequency is actually chosen for traffic economics. You simply are not close to that demand when connecting two <1 million cities in the west. So I am not sure it is a very good example.

            Expansion of regional train would probably facilitate to lots of 1-hour commutes (though that is also perhaps not ideal for other reasons, e.g. it simply consumes a lot of energy to move 300 km every day), as very many mid-sized cities in Germany would be within regional-train commuting distance of those bigger cities you mention. And that kind of commuting patterns creates way more demand for travel. And large parts of Germany (in particular in the west) have a very non-monocentric pattern of urban concentration, which would benefit much more from regional train than very fast intercity connections.

          • Herbert

            Germany has with the most charitable count four cities of one million or more. Berlin, Munich, Hamburg, Cologne.

            Cities of roughly half a million are much more common and together represent more economic strength and traffic demand. Frankfurt, Germany’s air hub is only the fifth largest city.

          • Alon Levy

            I should blog about this at some point, but, metro areas in Germany aren’t the same thing as in the US or Canada or France, with one core and surrounding suburbs. Berlin works like this, and I think so do Hamburg and Munich, but the Rhine-Ruhr really doesn’t, Frankfurt doesn’t (Darmstadt and Mainz are way more historically important than anything comparable in, say, Chicago), the Rhine-Neckar doesn’t, Hanover doesn’t, etc.

            In the US the closest analog to how German metro areas function is Boston. Are Providence and Worcester really Boston suburbs? Who knows, but Providence has its own NEC stop and it’s fairly popular.

          • Herbert

            I think there are several “Metropolregionen” in Germany with lower population density than the German average.

            Sonneberg is a member of the Nuremberg metropolitan region. It is in no way a suburb of Nuremberg despite a direct RE link.

          • adirondacker12800

            …. metro areas in Germany aren’t the same ….. in Germany with lower population density than the German average….

            They let the people from the suburbs use the main railroad station just like they let people from the suburbs use the airport.

            I live someplace with the density of Wyoming but when I want to get to Philadelphia that doesn’t matter because Philadelphia doesn’t. Someone in New Haven who is living car free appreciates that the bus goes to Union Station. That doesn’t stop me from using the long term parking lot in Saratoga Springs. Unless I want to charter a private jet I can’t fly out of Saratoga Springs or Glens Falls. American has vestigial service from New Haven. Neither of us, is going to get urge to take a train, even a very high speed one, to Chicago. They might consider something out of Newark but unless I’m going to Portugal, probably not. Utica, itty bitty Utica, will probably have high speed trains someday. Because it’s along a line that connects tens of millions of people. The density of Watertown or Auburn doesn’t matter because they’ll be getting on a bus that goes to the bus station in Syracuse that has the airport on one side and the train station on the other. Saratoga Springs will likely have a high speed railroad station, like Utica, because it’s on the way that connects ten of millions of people. Thriving metro Fargo North Dakota, won’t. There aren’t enough people out there. And it’s the same size as Saratoga County. Which has the Glens Falls MSA next to it. Which are all part of the Albany CSA. There are no CSAs in North Dakota…. metro areas are different, why do you find this surprising?

      • Martin

        Apparently, Korea runs 18 car trains though. So that is still some massive demand.

    • Matthew Hutton

      400km in 4 hours is a speed you can do on two lane highway in rural Arizona or Utah.

    • Nilo

      The ROW east of New Haven in Connecticut is barely low speed rail given how slow the trains have to travel at various points.

    • Alon Levy

      Yes? The average speed of the express trains in France that the TGV replaced, like Mistral, was 100-120 km/h. The black segments on my map in Bavaria average in that range. Zurich-Basel averages 100-something. It’s pretty standard for legacy rail that’s modernized but not upgraded to continuous 200 km/h the way Berlin-Halle, London-York, London-Birmingham, Stockholm-Gothenburg, and Stockholm-Malmö are (all five average 140, except London-York, which scratches 160).

  3. electricangel

    NY-BOS is about as far as Milan-Florence. You can do the latter trip in1:39; the fastest you can do the former is about 3:50. The price is more than double in the low-cost USA.

  4. Reedman Bassoon

    Alon — no comment about California? It is the ultimate “build only a piece of HSR” example.

    • Alon Levy

      I hope the example of the mountains between Bavaria and Thuringia can make it clear what California should have done: prioritized LA-Bakersfield above all else and build out from there. Not to mention, Los Angeles is far and away the largest city in the state, so the first segment should ideally touch it to maximize travel time benefits for people going to or from it.

      The “it’s okay to leave urban areas slow” point has some applications to Caltrain. The average speed on the Caltrain corridor can be increased, but going beyond about 130 km/h SF-SJ (i.e. 37 minutes) is extremely hard; the 30-minute promises from Prop 1A are at this point lolzy. There are more valuable time savings elsewhere…

      • Nilo

        In positive news though, the lack of dealing with LA to Bakersfield until later lets us convince the powers that be that they should go back to building it through the Tejon Pass.

      • adirondacker12800

        Money dropped down out of the sky and the only place they were prepared enough to spend it was between Bakersfield and Fresno. Spending it to get to Los Angeles wasn’t an option.

      • Martin

        There was never money for the LA to Bakersfield segment, so there never was a choice to start there.

        Tejon Pass is not such a great options given that Palmdale route makes connection with DessertEpress (Victorville Vegas) line much more feasible. Heck, one could see Virgin + California split the cost of that segment since it makes the other lines more valuable.

        • Alon Levy

          I want to say that detouring through Palmdale to serve Vegas is, relative to the sizes of LA and SF, about comparable to detouring the north-south HSR mainlines in Germany to serve Erfurt and Würzburg, but Germany has done both and suffered the attendant construction costs for it.

          • Onux

            I would disagree and point out that Erfurt/Wurzburg are at least the capitals of their respective state/district, far enough from other major cities to have their own economic identity. The comparison would be to Central Valley cities like Fresno and Bakersfield. Palmdale by comparison is bedroom community for LA. One can debate the German approach of serving these intermediate cities, versus the French approach of bypassing them (we know the SNCF proposal wouldn’t have served Fresno/Bakersfield), but going to LA via Palmdale is like routing Munich-Frankfurt through Landshut – there is just no point.

          • adirondacker12800

            I thought Erfurt was the center of rail universe in Gemany and all trains must be timed to meet there…. So that people can take longer trips that involve changing trains instead of just taking the train that goes to where ever they are going.

          • Alon Levy

            Erfurt wasn’t my decision, nor that of rail activists who’ve been here longer than me, who would have preferred a direct Leipzig-Nuremberg route. But now that it’s here, future investment should be planned around what’s been built. Same thing with other questionable HSR lines like the Berlin-Hanover line that bypasses Magdeburg (and was built explicitly as a cohesion project for East Germany, back when Berlin was incredibly poor) and the Hanover-Würzburg main line with several times the tunnel length of the entire LGV network. The worst thing about Erfurt isn’t the time cost of the detour, it’s the cost of tunneling – a direct route would have passed where the mountains were lower and thinner.

          • Alon Levy

            No, it’s not about Hamburg-Munich or Berlin-Frankfurt. Hamburg-Munich today goes via the depicted 250 km/h line to Würzburg. Berlin-Frankfurt slogs through really slow tracks across the mountains to Fulda and then from Fulda to Frankfurt. During the political debate over the design, it was not DB that wanted Erfurt – it would’ve been happy building a direct line from Leipzig to Nuremberg. It was Thuringia demanding the line serve its capital.

            You’re right that the tracks as they are today could be used in the future for Hamburg-Munich and Berlin-Frankfurt, but that would require additional investment that is nowhere on the horizon for reasons that begin with au and end with sterity. There are a bunch of plans for 250 and 300 km/h lines – Stuttgart-Ulm is under construction, Karlsruhe-Basel will be built any decade now, Berlin-Hanover is supposedly slated for 300 km/h upgrades. These do not include the key Göttingen-Erfurt segment without which Hamburg-Munich will remain faster on the current route, or the Erfurt-Fulda route across the mountains.

          • adirondacker12800

            It was Thuringia demanding the line serve its capital.

            How many more times are you going to whine the democracy isn’t doing what you want? I’m not going to go investigate this. If they didn’t get Thuringia on board, nothing would been built. You could be whining about how long the bus trip is.

          • Alon Levy

            You don’t need to investigate it yourself, it was the subject of a big expose in Der Spiegel. (IIRC finding that article was the start of Connor Harris’s career in rail activism?)

        • Eric

          Actually, Tejon makes viable the cheapest possible Las Vegas connection, which is a wye at the junction of I-5 and CA-138.

          This route adds a few minutes to LA-Vegas, but the gains to LA-SF and SF-Vegas make it worthwhile.

          • Martin

            Few mins for LA-Vegas is probably close to half an hour given the distances. However, let’s consider total track required to build LA-SF-Vegas tracks at both locations. If we’re trying to optimize for cost, Palmdale just makes more sense. It’s about 50 miles less of mountainous right of way.

          • Eric

            No, Tejon is one mountain crossing while Bakersfield-Palmdale-LA is two mountain crossings. Clem has done the math and shown than Tejon is easier in terms of mountain crossings.

          • Martin

            @Eric, SF –> LV and LA –> LV via Tejon is TWO crossings, isn’t?

          • Eric

            No, because halfway across Tejon you have a flat connection to LV.

        • adirondacker12800

          Las Vegas is a very popular resort. And a convention destination. Californians love to go there. Las Vegas airport and San Francisco airport are the country’s 7th and 8th busiest airports. Itty bitty Orlando is 12th. Busier than Miami. Which is also a resort. Resorts have a lot of travel demand. While there are things to see and do in Erfurt it’s not a big vacation destination. Metro Sacramento is slightly smaller than metro Las Vegas. Sacramento has the advantage of being the capital of a very large state. Sacramento’s airport is 40th on the list. Las Vegas is anomalous.

          • Nilo

            The time savings and money savings on the Tejon Pass alignment make it very much worthwhile. I think Clem estimated it as a 12 minute savings coupled with saving a couple billion?

        • Onux

          Picking any CAHSR route based around a connection to Vegas is not a logical process. There are in round numbers 14M people in N. California and 21M in S. California; Vegas is about 2M. Although Vegas is an anomaly with very high air traffic for its size, it draws from the entire country, and the travel market between Vegas and California is not significant compared to the NorCal/SoCal route. The CAHSR route should be based on what is best for the SF-LA market, which is Tejon.

          When you do go to Vegas, the route should be via Cajon Pass, following I-15, leveraging the planned HSR route to San Diego. Trying to connect Vegas to SF via HSR is a fool’s errand. The shortest possible route via Tehachapi (from Mojave to Barstow) is 590mi/950km. This is an at best 4 hr trip competing with a 60 or 90 min flight. Ridership would be very low. But sending all traffic to Vegas via this route would kill ridership from the Inland Empire (Riverside and San Bernadino) and San Diego. Having to go through LA/Palmdale/Mojave to Vegas adds 140 mi of travel from these places. All of SoCal should be 1h30 to 2h30 from Vegas by HSR via Cajon, this would dominate existing flight and driving options. Adding an extra hour plus of travel with a Palmdale detour would collapse ridership potential compared to a 45-75 min flight.

          Since there are 50% more people in LA/IE/SD (close to Vegas) than there are in NorCal including the Central Valley (far from Vegas) there is no point trying to build a marginal route for fewer people when a great route for more people exists. I-10 and I-15 represent the shortest route to Vegas from LA, SD and everywhere in between, and follow population corridors through Ontario-S Bernadino and Escondido-Temecula as well. There is, by comparison, no population along CA-58.

          Note, trying to split the difference and running the connection from Palmdale to Victorville doesn’t help. You still have millions of people between LA and SD who have to start their Vegas journey by going West when they want to go East, adding about 110 mi to the trip. In exchange, you make the SF-Vegas trip 50 mi longer, depressing that tiny rail market further.

          • Martin

            If we look at air passengers, there were about 3 million passengers visiting Vegas between the SF and LA area airports. It’s a significant number of people, so I don’t think it’s “Logical” to ignore it. At say $100 per roundtrip ticket, the $300 million will pay for a lot of track maintenance.

          • Onux

            The yearly interstate passengers between N Cal and S Cal airports is about 23 million. Plus about 103 million vehicles cross one of the mountain passes between the two every year. Not all vehicles are cars, and not all car trips are served by HSR, but those that are usually have more than one person per vehicle and you are still looking at 7x (air alone) to 30+x as many people travelling SF-LA as from California to Vegas. 3M is a significant number in isolation, but not in comparison. Ignoring something 1/7th to 1/30th the size of other factors is indeed logical. All the more so when the LA-SF route goes NW, while the LA-Vegas route goes NE. Going the wrong direction for not a lot of people is not good planning.

            Vegas should absolutely be connected to SoCal via HSR as part of a wider California network. Its route should not be mixed with that of the core network, which is SF-LA.

          • adirondacker12800

            If you want to connect the most people with the fastest routes to the most places it’s Tehachapi and Cajon. It’d be fabbbbulous if they reopen any of the stuff railfans think is “better”. Add ten, twenty years to getting anything done. Sounds great!

          • Nilo

            Love to in the face of overwhelming evidence insist that CAHSR made the right choice in picking Tehachapi over Tejon. Somehow you think shaving 11 minutes and billions of dollars off cost just isn’t relevant? Clemof of course states this much better than I could, but really, come on.


          • Onux

            Agreed Nilo, connecting the “most people” (SF-LA) with the “fastest route” (fewest miles) is Tejon, hands down.

          • adirondacker12800

            There are people in places other than Los Angeles and San Francisco. And places people in either them want to go that isn’t the other one.

          • Nilo

            This is true, which is why the connections to San Diego and Sacramento are more important than the one to Las Vegas. Coincidentally both SF-SD and LA-Sac are helped by a Tejon pass alignment.

          • Nilo

            This was an impressive dodge since you looked at the same page I did on wikipedia presumably for San Diego.

            577k San Diego-Las Vegas
            434k San Diego-Oakland
            581k San Diego-San Jose
            921k San Diego-San Francisco

            Of course Sacramento is a major destination from the Bay Area, but you’re right Las Vegas is certainly a more important destination from LA-Sacremento. Overall minimizing the LA-SF time allows you to minimize the SD-SF time, which more or less minimizes the LA-Sac and SD-Sac times. IMO LA-SF and SD-SF are just so much more important than LV that they should be prioritized before anything else.

            Once you build those segments (plus Sacramento for the politicians), building a fancy line to Las Vegas should be politically easy.

          • Alon Levy

            [Slight interjection: the travel market to Vegas should be somewhat financially discounted – somewhat – because it’s so peaky that the operating costs per rider are higher.]

          • adirondacker12800

            And when they build Cajon you’ll minimize travel time between Sacramento and San Diego because they will able to avoid the congestion in and out of Union Station Los Angeles. You won’t if it goes over Tejon. The Riverside-San Bernandino-Ontario MSA is almost the same size as the San Francisco-Oakland MSA. By the time they build Cajon it will likely to be bigger. And they get everywhere, including Las Vegas faster. By avoiding Los Angeles. Like the people who want to go from Sacramento to San Diego will be able to do.

          • Onux

            Adirondacker, why do you keep bringing up Sac vs LV, or Sac-SD. Sac is a minor market; as Nilo’s statistics on SD air travel make clear the major driver of interstate travel in CAis Bay Area to LA Basin. The shortest, cheapest, fastest route from SF-LA is through Tejon, thus it should be the lynchpin of the HSR plan. I-5 sees 172k vehicles/day through Tejon, while CA-58 over Tehachapi only sees 45k.

            Building Tehachapi/Cajon to “avoid the congestion” is a non-starter. First, there is no “congestion” from higher ridership on rail routes as on freeways. Trains are scheduled and run as fast as the schedule/infrastructure allows. Putting more people on the train won’t slow it down. Second, LA is the largest destination for people from SD, Sac, LV and everywhere else, so you would not want to bypass it and lose ridership. It would be like building HSR from Trenton to Bridgeport and not putting a station in Manhattan. Even as an alternate, with some trains to LA, it doesn’t make sense. All trains between the north and south of the state need to stop in LA Union.

            The Inland Empire (Riverside/San Bernadino) is the best example of why not to use MSAs in the US, because they are county based. Both Riverside and SB counties run to the Nevada/Arizona state lines, so the MSA population includes people in Barstow, Palm Springs, Needles, etc. that are not related to the travel market in those cities. If we compare urban areas, SF-Oak is 3.28M (with contiguous San Jose adding 1.66M) while Riv/SB is 1.93M. LA is 12.1M Once again, the Bay Area-LA is the largest market in the state.

            If in some crazy future Riverside grows as big as LA, then you can connect the Cajon line along I-15 to the Tejon line along I-5 via CA-138 as Eric noted. No Tehachapi needed.

            I am not bullish on SF-SD as a major route. The distance will be about 582mi as CASHR plans it; switching to Tejon and taking the direct path down the coast can get this to about 538mi. 850-950 km is a long HSR trip. Tokyo-Hiroshima gets 40-60% mode share at 821km, but Tokyo Fukuoka gets only 8% at 1069km.

          • adirondacker12800

            First, there is no “congestion” from higher ridership on rail routes as on freeways.

            So the 14 tracks of railroad in the general vicinity of Herald Square in Manhattan are silly. They should have just sent the Hudson and Manhattan to Long Island City and be done with it. But then they didn’t ask you and went and built four tracks from Herald Square to Long Island City instead. Silly them. Even sillier, they went to expense of digging a subway under Sixth Avenue. They could have just sent elevated trains out to Queens and had them dive into the tunnel downtown. Four tracks of subway between Jersey City and Manahttan. What were they thinking? While they were busy building two from mainline tracks from Newark. Though to them they were all mainline tracks. There were ticket offices, selling intercity tickets, and baggage services on what is now PATH. Silllllllly them. And the thought occurred to them to send El trains to Queens. From Second Avenue. Or the ten tracks in East New York, Brooklyn. And it’s scandalous that the LIRR and NJTransit have places where there are six. Don’t peer at Sunnyside in Queens or Kearny in New Jersey. There’s more if you can disentangle what is going on. Or in Tokyo, London, Paris….. If somebody who wants to go from Bakersfield to San Diego is in the seat, someone who wants to go from Fresno to San Bernardino can’t sit in it. Or someone who wants to go from Los Angeles to San Francisco either. Unless you think California, Nevada and Arizona are going to forever be like they are right now. And ever and ever and ever. And no one in Fresno wants to go to San Diego and no one in Bakersfield wants to go to Sacramento. And none of them want to go to Las Vegas or Phoenix.

            Google maps says it’s 163 miles from San Bernardino to Bakersfield via Burbank and 165 via Palmdale. It wouldn’t be this tidy but 165 miles at an average speed of 165 takes an hour and 164 miles at an average speed of 82 takes two hours. If someone who wants to go from Los Angeles to Las Vegas is going through Ontario that means they aren’t taking up a seat in Burbank, someone who wants to go from Los Angeles to San Francisco can sit in it. Shift all of the stuff that doesn’t need to be in Burbank away from Burbank people need to pass through Burbank can be there. Or Los Angeles is never going to be a big enough destination to need the capacity because everything is going to look just like it does today forever. And ever and ever..

  5. Gok (@Gok)

    DB’s web site suggests there are only six 4-hour trains between Berlin and Munich per day. How would this €5 billion expenditure on new track compare to increasing frequency instead?

    • Alon Levy

      The Deutschlandtakt plan for 2030 is two trains per hour, one 300 km/h one doing the trip in 3:49 and one 250 km/h doing it in 4:31 with more stops. My belief is that getting Berlin-Munich down to about 2:45 will fill 2 express tph and potentially also 1-2 local tph doing the trip in ~3:45 with more stops (like Ingolstadt and Leipzig).

    • michaelrjames

      Berlin-Munich is used by about 2m air passengers annually. It is the 6th busiest city-pair in Europe. But of course other German pairs are significant like Frankfurt-Berlin at 2m or, using the same rail line if it was LGV, Munich-Dusseldorf at 1.6m, Munich-Hamburg at 1.8m, Frankfurt-Hamburg at 1.4m etc. No accident that in the Wiki list of top 30 city-pairs by air the only French ones are Paris-Toulouse 2.4m (the busiest air route within Europe) and Paris-Nice 2.2m, for the obvious reason that TGV hasn’t reached these two cities but equally, why fast train is inevitable despite the difficulty and expense.
      In the past London-Paris would have been #1 but isn’t even on the list anymore; one suspects the same will happen to London-Amsterdam #14 with 1.6m in 2016 before the Eurostar service began. London-Cologne may have never been on the list but London-Frankfurt at 1.5m (#21) is a sure bet for Eurostar one of these days or decades.

      • herbert

        You’re confusing city pairs and airport pairs, a mistake I’ve mad myself in the past.

        Paris has only one airport for domestic flights (ORY), but London has several, so the busiest routes in Europe are LON-DUB and LON-AMS. The latter at least is getting credible competition with Eurostar…

        • Alon Levy

          Alas, Belgian high-speed lines aren’t really high-speed (Brussels-Aachen averages 140 km/h, same as Berlin-Halle or London-Manchester), so Belgium forms a considerable barrier between London and points east. Thankfully France connects Paris to Saarbrücken and Strasbourg in 1:47 each so in that direction things are fine until you hit the German border.

          • Alon Levy

            The fifth and sixth should be to complete the gaps between Brussels and Leuven and between Brussels and Antwerp, in some order.

          • Diego Beghin

            A high-speed bypass of Leuven and Brussels-Antwerp HSR would both save about 5 minutes each. It would be nice but won’t make the avergae speed anywhere near 200 km/h. The reasons the lines are so slow is that they need to make many stops. A high-speed bypass of Liège could save up to 15 min but is it worth if for the one international train every two hours that would use it*? And you’d lose travelers from Liège itself, which isn’t that much smaller than Brussels (but ok, they have far fewer Eurocrats who need to go back home every weekend).

            *That’s also the reason for no Leuven bypass, the national IC trains stop there.

          • Alon Levy

            Is there a way to speed up the approaches to Liege without skipping the city? In Berlin the trains go 160 km/h south of Südkreuz.

          • Diego Beghin

            You can do 120 km/h almost immediately East of Liège Guillemins, and 4 km further East at Chênée you can go 200-260 km/h. The real problem is the Western approach from Brussels, which goes through some high grades very close to the station. I couldn’t find any specific numbers, but from Google* it seems that going into Liège there’s an elevation loss of ~120 m in just a few km, with maybe 4-5% grades. Speed is limited at 70 km/h on the way down because otherwise the trains might have trouble braking in time. Before EMUs were a thing, there were dedicated power cars which would be temporarily coupled to the trains to make them go up the hill. No idea how cheap/expensive it would be to fix that, some tunnels might be needed? Or maybe the speed limit is overly conservative and could be increased if all obsolete rolling stock is kicked out of the route? The hill with high grades seems to be full of development so a bypass isn’t easy.

            *A nice trick to look at elevations is to check a bike itinerary.

          • Diego Beghin

            PS: According to the wiki, the speed limit is 70 km/h on the way down and 140 km/h on the way up. And I know from personal experience that loco-hauled trains are still used on the route. Do they have worse braking than EMUs as well as worse acceleration? Kicking them out would at the very least simplify operations on the way up…

          • Alon Levy

            I think the braking rates are similar? TGVs actually have higher power-to-weight ratios than ICEs, so in the power-limited portion of the speed-acceleration curve they can electrically brake faster. Then in the traction-limited portion brakes aren’t symmetric with acceleration anyway.

            70 km/h on what looks like a 4% downgrade is really conservative. The Frankfurt-Cologne line runs trains at 300 km/h with 4% grades. Maybe there are some low-speed switches farther down the route where running away at 140 is catastrophic, but the curve near Sainte-Marguerite looks like 670 meters, on which 140 km/h isn’t great but isn’t going to cause a derailment. Then east of the curve the grade moderates to the point that braking shouldn’t be a problem.

          • Diego Beghin

            I can certainly believe that this speed limit is overly conservative. The SNCB makes some really poor decisions, like buying more locomotives to haul double-decker trains through Brussels-Central, which is precisely where long dwells and poor acceleration destroy capacity.

        • michaelrjames

          OK, but doesn’t change the point. And both those routes are across water so … and we’ll see how the Amsterdam route fairs (though of course the reverse route, back to London, is much longer because the anti-Schengen, anti-EU Brits insist on passport control at Lille where all pax have to de-train).

  6. Paul

    When I took rail classes in Switzerland, they used speed vs. space plots to talk about the impact of upgrades on different parts of the line. The “toy model” is really the best case where the upgrade is one long segment. On the other hand, if you start by building short high-speed bypasses the full benefits of those aren’t realized until they start to be connected into longer segments because of the acceleration/deceleration losses at the beginning and end of the segments. I thought those visualizations were helpful for seeing which upgrades were not fully used and which speed reductions (e.g. tight curves, station approaches) had the most impact.

    • Alon Levy

      Yes! And for all the crap I give DB about neglecting speed and having bad punctuality (even worse than SNCF!), it understands it and has reasonably fast urban approaches. The speed zones through Berlin are generous, 120 km/h between Hauptbahnhof and Südkreuz and 160 km/h south of the Ring. This isn’t New York Penn Station and its 10 mph throat.

      • herbert

        DB has the most complex and most congested rail network in Europe, perhaps the world. In addition to its geography forcing a lot of freight through the country, it has barely any dedicated HSR lines entirely free of freight and slower local trains.

        When Taiwan opened a new HSR system, they flew in German train operators to train theirs as they were deemed the best in the world having to deal with the German network

        • Alon Levy

          China and Japan both have higher passenger rail traffic density than Germany. (And China has, on the same network, higher freight density than the US.)

          • Herbert

            Japan built an entirely new network just for Shinkansen and has negligible freight traffic.

            China’s railways date mostly to the 21st century, Germany’s to the 19th

          • Alon Levy

            France doesn’t run freight on the LGVs either. And by Chinese standards Germany has negligible freight traffic.

          • Brendan Dawe

            I was surprised to see that, measured by tonne-kilometers per route-kilometer, Germany’s freight density is an order of magnitude smaller than China, around an eight of the United States, and a twentieth of Russia’s freight density. A peer nation for freight traffic density would be South Korea, a state with no rail-transitable frontiers and a third the area to cover as Germany. Yikes!

          • Chris

            German rail freight traffic is very unevenly distributed. Most branch lines have no freight traffic at all, and even some mainline have very little, while a few transit routes are completely overloaded.

          • Alon Levy

            How overloaded is overloaded? Is the problem too much freight traffic itself, or too much mixing of slow freight traffic and fast intercity passenger traffic? Because the latter gets resolved if passenger trains are moved to the NBS the way China’s moving intercity passenger rail to the high-speed lines.

          • Herbert

            Do you remember what happened when the Rastatt construction oopsie shut down the upper Rhine line? It paralyzed freight traffic throughout the blue banana. In part because there were no electrified routes through which to reroute.

  7. adirondacker12800

    They are upgrading it. Whether or not any individual train goes faster is a good question but they are replacing all of the switches in Interlocking A, west of the platforms, as part of the Trenton-New Brunswick work. You then start to consider that if it’s a bit faster a slightly longer train can clear the interlocking in the same amount of time. … I seem to remember to 30 mph switches. Near term it will mean the trains move through quicker which means they can run slightly longer trains in the same amount of time. Anything beyond that needs new tunnel under the Hudson because even with lousy slow service the tunnels are at capacity. The New Portal bridge probably gains more. The speed limit between the western tunnel portal and Secaucus can be raised from 60 mph.

    • Alon Levy

      First, Shinkansen trains are much bigger than conventional cars – they have 5-abreast seating and are 25 meters long whereas conventional cars are only wide enough for 4-abreast and 20 meters long. Per unit of train mass, they’re very close.

      Second, conventional cars on JR East are mostly low-speed commuter lines, which make a lot of stops and have many acceleration cycles, but also run at 100 km/h. These aren’t the 160 km/h legacy intercity railways that high-speed rail replaces.

      • yuuka

        The JR East legacy intercity network is kind of like the Amtrak Northeast Regional. While most long distance trips are taken by shinkansen, there are still some lines like the Joban (Hitachi, Tokiwa) and Chuo Main (Azusa, Kaiji) that are run by named limited express trains. These operate out of the same depots as the commuter fleet, so I’m not surprised they’re bunched together.

        • Alon Levy

          Yeah, but the vast majority of non-Shinkansen ridership on JR East isn’t the intercity network, it’s the Kanto commuter rail network.

  8. herbert

    The part of the “Nuremberg-Ebensfeld upgraded line” around Bamberg is far from done. They’re building a new tunnel between Fürth and Erlangen for freight trains, they’re building (after years of fights about alignment and surface vs. underground) a new four track stretch through Bamberg… If what can be heard semi-officially is halfway true, they hope to shave some time off the Nuremberg-Erfurt part of the journey. As to why they did not build a connecting curve from Ingolstadt to Munich airport or upgrade the old Halle-Berlin railway to 300 km/h I have no clue whatsoever….

    • Alon Levy

      Yeah, I don’t get why they keep doing mixed infrastructure here. The best industry practice in the world when it comes to combining intercity passenger and freight trains is China, and there the high-speed lines are passenger-dedicated (as in Japan and France) while freight gets to use the legacy network. The Hanover-Würzburg HSR line here has 120 km of tunnel, much of which are due to the need for freight-friendly grades and wide enough curve radii to allow 280 km/h with low superelevation. And they keep doing this, like with Stuttgart-Ulm. Bastante. Just have freight share tracks with slow regional lines, the speeds are pretty similar.

      • gertikan

        They keep doing this because it virtually gives the projects a benefit-cost ratio larger than 1. Of course you won’t see many freight trains on those HSR tracks. Incidentally, that’s what they did on the Nuremberg-Munich line. The ‘environmentally conscious’ freight trains pushed the expensive alignment above 1. And immediately after opening the line, DB forbade any freight trains from using it.

        • Herbert

          Why the benefit cost ratio laws aren’t significantly changed is beyond me.

          For example there are planned extensions (e.g. U2 Nuremberg south to Stein) that fail to reach 1.0 because only the ridership on the newly built part “counts”, not the increase in ridership across the entire line – exactly the opposite of what Alon says re subway extensions

  9. Hugh B

    2 questions tangentially related to the article but important nonetheless:
    1. At what point do you think HSR speeds will plateau as publicly used automotive and airplane speeds have?
    2. What opportunities for improved passenger rail service do you see with the continuing collapse of the coal industry?

    • Alon Levy

      1. My guess is high 300s. Japan is raising speeds from 320 to 360 but it’s glacially slow and has a lot of starts and stops. China has trainsets capable of 380 but is only building up to 350 right now. The current state of the technology is that at 350-360, around two-thirds of the train’s maximum power consumption is spent on fighting air resistance (vs. around one-third at 300), and this is after decades of engineering the perfect noses for reducing drag.

      2. I don’t know – can you elaborate on this? Do you mean that the collapse of American coal will reduce the amount of slow freight traffic on rails, enabling the tracks to be used by passenger trains and faster freight trains?

      • adirondacker12800

        There probably always will be a small amount of coal, I’m not a metallurgist, metallurgical coal is the cheap way to make virgin steel. Cheap natural gas is displacing coal. Railroads run on tight margins. If they aren’t making money hauling coal, some of the lines will close. And they won’t have the money to run fast freight.

    • Eric

      Re #2: I assume few opportunities, because the routes which currently carry coal are generally not very useful for passenger service (mostly low speed intercity lines, with sharp curves and numerous level crossings).

      There are some freight railway corridors in urban areas which should be mostly or exclusively used for passenger rail. Particularly LA Metrolink, Chicago Metra, the Capital Corridor in California, maybe the coastal line in South Florida. However my impression is these are not used much for coal.

      • adirondacker12800

        If the class 1s aren’t making money hauling coal they don’t have the money to run fast freight on the line 1,000 miles away from the mines that doesn’t carry any coal.

  10. yuuka

    Meanwhile you have JR Hokkaido, racing against time to build the Hokkaido Shinkansen to Sapporo before they run out of cash to burn on unprofitable rural lines and require government bailouts.

    In the meantime you have 1tph from Hakodate down south, and less than a train per hour stopping at the intermediate stations of Kikonai and Okutsugaru-Imabetsu. That will probably double or so past opening all the way to Sapporo but I don’t think it’ll still be very heavily used when it’s all said and done.

    Getting some of those passengers flying 500 seater 777s on JAL and ANA must sound really nice, nice enough for them to want to hold through this episode of costs going up and ridership going down (at least for the financial statements circa 2017 that I’m looking at here: http://www.jrhokkaido.co.jp/corporate/mi/senkubetsu/29senkubetsu.pdf).

      • yuuka

        That’s true, but it’s also 1000km long. For reference, China’s G17 train, run by Fuxing vehicles at 350kph, makes the 1000km non stop journey from Beijing to Nanjing in 3h 13min, and 300km to Shanghai takes another hour. Japanese trains make more stops, which I’d think would be politically impossible to remove, so good luck with equalling that even with 360kph trains.

        Given that Hokkaido is, well, Hokkaido, where weather tends to be considerably more unforgiving which drives up operation costs, I guess they’re lucky that JR East is doing the necessary R&D for them (the H5, is basically an E5 with a JR Hokkaido paint job). Also, looking at some updated data, it appears that the government is taking over more of the capital costs as well, so the operation just needs to break even.

        Although given the inelastic pricing of shinkansen which means that it usually ends up more expensive than making the trek out to Narita to fly low-cost (which can also be faster too!), and that you can’t book more than 1 month out, I can see why.

        (in fact, all these management issues with shinkansen demand is something I’d slip Alon, if only everything I can find wasn’t in Japanese)

        • Herbert

          But isn’t aviation more susceptible to weather delays than rail?

          That at least would speak in favor of trains, wouldn’t it?

          In China high speed rail has become viable on ultra long distances not only because of sleepers and low prices but also because China’s airspace is severely restricted by the military which combined with a tough safety culture (remember, China was the first country to ground the 737 Max despite being one of the largest customers) makes aviation prone to delays

          • adirondacker12800

            And they should have shut down LaGuardia when Idlewild, now JFK, opened. Very reasonable people who know what they are talking about say the way to get more people in and out of New York City airports is to close LaGuardia. But they won’t. And Senators from the hinterlands who like to fly from the hinterlands or D.C. into any of them, warp the market to their liking.

          • Alon Levy

            You’d expect all these politicians would manage to get the train that connects where they get money to where they spend it to be faster than 2:50.

          • adirondacker12800

            You would think but then they would have to admit that they aren’t stalwart Real Americans(tm) that drive everywhere. That don’t think trains are a plot to sap the precious bodily fluids of Real Americans(tm) to soften them up for COMMMMMUNISMMMMMM!
            Or they are too stupid to consider it. One of the recent scandals was that someone hired a chartered jet to get to Philadelphia. Unless it was for a conference at an airport hotel and a very tightly scheduled conference at the other airport, I can’t imagine why anyone would do that, but they did. …..and on the third hand trying to impress donors in New York or Philadelphia with you used Acela … makes them think you are a spendthrift who couldn’t schedule this for a Regional… They probably keep it all mum.

  11. Pingback: The High-Speed Rail Germany Needs | Pedestrian Observations

Leave a Reply

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

WordPress.com Logo

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

Google photo

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

Twitter picture

You are commenting using your Twitter 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.