The Dutch government has proposed a Hyperloop test track in the Netherlands. It might also be incorporated in a future commercial line, between Schiphol Airport and Lelystad Airport. This post looks at the constraints and alternatives. At this blog, posts on infrastructure inside the Netherlands are normally in Dutch. This one is in English, because of international interest in Hyperloop projects. (For the same reason, the official report is also in English.)
Amsterdam and Flevoland: click to enlarge…
Is this a realistic plan? The short answer is no. Even if the technology was available tomorrow, the traffic on this specific route is unlikely to justify the investment. The main constraint is the limited function of the proposed line, as a pure airport-to-airport link. That is explained in more detail below.
However, that probably does not matter anyway, since the entire project may be no more than an election gimmick. The Netherlands has a history of ‘electioneering by infrastructure’ – voters are promised new rail lines, new stations, new metro lines, and new tram lines. In this case, the government is also suggesting that the Hyperloop could be extended to the north of the country. This too is a tradition: politicians repeatedly propose a northern high-speed line, but it never gets built. The current Hyperloop proposal is timed to coincide with campaigning for municipal elections, in early 2018.
The only thing that might get built is a section of test track, east of Almere. For that, there are no constraints except the finance. There is an alignment available (a reservation for a road which was never built), and it runs through open flat farmland. You can find it as ‘Vogelweg Zeewolde’ on Google Street View. The official report claims that the test track could be incorporated in the later commercial line, but it seems to be in the wrong place for that.
Function of the proposed commercial line
The proposed Hyperloop would connect Amsterdam Airport Schiphol with Lelystad Airport, which is currently being expanded. Schiphol Airport is very close to Amsterdam: only 9 km from city centre to runway. That is a great advantage for passengers, but an airport in a densely populated urban region brings many problems: congestion, development pressure, and environmental damage. Relocation to an artificial island in the North Sea has been suggested for decades. The prohibitive cost of that option led successive governments to search for alternatives on land.
Some traffic was relocated to regional airports, but Schiphol continued to grow. The government concluded that a relief airport was needed in proximity to Schiphol, and Lelystad is the only candidate. It is already owned by Schiphol Airport, so there are no problems in coordinating strategy.
Lelystad Airport is located in open farmland in the Province of Flevoland. The land was reclaimed from the sea, which has a specific advantage: there are no historic towns and heritage sites to obstruct expansion. Since this is a fully planned landscape, there is no urban sprawl either. The airport is located about 4 km outside the city of Lelystad, a new town with 77 000 inhabitants.
The plan is for an expanded Lelystad Airport to take the low-cost carriers and summer tourist traffic. Schiphol would keep the higher-value long-haul and transfer traffic. However, there is a problem with this strategy: no tourists want to go to Flevoland. Most of them want to go to Amsterdam, and for that reason a rail link to Lelystad airport has been discussed, although there are no concrete plans. The existing rail line Amsterdam – Lelystad – Zwolle passes 5300 m to the north of the terminal.
The proposed Hyperloop would link the two airports directly, with no intermediate stops. That in itself has major consequences for the probable traffic. It also matters exactly where the stations are located. The line might be a airside-to-airside link, which would create a ‘single airport’, from the passengers’ perspective. On the other hand, all non-airside passengers will face airport check-in and security procedures, which would cancel the advantage of high speed. If the line is not an airside-to-airside link, transfer passengers must check out, and then check in again. Whatever the arrangement, the line will serve the two airports, and nowhere else.
Most importantly, it will not serve Amsterdam. Anyone in Amsterdam who wants to use this Hyperloop will first go to Schiphol Airport. Unless their destination is Lelystad Airport, they will need to change mode again, at the end of the Hyperloop. Given the relatively short distances, Hyperloop cannot offer faster journey times if passengers must detour. More on this below, where the line is compared with existing and improved railways.
The two airports are 55 km apart in a straight line, terminal to terminal. At Lelystad Airport itself, the choice of alignment is open: it can exit the airport either north or west, or possibly on the south side. The proposed Hyperloop test line is near the airport, but is not aligned with the future terminal. An extra curve, and an extra line along the N302 road, would be required to incorporate it into the commercial line.
At the western end of the proposed test track, there is no self-evident route toward Schiphol. If you simply extend the line of the test track, it points at Rotterdam. The Hyperloop line to Schiphol must therefore turn further north, and that means it must pass through the city of Almere. Even as it turns, it will cross planned new housing developments. At some point it must also cross the lake, which separates reclaimed Flevoland from the older land to the west.
These constraints imply, that the Hyperloop must follow existing infrastructure around Almere. Notably, there are very limited options for bridges across the lake. The government report suggests, that the line will follow the motorways A6, A1 and A9. That rules out the Vogelweg alignment, which is 4 km south of the A6. (In other words, the report contradicts its own claim, that the test track can be incorporated in the later commercial line.)
The Hyperloop will probably cross the lake alongside the existing rail and motorway bridges, at Muiderberg. There is very little flexibility here, since there are urban areas on the other side: Naarden, Bussum and Huizen. It is an upper-income area, with expensive houses, protected landscapes, and nature reserves. The area is also unsuitable for tunnelling: it is a low ridge of relatively loose sand and gravel.
However, following the motorways brings new problems, as shown by initial studies for a Hyperloop in California. Very few motorways run in straight lines, and their curvature is too sharp for the Hyperloop, running at over 1000 km/h. Instead of running neatly alongside the Californian freeways, or in their median strip, the Hyperloop alignment would need to cross the freeway often, to maintain its own curvature. There are some straight motorways in the Netherlands, but the motorway through Almere is curved.
The problems are worse on the other side of the lake bridge, heading toward Schiphol Airport. The bridge is aligned southwestwards, again toward Rotterdam and not toward Schiphol. The existing motorway turns 60 degrees westward after the bridge, from the A6 to the A1. It then turns 50 degrees west-southwest, to pass through suburban Amsterdam, where it is numbered A9. The A9 then turns 90 degrees to run alongside the A2, and then 120 degrees, to run west-northwest to Schiphol. There is no way a Hyperloop line can follow all these bends (in blue on the map)
There is an alternative: the unbuilt alignment of the A9, which would have run in a wide curve south of the town of Weesp (shown in green). A motorway on this alignment was planned for decades, but rejected on environmental grounds. A tunnel was subsequently rejected on cost grounds. It is possible that the Hyperloop could descend into tunnel from the bridge at Muiderberg, and follow this alignment. That would however cancel one of the claimed advantages of Hyperloop, namely that it needs no expensive tunnels.
The A9 motorway continues through Amstelveen, a suburban municipality. Here too there are curves, but the main problem is limited space. The motorway is jammed in between housing, and some demolition would be required to accommodate the Hyperloop.
After Amstelveen, the motorway curves around Schiphol airport, and then passes under a runway, to reach the terminal zone. Again, the curves in this route are too sharp for Hyperloop, and in any case a direct alignment would be shorter. A possible route into the terminal area is a tunnel under the two eastern runways, about 5 km long.
Assuming it exited Lelystad Airport on the west, this alignment could be about 60 km long, airport to airport. That is probably close to the shortest possible surface alignment, but it does mean that the Hyperloop will definitely avoid Amsterdam. The consequences for travellers are evident, by comparing the Hyperloop with the existing railway.
Schiphol – Lelystad by train
Historically there was no railway through Flevoland: it was under the sea. The land was reclaimed mainly to grow crops, but planners did realise that it offered a faster rail route, from Amsterdam to the northern Netherlands. After long delays, the railway through Flevoland was completed in 2012.
Starting from Schiphol Airport Station, which is under the terminal concourse, a four-track line runs toward Amsterdam. It then splits into two routes. One runs to Central Station, and the other to the relatively new southern station, Amsterdam-Zuid. At Weesp, to the east of Amsterdam, the two routes rejoin.
After Weesp, the line is called the Flevo Line: it crosses the lake into Flevoland, alongside the A6 motorway. It passes through Almere, and continues to Lelystad Station. There is no branch to Lelystad Airport at present, so passengers use a local bus. (The Intercity trains continue to Zwolle, Groningen and Leeuwarden).
A branch to Lelystad Airport was proposed here earlier (in Dutch). Operationally, the simplest option is a 7 km north-south branch from Lelystad Station. It is shown in blue on the map (north is on the left). The branch would carry a shuttle train service, which would take about 6 minutes. Existing Intercity services through Lelystad would be unaffected.
This variant could be combined with a west-to-south curve, which would allow trains from Amsterdam to run directly to the airport (shown in yellow). It could be used by a dedicated service between the two airports.
If a branch to the airport was built, without further improvements, journey time between the two airports would probably be 44 minutes. That is 2 minutes more than the present Schiphol – Lelystad time. Obviously, the Hyperloop would be much faster, under 10 minutes if the promised speeds are implemented. On the other hand, the existing train service stops in Amsterdam, and allows interchange with other Intercity services, for instance to Amersfoort and Alkmaar. Connections will also be improved by the new north-south metro line (2018), which will improve access to the city centre. In Almere there is interchange with regional services to Zwolle and Hilversum. In any case, trains through Schiphol to Lelystad already serve other cities, such as The Hague.
The proposed Hyperloop will not offer any intermediate connections at all. It will offer no through service. It will not take you anywhere, except its two terminal stations, Schiphol Airport and Lelystad Airport. Using this Hyperloop route for longer journeys will carry specific time penalties: exit first mode, transfer, check-in procedure, exit Hyperloop, and transfer to third mode.
Upgraded and new rail lines
The estimated 44-minute train time between two airport stations could be cut, by adding additional rail infrastructure. Some is already under construction: four tracks on the southern route through Amsterdam. Increased line capacity reduces scheduling and operational conflicts, between slow and fast trains. It does not however increase the line speed as such. Both routes through Amsterdam have maximum speeds of 120 km/h, with 80 km/h through some stations and junctions. The rest of the route to Lelystad has line speeds of 130 or 140 km/h.
The existing rail line east of Almere can probably be upgraded for higher speeds, 160 km/h or perhaps 180 km/h. That would cut journey times for a potential airport-to-airport shuttle train, although not spectacularly. A much more substantial improvement was proposed here earlier: a new high-speed line through Flevoland, the ‘Flevo-HSL’. It could also serve as the main route to Berlin, with more high-speed lines (HSL) east of Zwolle. The proposed alignment runs close to the A6 motorway, before turning east towards Zwolle. The Flevo-HSL alignment is notable for passing through Almere in a straight line: it does this by following a high-voltage power line. Trains would not stop in Almere, and the line would bypass Lelystad.
This Flevo-HSL is also compatible with extension to the north-east, to create the long-planned high-speed route to the northern Netherlands. In principle, this proposed Flevo-HSL could run through Lelystad Airport. The question is, whether the airport is important enough to warrant this route variant, and why every high-speed train to Berlin or Groningen should stop there. If only some trains would stop, then a branch line seems preferable.
An additional proposal, which only makes sense in combination with the Flevo-HSL, is a fast exit line from Amsterdam to the Muiderberg bridge, alongside the A1 motorway. This would connect to both routes from Schiphol through Amsterdam. These three proposals – fast exit line, Flevo-HSL, and high-speed airport branch – could cut 15-20 minutes off the 44-minute journey time, airport to airport. That further limits the relative advantage of a Hyperloop.
An additional rail route to Lelystad Airport can be created by extending the proposed airport rail branch south, along the N302 road. This new line would connect Lelystad to Harderwijk, on the old main line to Zwolle.
This would cut journey times from Amersfoort and Utrecht. The proposed Hyperloop can offer no such regional connections, unless it is extended well beyond the original proposal. (And they would have to figure out, how it would stop at intermediate stations).
An isolated non-compatible line is inflexible, compared to an existing railway network. That was the main reason to abandon two Maglev proposals in Germany, the Ruhr Express and the Munich Airport Link. And in fact a Maglev train can stop at intermediate stations, but a Hyperloop can not. Vehicles must leave the running line, slow down, and enter an airlock.
Using existing rail technology, Lelystad Airport can be connected to the ‘legacy network’ in three directions. With these new links, trains through Lelystad Airport could realistically serve about 20 stations, without a change of train. Hyperloop, by definition, is incompatible with all existing transport systems, and therefore has no network synergy.
Single lines with new technology make more sense, where there is no ‘legacy network’ – for instance in Abu Dhabi. But what about extending the one Hyperloop to other places? Again this is only rational, if there is no ‘legacy network’. Replacement of the entire Dutch railway network by Hyperloop lines would not make sense, even if someone else paid for it. The Hyperloop is intended to combine ‘low cost’ infrastructure with high speed, and that led to the choice of small vehicles in small tunnels. Inherently, Hyperloop cannot match the capacity of modern heavy rail systems.
Low capacity might be appropriate for airport-to-airport transfers, since this a limited subset of total passenger travel. Even is capacity is sufficient, however, a dedicated airport-to-airport line, with incompatible technology, can not be effectively incorporated into the rail network
Like any isolated line, the selected route is also vulnerable to changes in travel patterns. In this case, the proposed Hyperloop has a very specific function, determined by a specific airport strategy in the Netherlands. That strategy can change. The Schiphol company might lose control of Lelystad, possibly due to European Commission anti-monopoly policies. If the airports split and become competitors, Schiphol might try to grab all traffic, and Lelystad Airport might simply close. The government might also revive old plans for a ‘second Schiphol’, for instance at Dinteloord, between Rotterdam and Antwerp. Outbound air passengers might switch to airports across the border in Germany, such as Weeze and Düsseldorf. All such changes could sharply reduce travel on the dedicated Schiphol-to-Lelystad link, especially if it was an airside-to-airside link.
There is no Hyperloop anyway…
The Dutch government has made its plans, or at least its suggestions, on the assumption that Hyperloop wil soon be available for passenger traffic. That is wholly unrealistic. Hyperloop is essentially an unproven technology, with research in the initial stages. The technology has already been significantly altered since the earliest concepts, in response to major technical obstacles. That will probably happen again, before a passenger-carrying prototype starts testing. There are known major issues, such as maintaining a near-vacuum tunnel, high acceleration, uncomfortable ride, noise, and passenger safety and evacuation on failure. There are known operational constants, such as complex switches, and unloading via airlock. The infrastructure, a long metal cylinder on pylons, is untested at 100% scale, and its construction costs are unknown.
All these issues are well documented and well commented. Hyperloop proponents are usually emotionally committed to what they see as a utopian technology, and tend to dismiss criticism. All technical problems, they believe, will be overcome. The heroic status attributed to Elon Musk has a significant role here: if he says it can be done, then his admirers think it will be done. That is not, however, a rational basis for infrastructure planning, in the Netherlands or elsewhere.