The effects of closing down the Chain Bridge - case study from Budapest

This article was originally co-authored by Patrik Szigeti and Gy?rgy Jablonszky and published in Qubit.hu , while translated from Hungarian with the help of DeepL.

https://qubit.hu/2021/07/07/a-tomegkozlekedes-a-vartnal-kevesbe-sinyli-meg-a-lanchid-lezarasat-a-gyalogosok-annal-inkabb

The closure of the Chain Bridge has been one of the most talked about transport news stories of recent times - the bridge's prominence is something we experience every day in Budapest, whether it's public transport, a walk along the Danube or complaints from annoyed taxi drivers. But how much impact does the closure actually have? What do the figures show?

Of course, there are many different ways of looking at this question, of which we will focus on public transport and pedestrian traffic in the context of green transport. We have used the timetable database published by BKK and the OpenStreetMap community map to show the role of the Chain Bridge in public transport and the measurable and tangible consequences of the closure in our daily lives.

Datafied cities

The transport system as a series of links connecting cities is also worth looking at from the perspective of network science. This idea is not so new - in fact, it dates back to the first known graph-theoretic problem, posed and solved by the 18th century Swiss mathematician Leonhard Euler, who studied the bridges in the city of K?nigsberg. In the nearly three centuries since then, graph and network theory have become disciplines in their own right, while in recent years data-driven urban planning has flourished.

This has been driven by the acceleration of the spread of geospatial information technology and digital urbanism, which has led to the emergence of platforms and standards, both private and community, such as OpenStreetMap (OSM), which is a community map, or the General Transit Feed Specification (GTFS) format, which brings together public transport databases from major cities around the world.

Budapest is not excluded from these global trends. As far as public transport is concerned, BKK's GTFS static timetable database has been publicly available since 2011, so that, for example, the data of the routes are also displayed in the Google map trip planner. In addition, the Futár app, which has become very popular in recent years, also provides great data: it publishes BKK's real-time journey data.

As for the community mapping of Budapest, in the last three years alone, volunteer hobby mappers have contributed to the city's digital footprint on OSM with more than 50,000 new intersections (coordinate-defined points) and nearly 100,000 new road segments, according to the OSM History Analyzer. Thanks to this enthusiastic work, nearly 500,000 data points will be available to Budapest analysts by July 2021.

The public transport network

Budapest's public transport can easily be shifted into a network mindset, for example by considering stops as nodes and the frequency of services between them as a link strength. In this formulation, the public transport network of Budapest had 5 373 nodes and 6 594 wires in the seven days before the closure of the Chain Bridge (see figure below), which together form 141 007 fixed BKK routes with 6 039 413 stop touch points, counting separately the opposite direction of travel for each stop.

A quantity often used in network science is the so-called intermediacy, which quantifies the role of a node as a network bridge: it determines the extent to which a node plays an inescapable role between different parts of the network.

In other words, high intersections are a bottleneck in the transport network, while low intersections can be easily bypassed. Based on this, the 20 highest public and "most bypassable" interchanges in Budapest:

In addition to the role of nodes, the links between them also provide interesting information. As a reminder, a link represents two consecutive stops, A and B, where the stronger the link between A and B, the more (arbitrary type) of flights stop first at A and then at B. This shows that the busiest stop pairs are almost without exception on the M3 metro line. The best of these is the section between Kálvin tér and Astoria, where more than 6 000 trains pass in a week. In contrast, if we look at all these consecutive A-B stops at the city level, the median observed frequency between them is only about 600 trips, corresponding to an average frequency of 15 minutes. Although the number of trips connecting the two ends of the Chain Bridge is close to 2,000, placing it in the top 5 percent, it is important to point out that the sheer number of trips between two consecutive stops is subject to biases due to the nature of the data source used. For example, although the M3 feeder line has three times as many services, the number of passengers carried is proportionally much higher - while the former is characterised by articulated buses with a capacity of 135 passengers, the latter by normal-sized buses with a capacity of up to half that.

Impact of the closure on BKK services

The renovation of the Chain Bridge will affect several bus routes: buses crossing the bridge will partly use a shortened route and partly cross the adjacent Elisabeth Bridge, where a relief bus service has been launched, which will open up the city centre from Buda along Kossuth Lajos Street and Rákóczi Road. In addition, a ferry runs every half hour during the day between the two banks, but this is mainly of tourist rather than practical importance. But what impact will this have on the public transport network as a whole and on the accessibility of the stops connecting the two sides of the Chain Bridge?

The centre of gravity of the public transport network in Budapest, calculated on the basis of the number of lines (in our case the geographical coordinates of each stop, to which we add the number of lines on the sections concerned as weights), has shifted slightly to the south-east as a result of the diversions (this is due to the transfer of most of the lines on the Chain Bridge to Elisabeth Bridge). The shift is much smaller, however, if we consider not only the number of lines but also their capacity, i.e. ultimately the number of passengers carried. The overall network effect is therefore relatively small, but when looking at Budapest as a whole, it makes a small difference to the location of the transport centre.

The situation is different if we look at the diverted section, i.e. the journeys between the two ends of the bridge. For example, the 105 service between Clark ádám tér and Széchenyi István tér took about two minutes before the closure, according to the timetable. As a consequence of the diversion, this route now takes five times as long, crossing the Elisabeth Bridge and passing the E?tv?s Square stop.

However, as you move away from the bridgeheads, the effect is diminished by the other routes on the network. Although the journey time from Clark ádám tér to Deák Ferenc tér is similar to that of the 105 bus, with alternative connections, it is still more than double the time before the closure. However, the journey time from Apor Vilmos Square (the Buda terminus of the 105) to Deák Square is only 5 minutes (30%) longer than before the closure.

From Buda to Pest and back - would you rather walk?

As we have seen, although bus services will be affected by the closure of the Chain Bridge, it seems to be enough to walk to the bus stop a little earlier and take the bus to your destination. But what if you want to get from Buda to Pest, or vice versa, on foot? The Chain Bridge, at 375 metres in total length, did not pose insurmountable obstacles for pedestrians, but its removal left two alternatives: the Margaret Bridge, almost 2 kilometres from bridgehead to bridgehead, and the Elisabeth Bridge, 1 kilometre away.

To answer this question, we divided Budapest into "neighbourhoods". These are neighbourhoods of historical origin, which further subdivide the official districts. In total, there are 203 neighbourhoods in Budapest, the largest of which is Soroksár, which, for example, makes up the majority of the XXIII district, together with Soroksár suburb and Millenniumtelepp. But also separate neighbourhoods such as Lipótváros, Csepel Belváros or Küls?-Ferencváros are considered to be separate neighbourhoods. The smallest neighbourhoods include the northernmost part of Budapest, the Székesd?l? in District IV, or the rapidly developing Infopark, which is a cluster of office buildings.

Taking these neighbourhoods as a starting point, we examined how long it would take to get from the different neighbourhoods in Buda to the centre of the closest neighbourhood in Pest, and vice versa, on foot before the closure (see figure below). This gives a more nuanced picture of the impact of the closure compared to the district division, as, for example, while the northern half of the split fifth district, Lipótváros, is closer to the Chain Bridge and residents there will have to walk in either direction, from the southern part of the city centre, the Elisabeth Bridge and the Liberty Bridge are convenient options for walking to Buda.

As expected, despite the loud complaints of city centre residents, most of Budapest, 180 out of 203 neighbourhoods to be precise, will not be affected at all by the closure. The people of Pest fared better, with only Terézváros and Lipótváros having a longer walk to Buda, the former by an average of 5 minutes, the latter by 13 minutes more. In Buda, however, an interesting trend is emerging: from Clark ádám Square, out of the city centre, towards Buda?rs and Budakeszi, the walk to Pest is longer from all neighbourhoods.

The data shows that those who want to walk from the Buda Castle district to Pest are the worst off, with a 22-minute increase in walking time between the two areas, or 67%. It may therefore be worth using public transport here. In addition to the Castle, Krisztinaváros, Víziváros and Lipótváros are the most affected, with a one-and-a-half times increase in walking time due to the closure.

So the figures show that the impact of the closure on the public transport network is not as strong as you might think: for journeys between Buda and Pest, the main increase in journey times is between the two bridge ends, but it is not drastic here either, while it drops to zero as you move away from the Danube. The reasons for this observation are presumably strongly related to the robust design of the BKK transport network and the nature and effectiveness of the deliberate alternative routes introduced by BKK to remedy the closure.

Following the public transport mazes, we also looked at the pedestrian options. Looking at the walking times between Buda and Pest before and after the closure, we can say that the time to get from Buda to Pest and from Pest to Buda from the immediate vicinity of the Chain Bridge has increased significantly, as everyone is forced to walk to the nearest bridge if they want to cross the Danube on foot. It is also interesting to note that, while the Pest side is only minimally affected by the closure in terms of walkability, the Buda side is affected all the way to the western agglomeration.

Overall, the closure of the Chain Bridge will cause relatively few difficulties for public transport users and slightly more walking for pedestrians. Two important questions remain to be answered. The first is how the closure will affect cycling and driving; the second is whether it is worthwhile to swap walking for cycling and driving for buses.