Revised: high-speed line Vinkovci – Novi Sad

Revised with new maps: High-speed line Vinkovci – Novi Sad.

Vinkovci – Šamac cutoff line

The main north-south route through Bosnia was completed in 1966 in Tito’s Yugoslavia. It runs through Samac, Doboj, Sarajevo and Mostar, to Ploce on the Adriatic. However, it starts at the village of Vrpolje, which is on the main Zagreb – Belgrade line, but otherwise ‘in the middle of nowhere’. This illogical route was included in Pan-European Corridor Vc, only because nothing better was planned. No major work was done anyway, and these corridors are now under review.

The logical place for the north-south axis to start is Vinkovci, which is already a main railway junction. It would become more important with the high-speed lines (HSL) proposed here earlier. The Drava plain high-speed line, and the roughly parallel Sava valley high-speed line would converge at Vinkovci, and diverge again as high-speed routes to Belgrade, and to Novi Sad. The HSL Pécs – Osijek – Vinkovci would be the main north-south route into the region. Traffic form the Zagreb direction would however use the proposed Derventa cut-off line.

A new cut-off line from Vinkovci to Šamac would shorten the route between Vinkovci and Doboj, and create a much more logical junction with all other lines in the region.

There are no topographic obstacles to the line, although the area is low-lying. For about 15 km between Vinkovci and Cerna, the new line would parallel the existing branch line to Županja. The new line would also parallel the planned Sava – Danube canal (it has been planned since 1792).

Šamac is located at the confluence of the Bosna and Sava rivers. The original Austro-Hungarian line from Vrpolje opened in 1878: it terminated on the opposite bank of the Sava. The river was the border with Bosnia, which had just been occupied by Austria-Hungary. Only after the Second World War, was Šamac connected to Doboj and Sarajevo, replacing the earlier narrow-gauge route from Slavonski Brod. The line Šamac – Doboj follows the Bosna river: that is a logical route, but the connection to Vrpolje is not.

Click to enlarge: The original railway geography, from an Austro-Hungarian military map of around 1910, Šamac branch highlighted in blue…

Šamac is geographically in Bosnia, and is sometimes known as Bosanski Šamac. However, it is part of the Republika Srpska, which calls it simply Šamac. The settlement across the river, on the Croatian side, is known as Slavonski Šamac. The existing road/rail bridge (single-track) is sufficient for the limited traffic. A new line would need a double-track bridge, with a better approach curve.

At Vinkovci, the new cutoff line would approach from the west, alongside the local line from Županja (Croatian line L210). It would join the main line from Zagreb near the station. The low density in Vinkovci would allow a surface line with overbridges.

The alignment between Šamac and Vinkovci would be determined by local conditions. The planned canal is also intended for drainage, and runs through the lowest terrain, following smaller rivers. The new rail line can not entirely avoid these former marshes, which are clearly shown on the 1910 map. The best option is probably an alignment just west of Babina Greda, crossing both the canal and the motorway. It would then turn north-east, following the canal toward Cerna: this section would be on viaduct.

The new line is intended for long-distance traffic, and would have no intermediate stations. The local line to Županja would be left much as it is, with a third track alongside the new line in Vinkovci itself. The new line would be about 38-40 km long, from the Sava bridge to Vinkovci station. That is about 15 km shorter than the existing route via Vrpolje, and with a new alignment, the journey time will be much shorter.

Sava high-speed line, Zagreb – Vinkovci

This proposed high-speed line (HSL) parallels the existing Zagreb – Vinkovci line, through the Posavina region, along the middle course of the river Sava. The line is part of Pan-European Corridor X and upgrading is officially planned – but nothing on the scale proposed here.

West of Vinkovci, this proposed HSL would join the earlier proposed Drava plain high-speed line, from Maribor and Vienna (Wien). From Vinkovci two high-speed routes would continue to to Beograd (Belgrade), one along the motorway and one via Novi Sad. The new HSL would also connect to other new lines, into Bosnia: more on that later.

The existing line is 255 km long, and consists of Croatian lines M102, M103, and M105. It was not built as a single line, and was not completed until 1929. During the first and second Yugoslavia (1918-1992), it became the national rail axis, and a transit route to south-eastern Europe. Nevertheless, the line is single-track to Novska: there is a parallel line to Novska via Sisak, south of the Sava. The main Yugoslav highway, now the Croatian A3 motorway (E70), was built roughly parallel to the railway.

The line runs between the Sava and the Slavonian hills. Between the two is agricultural land, with many villages. The image below shows a typical section, near Novska. From left to right are the marshland of the Sava (liable to flooding), the motorway, the agricultural land, the railway, the old highway with settlements, and the hills.

Click to enlarge…

The terrain is flat, and much of the line consists of straight sections. However, they are separated by sometimes sharp curves. A new high-speed line can run alongside the existing line in some places, and along the motorway in others. The existing line should also be upgraded, to allow for intensification of services.

Long passenger rail lines have a problem: slow trains obstruct fast trains. If all trains stop at all stations, travel times will be too long. Express trains can overtake stopping trains at some stations, but that works best with a few trains per day. Frequent services require each train to arrive on time at the overtaking station: any delay disrupts all following trains. Adding reserve time minimises delays, but leaves trains standing at platforms for too long. The longer the line, the worse the problems.

Splitting the services is one solution: terminating trains must have their own terminal platforms or reversal tracks, or they will block through trains. The pattern of services partly determines the necessary infrastructure on the Sava valley line.

From Zagreb to Dugo Selo (21 km), there would be separate fast and slow tracks. High-speed trains would not stop at Dugo Selo, all others would: that requires a 6-track station there, or a bypass. Urban-regional services (S-Bahn) would terminate at Dugo Selo. There is no point in extending them another 18 km, to serve one more station at Ivanić-Grad: it can be served by regional and inter-regional trains.

Regional services from Zagreb, every 30 minutes, would stop at all stations after Dugo Selo, and terminate at Slavonski Brod, 190 km from Zagreb. A separate regional service would operate from Slavonski Brod to Vinkovci (65 km).

There were 21 stations and 20 halts between Dugo Selo and Vinkovci, but only four towns. Some villages are too small to justify a station, on an upgraded line. 20-25 stations are enough for this line, giving an average station spacing of around 10-12 km.

Fast inter-regional services (every 30 minutes) would use the fast tracks to Dugo Selo. They would serve only Ivanić-Grad ( 39 km from Zagreb, population 15 000), Kutina (population 15 000, 79 km from Zagreb), Nova Gradiška (population 16 000, 140 km from Zagreb), and Slavonski Brod. The inter-regional service would terminate at Vinkovci. To allow overtaking of regional services, all the remaining stations should have separate through tracks and platform tracks. Services diverging from the line would conform to this pattern of stops. (More later on the branch lines, and new lines).

The high-speed line itself will start at Dugo Selo, or possibly at the Zagreb ring motorway: it would continue all the way to Vinkovci station. Much of it would simply be built parallel to the existing line, on its southern side. At some places, the HSL can better diverge from the existing rail alignment.

On the first 45 km, the HSL could bypass Dugo Selo on the south, and Ivanić-Grad on the north. An alternative is a line along the A3 motorway, starting near the Zagreb ring. The Dugo Selo – Novska section was built in 1897 as a local line: it has many straight sections, with curves between them.

Click to enlarge: The base map is an Austro-Hungarian military map of around 1910, using some German place names.

At Kutina, the HSL can run about 300 m south of the existing station. The old line then turns away from the Sava, to Banova Jaruga station, the junction for the branch line to Pakrac. The HSL can certainly bypass this section, along the motorway: in fact the whole line can be relocated, since Banova Jaruga has only 750 inhabitants. At Lipovljani, the relocated line and the HSL would both rejoin the existing alignment. (With about 3 km of extra track east of Kutina station, the Pakrac branch can be kept operationally separate from the main line, and upgraded as a regional line from Kutina).

Click to enlarge…

Near Kutina, a new line across the Sava from Sisak would join the HSL, and the main line. A similar link is officially planned, but from Sunja to Novska. Both variants allow the Sisak line to duplicate the Sava main line, as a route out of Zagreb. (It is used in this way already, but the L-shaped route Sisak – Novska is inconvenient). With a new cross-Sava link, Novska would lose its role as a junction station, except for a local line to Dubica.

The HSL would continue alongside the main line, to Okučani. On this section, a new line to Banja Luka can diverge south, crossing the Sava at Gradiška, formerly Bosanska Gradiška. (This line was already planned in the first Yugoslavia, before the Second World War).

From Okučani to Zapolje, the HSL can follow the motorway, bypassing Nova Gradiška. East of Nova Gradiška, the old alignment is generally better (straighter), and suitable for a parallel HSL.

Click to enlarge…

A short bypass at Nova Kapela would avoid two curves, and the junction station (lines to Požega and Našice). This junction would also disappear, with the construction of a new tunnel between Nova Gradiška and Požega. The junction station would then be Nova Gradiška itself, a more logical arrangement. On the main line, another short new section is needed, to avoid a sharp curve at Sibinj.

Between Nova Kapela and Slavonski Brod, a new cut-off line can diverge toward Derventa and Doboj. This line would be used by high-speed trains from the HSL, and inter-regional trains from the upgraded main line.

Click to enlarge…

At Slavonski Brod, the motorway passes through the built-up area. So does the rail line, but it is straight. The best option here is to widen the existing alignment, with a tunnel for the HSL under the station zone. Slavonski Brod (population 62 000), is a the largest urban centre on the line. This is a traditional crossing point on the Sava, and the urban area extends across the river. That part is now in the Republika Srpska, and it is called Srpski Brod by Serbs, and Bosanski Brod by Bosnians.

Brod was the junction for the narrow-gauge line to Sarajevo, in Austro-Hungarian times, and is a possible junction with a regional line on a similar alignment: more on that later.

Click to enlarge…

East of Slavonski Brod, the rail line turns away from the Sava, toward Vinkovci. This is an almost straight line across an open plain, and the HSL can simply follow the existing 65-km alignment. About 20 km west of Vinkovci, the Drava plain HSL would join the HSL from Zagreb.

On the main line, the existing junction station at Strizivojna-Vrpolje would lose most of it functions. The Drava plain HSL would serve Djakovo, and a new cut-off line would link Vinkovci directly to Šamac (Slavonski and Bosanski Šamac). The old line would be retained for freight, and possibly a regional service Osijek – Djakovo – Šamac.

Vinkovci is not a large city (33 000 inhabitants), but it is a railway junction, and would retain that function, for the proposed high-speed lines.

The connecting HSL, from Vinkovci to Belgrade, is not described here. With a new alignment along the motorway, it would be about 150 km long. (The alternative HSL route via Novi Sad is not much longer, about 165 km).

The total route from Zagreb to Belgrade would be about 400 km long, with about 95% on new alignment. There are no topographic obstacles on the route: the line crosses the Sava once, in sight of Belgrade station. A line speed of 300-350 km/h is therefore feasible, and a journey time of under 2 hours, with one stop at Vinkovci.

Sombor – Vinkovci inter-regional line

This inter-regional rail line is an extension of the earlier proposal to upgrade the 60-km line Subotica – Sombor. That line connects to the Subotica – Szeged line, which is part of the proposed east-west high-speed route Baja – Subotica – Szeged. These lines would be upgraded to the standards of the German Ausbaustrecken, for 200-230 km/h.

On the diagram: proposed high-speed lines in blue, inter-regional lines in dark green, regional lines in light green, Sombor – Vinkovci options in red. Not all lines are shown.

Note that there is a later alternative for the proposals described here: a direct line Sombor – Osijek via a Danube bridge at Apatin.

The old Szeged – Subotica – Sombor lines form part of a longer rail route, the Alföld – Fiume line. It was built in 1870, as a single line from Oradea to Osijek, for traffic toward the port of Fiume (now Rijeka). At the time the entire line was in the Kingdom of Hungary: it is now divided between Romania, Hungary, Serbia, and Croatia. The line was at right angles to the radial lines from Budapest, which were also cut by new borders.

A Sombor – Osijek link was an integral part of the Alföld – Fiume line, but it was not a straight line, like much of the route. The Danube floodplain here is very wide, and uninhabited. The line turned south, to cross the Danube at Erdut. At first that was by wagon ferry – a bridge was not built until 1910. The line then ran west, parallel to the Danube and Drava, to Osijek.

The line is 61 km long. It is numbered line 20 in Serbia, to the Danube bridge, and then Croatian line R104, as far as Dalj. From Dalj to Osijek, it is line R202.

The line proposed here is a modern version of this link. It would not run to Osijek, which has a relatively isolated position in the rail network, but to Vinkovci, 30 km further south. It would therefore restore the function of the original Dalj – Vinkovci line, built in 1878.

Vinkovci is a classic rail junction: much more so with the proposed Drava plain high-speed line from Vienna, a high-speed line (HSL) from Zagreb, a rerouted line from Sarajevo via Samac, a HSL from Pécs via Osijek, and a completely new HSL to Novi Sad via Vukovar.

The new line Sombor – Vinkovci is intended to connect the proposed HSL Budapest – Sombor – Novi Sad to the new HSL junction at Vinkovci, and to improve connections to other regional lines.

Sombor has a population of 51 000 (municipality 97 000). Vinkovci has 33 000 inhabitants: it is that size because it is a rail junction, and not the other way round.

There are several possible alignments between Sombor and Vinkovci, all of them using some existing rail lines, including the rest of local line R202, Osijek – Vukovar. The main issue is where the line crosses the Danube. The wide floodplain at the Drava confluence is still an obstacle, and most of it is now a national park. One possible alignment, via Vukovar, would avoid it entirely.

The simplest option is to upgrade the existing lines, with a short new link toward Vinkovci. The alignment on the west side of Sombor, where it turns south, would be improved (new curves shown in red). The line south past Prigrevica and Sonta is almost straight anyway. The old railway bridge at Erdut would be replaced, and the curves into the approach tracks widened (in red on the map).

Click to enlarge: The original railway geography, and new alignments. The base map is an Austro-Hungarian military map of around 1910, without present borders, and with old Hungarian and German place names.

The line to Dalj is straight, and so is the line south to Borovo, but the line turns 90° between them. A wider curve, cutting a small part of Dalj, is the solution. The station for Dalj (population 4500) would be on the north side of the village, to allow interchange with services to Osijek.

At Borovo, there is another sharp curve, where the existing line (here M601) turns toward Vinkovci. The upgraded line could go straight on here, and then curve south-west, to join the proposed HSL from Novi Sad to Vinkovci.

Click to enlarge: The original railway geography at Vukovar, with the proposed HSL Vinkovci – Novi Sad, and the link curve to Borovo, in red.

The pattern of services would be simplified:

• inter-regional service Sombor – Dalj – Vinkovci on the upgraded line
• regional service Sombor – Dalj – Osijek, the existing route across the Danube
• regional service Osijek – Dalj – Odžaci – Novi Sad
• regional service Osijek – Borovo – Vukovar (no interchange at Dalj with the other three services).

Regional services on the existing Borovo – Vinkovci line (Croatian line M601), might be abandoned. The proposed HSL Vinkovci – Vukovar – Novi Sad would carry almost all traffic Vinkovci – Vukovar.

The disadvantage of upgrading the original alignment, even with with better curves, is the zig-zag route via Erdut. The line would be 75 km long.

Click to enlarge: Original alignment in black; HSL and new Borovo curve in orange.

There are two alternatives for this route. One is essentially a cut-off line between Sombor and Dalj, starting on the west side of Sombor. Where the existing line turns south, the new line would go straight on to Apatin, with a new station on its eastern edge. This option is shown as a dashed green line on the maps. It would replace the existing branch line Sombor – Apatin (not on the 1910 map, since it was built later).

Click to enlarge: Original alignment in black/white; Apatin cut-off in green; HSL and new Borovo curve in orange.

The new alignment would then turn south, aligning itself with the Dalj – Borovo line. It would cross the Danube, and join this line. The new section would be about 35 km long. That would create two rail lines, almost at right angles: there would be a new station where they crossed. (The best place is beside the existing station, north of Dalj). The line would continue to Vinkovci via Borovo, as described above.

There is however a ridge north of Dalj, parallel to the Danube (Daljska planina). It is not very high (100 m above the plain), but the line would need a 2500m tunnel – and an 800m bridge, and a 2500m viaduct over the flood plain. This variant would however shorten the route by 10 km, and also serve Apatin. There would be only three services: Sombor – Dalj – Vinkovci, Osijek – Dalj – Odžaci, and Osijek – Vukovar.

The third option would re-route the Sombor – Vinkovci route via Vukovar. Trains would use the existing line south from Sombor. Instead of crossing the bridge at Erdut, they would continue south over a new line, passing Vajska. Near the Danube it would join the proposed HSL from Novi Sad. Trains would cross the Danube into Vukovar, serving the HSL station there, and then continue for 16 km to Vinkovci.

Click to enlarge: Original alignment in black/white; HSL and new link in orange.

The total route Sombor – Vinkovci would be 70 km long. The new section could be built for high speeds. In fact that is the only logical option, for a straight line across a flat plain, with no stations. The line would not serve Dalj, and would not connect with other services.

That undermines the logic of this variant, because there is a simpler alternative. A link west of Odžaci could connect the two proposed high-speed lines to Novi Sad – from Budapest via Sombor and from Zagreb via Vinkovci and Vukovar. The link line would be about 25-35 km long, depending on the alignment. This HSL route Sombor – Vukovar – Vinkovci (passing Odžaci), would be about 80 km long, but journey time would be shorter, However, trains from Subotica could not use it without reversal.

Of the options described here, the shortest is through a tunnel under the Dalj ridge. The construction of a new viaduct, bridge and tunnel might however outweigh its advantages. In that case the upgraded original alignments are the best option. (The later proposal for a direct line via Apatin avoids these issues, but the price is a new line across a nature reserve).