Right- versus left-hand running

[AY1975]

For various historic reasons, some rail networks in mainland Europe drive on the right and some drive on the left (for full details of which countries do which, both on the roads and on the railways, see the Wikipedia entry at https://en.wikipedia.org/wiki/Left-_and_right-hand_traffic).

I know that SNCF generally drives on the left, except in Alsace and Moselle which formerly belonged to Germany. As far as I know right-hand running is the norm just about everywhere in Germany, but I've seen some footage of this years Trier area Dampfspektakel (see also the thread on this event at https://www.railforums.co.uk/threads/dampfspektakel-trier-2018.150387/) where the trains appear to be driving on the left.

Do some lines in Germany close to the French border drive on the left (which would be ironic, given that some parts of the SNCF network close to the German border drive on the right)? Or might they just have been running wrong line, or maybe single line working was in operation or both tracks were bidirectionally signalled?

 

[MisterT]

I don't know about that specific point, but here in the Netherlands, we tend to run at the left side on all lines to Belgium from the last station before the border. I guess that other countries have similar solutions in place.
But it might help that 99% of our network is bidirectional and we don't have any specific rules about running right or left. Just follow the signals and you'll be fine

 

[Warwick]

The line from the channel ports to Paris was left running due - it is said - to Stephenson overseeing the building of the line and he had the trains running sur le gauche as they say sur le continent.

 

[LNW-GW Joint]

The Schweers atlases show direction of traffic if different from the norm for that country, and the Luxembourg-Trier-Koblenz line looks to be RHR as normal in Germany.
But most lines are bi-di anyway, and it's quite common for trains to be switched "wrong-line" for a spell, for engineering work.
My recent trips through the Arlberg and Tauern tunnels were both "wrong-line" (LH running on RH routes).

The Wiki article isn't very comprehensive.
Austria is still LHR on the Südbahn (south of the Semmering pass), which extends via Graz and Ljubljana (Slovenia) to Trieste (Italy).
Vienna-Semmering only recently switched to RHR, also Vienna-Břeclav (Nordbahn), as part of the major Vienna area upgrades.
South of Ljubljana, the LHR route towards Zagreb switches to RHR just after Zidani Most, to match the rest of former Yugoslavia.
I think the whole Austrian route to the then Russian border (now in Ukraine, via Krakow and Lviv) was LHR until the 1920s.
This was originally the Nordbahn to Krakow and the Carl Ludwig Bahn on to the Russian border.
It's amazing what changes there have been to the rail network over the last 100 years in Eastern Europe, under different national ownership.

In Spain, lines built buy the Norte railway (CCHNE) were LHR because of the planned connection to LHR France at Irun/Hendaye.
The other main railway, the CFMZA (Madrid Zaragoza Alicante) built its lines for RHR.
If you travel today from Irun on the old Norte route to Madrid, the line switches from LHR to RHR when it takes the modern deviation at Pinar to reach Chamartin station.
The old route continues as LHR from Pinar to Principo Pio, the original Madrid Norte terminus, which is now a Cercanias route.
New Spanish high speed lines are RHR, so at Valladolid coming from the north on LHR you not only go through the gauge-changer, but you end up with RHR through the station and on to the HSL towards Madrid.
The cross-border Figueres-Perpignan HSL has a flyover on the French side of the Pyrenees tunnel, to switch from Spain's RHR to France's LHR.
There's another flyover on the Paris-Strasbourg LGV Est where it reaches the RHR Alsace network.

Going back a bit to the 1830s, the Leipzig-Dresden Railway, one of the first lines in Germany, was originally LHR because of British input, and so I think was the even older Nürnberg-Fürth line.

 

[MarcVD]

Regarding the change in Zidani Most, I was there last friday and can indeed confirm that. To my big surprise, in fact, because I did not know beforehand. Wonderful little station, by the way. Lots of traffic, including freights and loco hauled passengers. Was there for 2 hours and saw some 15 trains during my pause...

 

[LNW-GW Joint]

Regarding the change in Zidani Most, I was there last friday and can indeed confirm that. To my big surprise, in fact, because I did not know beforehand. Wonderful little station, by the way. Lots of traffic, including freights and loco hauled passengers. Was there for 2 hours and saw some 15 trains during my pause...

I've been through Zidani Most twice and both times failed to photograph the old steam loco near the river bridge!
But I did get the one at Zagreb station.
There are quite a lot of old locos on display on plinths in the Balkans and Romania, and a narrow gauge one inside Sofia station.

 

[edwin_m]

In much of the former Austro-Hungarian empire road traffic drove on the left, and some of it stayed that way until German occupation in WW2. Sweden (and Iceland) didn't change until the 1960s. So perhaps not surprising the railways run on the left in some of those places too. Madrid Metro runs on the left while the railways mostly run on the right, but in Paris it's the other way round.

 

[axlecounter]

More interesting is, imho, what kind of technologies were put in place to permit “wrong-line” movements and how (if!) they evolved. It’s an amazing extremely varied world in there

Nowadays there are many countries and lines where the track chosen is merely common practice and nothing more, technically running right or left is the same.

 

[edwin_m]

More interesting is, imho, what kind of technologies were put in place to permit “wrong-line” movements and how (if!) they evolved. It’s an amazing extremely varied world in there

A bi-directional line is essentially a single line and under traditional British signalling would have needed some form of token working. I don't think there have ever been bi-directionally worked multiple track block sections under this type of signalling, with possible exceptions of tracks at stations with boxes both ends (although I can't think of any where a train can run right through both ends of a station on the same track in either direction). There are however some short bi-directional sections under the control of a single interlocking where the interlocking itself prevents conflicting routes - traditionally relying on the signalman to confirm the train has left the section but probably more recently enforced by track circuits.

Large area power signalling (Centralised Traffic Control in the States) didn't come in until the 1920s but the related technology of electromechanical (relay-based) interlocking meant that safety-vital interlocking could be enforced between remote locations by the presence or absence of voltages on the cables between them. A very few British routes gained bi-directional signalling under power signalling schemes in the 1970s and 1980s (parts of the London to Bristol and South Wales main lines and between Leicester and Bedford).

The same technology used in a less radical way allowed various "tokenless block" systems which retained traditional signal boxes but used electric interlocking between them to avoid the need for tokens. These were used on single tracks such as Salisbury-Exeter but again I don't believe these were ever used to create multiple track bi-directional routes in the UK. Bi-dirctional running did however become widespread in many other countries and must have been based on relay interlocking.

Computer-based interlockings came in from around 1990 onwards and made bi-directional signalling slightly easier to implement as with most of them all the vital logic for the whole area was concentrated into one place. Many re-signalling schemes since then have included simplified, or occasionally full, bi-directional running.

I think the reason bi-directional running is rare in the UK even since the advent of power signalling comes down to several factors:
- Density of traffic - most of the time trains are so frequent that there is no scope to send one through "wrong line". Notable that even where it's installed it's rarely used.
- Shortage of capital meaning that "nice to have" features such as bidirectional running during engineering work or service disruption would not be provided.
- Tracks in the UK being closer together, meaning it's difficult to work on one track of a double track safely while the other is open to traffic, meaning that routes have to be closed completely for work rather than being worked single line.
- Tradition. British signalling practice evolved from the absolute block system which only really works on uni-directional lines.

 

[LNW-GW Joint]

Nowadays there are many countries and lines where the track chosen is merely common practice and nothing more, technically running right or left is the same.

And why is there such a rooted objection to bi-di running in the UK?
Even when it is available it is hardly used, and certainly not for routine working (eg for daytime maintenance).
You'd think it would be a priority to have full bi-di running on the principal main lines, especially the 2-track sections.
The northern WCML would surely be a different railway with bi-di working, with far fewer closures.
Maybe the arrival of HS2 will trigger some changes north of Crewe.
HS2 itself, as HS1, will I imagine be bi-di from the start.
ETCS ought to allow bi-di as part of its basic functionality (as it does on the Cambrian, and on high-speed/upgraded routes in Europe).

 

[30907]

And why is there such a rooted objection to bi-di running in the UK?
Even when it is available it is hardly used, and certainly not for routine working (eg for daytime maintenance).

Part of the problem is that in the UK the economic case for major resignalling was the removal of crossovers (which in any caae were typically trailing only) and therefore short sections of bidi (even Simbids) were impracticable. Result - not much use during normal traffic hours.
Whereas typically European installations were done with existing local signalling and unsimplified layouts - OK with typical frequencies, I'd be interested to know whether (eg) NS/Prorail make much use of bidi for maintenance?

 

[axlecounter]

And why is there such a rooted objection to bi-di running in the UK?
Even when it is available it is hardly used, and certainly not for routine working (eg for daytime maintenance).
You'd think it would be a priority to have full bi-di running on the principal main lines, especially the 2-track sections.
The northern WCML would surely be a different railway with bi-di working, with far fewer closures.
Maybe the arrival of HS2 will trigger some changes north of Crewe.
HS2 itself, as HS1, will I imagine be bi-di from the start.
ETCS ought to allow bi-di as part of its basic functionality (as it does on the Cambrian, and on high-speed/upgraded routes in Europe).

I suppose @edwin_m has already answered your question. I don’t personally see bi-directional as a merely “nice to have” feature, but as fundamental feature for an efficient and modern railway system, but I see the point. I may be totally wrong but my impression is that bi-directional has become widespread on smaller countries while it still struggles (here more, there less) on bigger countries, hence maybe confirming (at least partly) edwin’s point.

ETCS itself does nothing for bi-directionality. It just stand over whichever interlocking you choose. It could well be placed (with appropriate interfaces) over a 100 years old system with no bi-di, no block sections, no whatever.

 

[edwin_m]

I notice the North Wales Coast re-signalling is introducing bi-di within its own area. This is a route where the engineering access is based on single line working during the overnight periods where traffic is light but the route can't easily be shut entirely.

I think the point about ETCS is that (at level 2) bi-di running doesn't need any extra hardware, as there are no signals and the detection sections are probably equally suitable for either direction of running. As the interlocking will be computer-based it can be re-configured to have extra routes, which isn't trivial but re-wiring a lot of safety critical relays would be much harder. Clearly there would be extra hardware if the scope for bi-di made it worth putting extra crossovers in, but that's a slightly different issue.

 

[MisterT]

Whereas typically European installations were done with existing local signalling and unsimplified layouts - OK with typical frequencies, I'd be interested to know whether (eg) NS/Prorail make much use of bidi for maintenance?

It is used almost every day/night, when maintenance takes place, e.g. when driving the early morning or late and night trains.

 

[LNW-GW Joint]

The move that impressed me most in Austria was on a fast Railjet overtaking a long freight train moving upgrade in the same direction.
This looked to be part of the normal timetable, not an engineering diversion.
Something you could imagine working over Shap and Beattock with bi-di installed.
It could avoid the ingrained WCML "flighting" of trains which gives long gaps in services every hour.

 

[MisterT]

I remember driving the all-station service between Nijmegen and Den Bosch (the Netherlands) a few years ago with an old Mat'64 train. Those trains were quite old already, without much of real traction control to prevent wheel slip, so I ended up running quite a few minutes late due to slippery tracks.
After the single track bridge just before the Ravenstein bridge, the Intercity service was already running right behind me.
The solution from the signalman was to let me struggle on the regular right track with my train, and the intercity service was given a path on the left track for overtaking, so it did just that while I was pulling out of the Ravenstein station

 

[edwin_m]

The move that impressed me most in Austria was on a fast Railjet overtaking a long freight train moving upgrade in the same direction.
This looked to be part of the normal timetable, not an engineering diversion.
Something you could imagine working over Shap and Beattock with bi-di installed.
It could avoid the ingrained WCML "flighting" of trains which gives long gaps in services every hour.

Might just work over Shap and Beattock with a heavy diesel freight, but if you work out how long an overtake would block the other line in most other places it just isn't possible without introducing huge gaps in the service in the other direction.

 

[MarcVD]

The belgian network is now entirely bi-directional, mostly with automatic block signals in both directions. Normal usage is on the left, wrong side running widely used during maintenance, and for overtakings. What is normally not allowed is both tracks used wrong way, i.e. when a train driver is running on the wrong main, he should never see a train coming in the other direction. On single track lines, signalling is often 'normal running' in one direction and 'wrong main running' in the other, as it allows to keep all signals and cabling on the same side of the track...

 

[Hophead]

Might just work over Shap and Beattock with a heavy diesel freight, but if you work out how long an overtake would block the other line in most other places it just isn't possible without introducing huge gaps in the service in the other direction.

I've had something similar about 5 years ago on the Brighton Main Line:
Northbound freight train stopped before Redhill station; up Arun Valley crossed over to the down slow south of Earlswood and continued thus, through Earlswood station into Redhill. The freight wasn't moving though. There's probably too many trains timetabled now for that to be possible (and, conversely, too few actually running to make it necessary).

 

[MisterT]

What is normally not allowed is both tracks used wrong way, i.e. when a train driver is running on the wrong main, he should never see a train coming in the other direction.

Just wondering, why would that not be allowed? Is it a technical limitation or just some archaic rule?

Some other kind of weird thing that I've seen once, was on the four track line between Utrecht and Amsterdam.
The tracks are, in normal circumstances, used as a two-lane road in both directions. The inner tracks are used for the all station services and the outer tracks for the intercity/fast services to allow planned takeovers.
In this case, the train I was driving was send to the left inner track, so running in the wrong direction on the all station track with the intercity service (it should be the outer right track for an intercity service).
It's was just a bit unusual, but then just after the station of Maarssen I saw the reason for this strange running: On the three other tracks, there were three trains running exactly beside each other in the other direction.
That was kind of a weird sight, and I still have no idea what those trains were doing

 

[gysev]

Just wondering, why would that not be allowed? Is it a technical limitation or just some archaic rule?

It is in the rulebook, although it does happen on short stretches (eg between Brussels Midi and Ruisbroek) during the rush houre.

 

[30907]

It is used almost every day/night, when maintenance takes place, e.g. when driving the early morning or late and night trains.

Thanks. In the UK that would have been done under traditional single line working rules - ISTR that the early southbound passenger trains between Carlisle and Lancaster used to have 15 minutes added to allow for this, as 0300-0800 was the only relatively quiet time over Shap on weekdays. This disappeared with the coming of Pendolinos.

 

[MarcVD]

It is in the rulebook, although it does happen on short stretches (eg between Brussels Midi and Ruisbroek) during the rush houre.

Some sections are explicitly allowed, I would say not more than 10 % of the network. For exemple a section before a divergence going uphill, to avoid stopping trains where it will be more difficult for them to restart if stopped at the signal. But it is a rule, not technically enforced. The idea is to avoid letting the driver to think he's routed towards another train - which is very possible if you drive on the wrong main, in the dark, and see headlights coming in front...

 

[DavidGrain]

The tunnel between Birmingham Moor Street and Birmingham Snow Hill is bi-directional but the only time I have seen any wrong road running was when the stock from a down Chiltern train to Snow Hill returned wrong road to Moor Street to go into the stabling sidings.

A couple of months ago I was on a train from Feldkirch in Austria to Friedrichshafen in Germany and we were regularly being switched between left and right running especially to serve stations.

I have been on a Chiltern train which ran wrong road through High Wycombe and I understand this is regular procedure when passing a slower train at that station.

 

[axlecounter]

What happened in Italy over the years is fun. They now have bi-directional lines with the most different peculiarities. Considering that normally it is alway left-hand running, you have:

- bi-directional lines where parallel and so-called right-right (two trains on their respective wrong) running is permitted
- bi-directional lines where only parallel running is permitted
- bi-directional lines where you can go wrong-way (right) but the left track has to be interrupted
- cheap bi-directional lines where you can go wrong way, the righthand track is equipped, but the station interlockings cannot make routes for the right track, so that right-bound trains must pass the starting signal at danger in order to go right
- and obviously non-bidirectional lines, where going wrong is still possible but only by interrupting the normal track and introducing the telephonic-block system between each stations pair, with a 90kmph speed limit and any signal passed at danger (almost never done but in emergency cases).

 

[talltim]

All three platforms at Chesterfield are bi-di, I'm not sure if the other line is. I can remember some inspired bi-di running (I think before platform three was built) when the lift for the subway was out of order. A southbound train had a wheelchair passenger aboard so it was swapped to platform 1 "wrong line". A northbound train train was then swapped to platform 2 to pass it.

 

[AlexNL]

Trains run on the left hand side on the southern part of the Dutch high speed line.

A few years ago, I was traveling from Breda to Rotterdam. While in the tunnel at Barendrecht, the train was sent through the crossovers, to the right hand side. In that area, a series of dive-unders and fly-overs will bring you from the LHS to the RHS on the normal lines and vice versa.

As my train had just gone through the crossovers, this meant it ended up on the 'wrong' side: we ran from Barendrecht all the way to Rotterdam Centraal over the leftmost track, at line speed.

Here's a map for reference: http://sporenplan.nl/html_nl/sporenplan/ns/ns_nummer/rtd-brd.html

 

[ac6000cw]

On single track lines, signalling is often 'normal running' in one direction and 'wrong main running' in the other, as it allows to keep all signals and cabling on the same side of the track...

As is also often the case in the US - signal heads for both directions mounted on opposite sides of the same post.

 

[MarcVD]

Yes but in Belgium it means you get steady signals in one direction and blinking ones in the other. And also, it look a bit weird to receive the "V" signal telling you enter wrong main regime on a single track line...

 

[AY1975]

At some stations on the border between a right-hand and a left-hand running country the switchover is more noticeable than at others when travelling on cross-border trains. For example, at Roosendaal in the Netherlands (right-hand running) close to the border with left-hand running Belgium, trains actually switch side in the middle of the station. But at Aachen or Strasbourg it is less noticeable from what I remember.