Is it cheaper to build a road bridge or a railway bridge?

[GrahamA]

We've seen the reopening of some lines over the years but places where road and rail cross can be expensive additions to the cost.

Level crossings are ruled out these days I believe, so if a line is to be reopened and the road is to be kept open I guess a bridge will become necessary. But a road bridge or a rail bridge, and is it always the same answer...

I'm thinking as an example the line between Uckfield and Lewes in E Sussex.

Whether or not you think that line would ever reopen, the path of it crosses the A26 south of Uckfield. The A26 here is effectively acting as the Uckfield bypass. It's a busy road at times but not a dual carriageway. I'm guessing that a road bridge over the railway would be a cheaper option here, but would it? And what if the A26 was a dual carriageway?

 

[edwin_m]

I'd say a road bridge will be cheaper in almost all situations. In particular, road can tolerate much greater gradients, so the approaches to the bridge will generally be much shorter. I can't think of any case in recent UK history where a railway has been raised or lowered to cross a road.

 

[furnessvale]

I'd say a road bridge will be cheaper in almost all situations. In particular, road can tolerate much greater gradients, so the approaches to the bridge will generally be much shorter. I can't think of any case in recent UK history where a railway has been raised or lowered to cross a road.

Additionally, unless you are talking tramway type loadings, the loads imposed on a railway bridge are far greater than those imposed on road bridges requiring a totally different set of design codes, and increased cost.

 

[bangor-toad]

Hi,
I'd agree with edwin_m that a road over bridge is cheaper in most situations.

The Uckfield bypass is one where it's not so obvious though. Immediately to the north of the old rail formation is the River Uck. The road crosses the river on an embankment and bridge. Any road over bridge of the rail formation would need a new, higher, river crossing as well. The waterway is more of a big stream so it's not a huge engineering challenge but it'd add cost & complexity.
The choice would be a new road alignment, probably a little to the west with a new rail & river bridge rather than reworking the existing river bridge.

The rail formation is quite straight and it may be far less disruptive to build up rail embankments and have a short rail over bridge.

Urban legend has it that East Sussex County Council committed to pay the costs of this bridge as park of the Uckfield bypass planning process. But mysteriously no-one has seen any documentary evidence of this...

If Uckfield - Lewes is ever re-opened this detail will need some decent consideration & modelling to work out the cost / difficulty balance.
Cheers,
Mr Toad

 

[Magdalia]

I'd say a road bridge will be cheaper in almost all situations. In particular, road can tolerate much greater gradients, so the approaches to the bridge will generally be much shorter. I can't think of any case in recent UK history where a railway has been raised or lowered to cross a road.

An interesting example, which hasn't been built yet, is the proposed East West Rail alignment between Hauxton Junction and Chapel Hill, which crosses the River Rhee valley and the A10 road (of nearby Foxton level crossing fame).

The 2023 consultation proposed a very high embankment, which became known as the Great Wall of Cambridgeshire, taking the railway over the A10. The 2024 consultation proposed the railway at a much lower level with the A10 crossing over it on a bridge.

 

[furnessvale]

The 2023 consultation proposed a very high embankment, which became known as the Great Wall of Cambridgeshire, taking the railway over the A10. The 2024 consultation proposed the railway at a much lower level with the A10 crossing over it on a bridge.

I hope that doesn't put the (relevelled?) railway at an increased flood risk.

 

[zwk500]

Hi,
I'd agree with edwin_m that a road over bridge is cheaper in most situations.

The Uckfield bypass is one where it's not so obvious though. Immediately to the north of the old rail formation is the River Uck. The road crosses the river on an embankment and bridge. Any road over bridge of the rail formation would need a new, higher, river crossing as well. The waterway is more of a big stream so it's not a huge engineering challenge but it'd add cost & complexity.
The choice would be a new road alignment, probably a little to the west with a new rail & river bridge rather than reworking the existing river bridge.

The rail formation is quite straight and it may be far less disruptive to build up rail embankments and have a short rail over bridge.

Urban legend has it that East Sussex County Council committed to pay the costs of this bridge as park of the Uckfield bypass planning process. But mysteriously no-one has seen any documentary evidence of this...

If Uckfield - Lewes is ever re-opened this detail will need some decent consideration & modelling to work out the cost / difficulty balance.
Cheers,
Mr Toad

Uckfield-Lewes cannot raise the railway very much because of it's closeness to the river. It's a stream most of the time but it has certainly flooded seriously before. The road crosses at a level where it would only need raising a small amount to bridge any new railway formation, and there are also proposals to rejig the town centre road layout to provide a road over rail bridge across the Bus station/car park area.

 

[stuu]

I'd say a road bridge will be cheaper in almost all situations. In particular, road can tolerate much greater gradients, so the approaches to the bridge will generally be much shorter. I can't think of any case in recent UK history where a railway has been raised or lowered to cross a road.

Recent is pushing it, but the railway to the Royal Ordnance Factory north of Bridgwater was raised to cross the M5 when it was built in the early 1970s, so it has happened

 

[dastocks]

I'd say a road bridge will be cheaper in almost all situations. In particular, road can tolerate much greater gradients, so the approaches to the bridge will generally be much shorter. I can't think of any case in recent UK history where a railway has been raised or lowered to cross a road.

Thameslink was lowered to pass under Ludgate Hill in central London around 1990, I think there were also alterations to street and rail levels levels north of Blackfriars where it crosses Queen Victoria St in the 21st century.

 

[edwin_m]

Thameslink was lowered to pass under Ludgate Hill in central London around 1990, I think there were also alterations to street and rail levels levels north of Blackfriars where it crosses Queen Victoria St in the 21st century.

True, but I think there were many factors affecting this other than cost. In particular the City wanted rid of the bridge to open up the view of St Pauls. Was Ludgate Hill raised a bit as well?

 

[swt_passenger]

True, but I think there were many factors affecting this other than cost. In particular the City wanted rid of the bridge to open up the view of St Pauls. Was Ludgate Hill raised a bit as well?

Yes, Ludgate Hill was sort of ‘smoothed out’, you can see at the present junction with Farringdon Road, (Ludgate Circus), the road level has been raised a few feet as it leaves the road junction, you can see the original pavement level nearer the buildings on either side of the road. It seems possible the ‘new’ gradient actually starts on the Fleet St side of the junction?

IIRC before the alterations the start of Ludgate Hill had a slight dip downwards under the old railway bridge.

 

[Hagbard Celine]

I can't think of any case in recent UK history where a railway has been raised or lowered to cross a road.

Galashiels at Currie Road and to a lesser extent Wheatlands Road.

 

[MarkyT]

I'd say a road bridge will be cheaper in almost all situations. In particular, road can tolerate much greater gradients, so the approaches to the bridge will generally be much shorter.

I agree. Also, roads can tolerate tighter curvature (at cost of traffic speed), which gives greater freedom in where to accommodate the ramps for a new bridge. Light rail structures, with curve and gradient limits closer to road norms, are clearly cheaper than those for heavy rail and likely wouldn't need a 25 tonne axle load capability!

== Doublepost prevention - post automatically merged: 12 May 2025 ==

True, but I think there were many factors affecting this other than cost. In particular the City wanted rid of the bridge to open up the view of St Pauls. Was Ludgate Hill raised a bit as well?

Such valuable city centre land and such complexity. Going down so steeply allowed the City Thameslink platforms to start further south. With the ramp north of Ludgate Hill, the majority of the platform would have been on the steep ramp. The platforms aren't entirely level now, but it's a lot better than it would have been if entirely on the ramp. Such a steeply graded station may not have been permitted at all.

 

[Russel]

Given the number of branch line reopening proposals that never seen to get very far due to cost, maybe it's time to look at what can be done with level crossings, maybe some kind of full redesign to make them as safe as possible, so they can be used on smaller branch lines to reduce the costs of bridges.

 

[Annetts key]

Where the railway is already on an embankment that is sufficiently high above the natural ground level, it's cheaper and more practical to put a rail over road bridge in. This occurred on the GWML just south of Bristol not long ago.

Although the loadings are very different, a steel railway bridge may be faster to construct than a road bridge. But I don't know about costs.

 

[MarkyT]

Given the number of branch line reopening proposals that never seen to get very far due to cost, maybe it's time to look at what can be done with level crossings, maybe some kind of full redesign to make them as safe as possible, so they can be used on smaller branch lines to reduce the costs of bridges.

The recently introduced Automatic Full Barrier Crossing Locally monitored design might offer a solution for such secondary lines. This has full barriers closing the road entirely, obstacle detector technology to prove the road surface within the gates is clear and is locally monitored by the train driver observing the crossing visually and the flashing crossing indicator before proceeding. The driver's indicator changes from red to white once the control equipment has gone through its expected sequence correctly, the barriers are proven down and intact, and the red flashing road signals are illuminated. Where a crossing is located next to a station, trains in one direction will always be starting from rest, and in the other will be approaching slowly, expecting to stop at the platform, so the necessarily low crossing speed restriction defined by the approach visibility should not be a problem operationally on minor lines where all trains stop at eth station concerned. Elsewhere on plain track this crossing type, like other locally monitored varieties, is limited to a maximum approach speed of 55mph where sighting is excellent. The warning boards on approach give sufficient notification to pull up short if the crossing isn't seen to be clear. The crossing area must be illuminated automatically on a train's approach to ensure the driver can observe the road at all times of day. The obstacle detection, local monitoring and full barriers should make such an installation just as safe if not more so for pedestrians, cyclists, and motorists as a typical light rail intersection with standard traffic lights alone. Another benefit of the AFBCL type is a short road closure time for each train movement, more like the ~30-60 seconds warning of a Automatic (half) Barrier Crossing Locally monitored (ABCL) than a typical Manually Controlled Barrier with Obstacle Detection (MCB-OD, actually fully automatic in normal operation!), that can sometimes be closed 2 to 3 minutes before a train arrives.

 

[Annetts key]

Given the number of branch line reopening proposals that never seen to get very far due to cost, maybe it's time to look at what can be done with level crossings, maybe some kind of full redesign to make them as safe as possible, so they can be used on smaller branch lines to reduce the costs of bridges.

I honestly don't see why new full width barrier crossings can't be considered. Those that are manually controlled by a signaller or crossing keeper are plenty safe enough.

Of course, the number of people who try to, or actually do jump off a road bridge onto the line below is likely ignored...

Yes, some fool did this at a bridge in my home town, they jumped, but they were not hit by the train, as it used a crossover to cross lines as it was diverging on to a single lead junction after the bridge.

 

[TPO]

Given the number of branch line reopening proposals that never seen to get very far due to cost, maybe it's time to look at what can be done with level crossings, maybe some kind of full redesign to make them as safe as possible, so they can be used on smaller branch lines to reduce the costs of bridges.

Although it's a controversial view, I tend to agree with the sentiment.

It's not insoluable, full barrier crossings plus a robust aproach to enforcement of infringements..... by which I mean red light cameras (with AI if appropriate) of good quality, looking at both front and rear so showing who was driving the road vehicle. Plus a mandatory ban of minimum 1 year plus a punitive fine on conviction, and 100% prosecution of miscreants. That approach basically stopped drink driving back in the day, and would be easier to implement at LCs as you wouldn't need police officers to be out catching the miscreants. The fines would probably go a fair way to offset the initial installation costs for the first couple of years, and starting with the bad actor crossings would be highly educational for the majority of road vehicle drivers.

It would be a less costly way of achieving the end of reducing/preventing collisions between road and rail vehicles at level crossings. And why should the railway have to fund the poor behaviour of road users, these are road costs not railway costs, they should be paid by road users not passengers or freight customers. Or at the very least, the resultant costs of replacing LCs especially on lines that could otherwise re-open should be paid by the Highways authorities rather than the railway.

Yes, somewhat controversial, but history/experience tells us that the only thing which really modifies road vehicle driver behaviour is reliable enforcement. It's the difference of behaviour achieved when a speed limit is enforced by average speed cameras vs when it's not.....

TPO

 

[edwin_m]

Where the railway is already on an embankment that is sufficiently high above the natural ground level, it's cheaper and more practical to put a rail over road bridge in. This occurred on the GWML just south of Bristol not long ago.

Yes I agree in the case of the railway being on an embankment. Another one was near Selly Oak a few years back.

Yes, Ludgate Hill was sort of ‘smoothed out’, you can see at the present junction with Farringdon Road, (Ludgate Circus), the road level has been raised a few feet as it leaves the road junction, you can see the original pavement level nearer the buildings on either side of the road. It seems possible the ‘new’ gradient actually starts on the Fleet St side of the junction?

Such valuable city centre land and such complexity. Going down so steeply allowed the City Thameslink platforms to start further south. With the ramp north of Ludgate Hill, the majority of the platform would have been on the steep ramp. The platforms aren't entirely level now, but it's a lot better than it would have been if entirely on the ramp. Such a steeply graded station may not have been permitted at all.

The vertical curve is very obvious from the platform so I'm pretty sure it starts north of Fleet Street. I'm not sure if there's still a formal restriction on the gradient in a station (unless it's somewhere trains are left unattended) but with the steep gradients each way going out of City Thameslink the worst that could happen if a train was left there unbraked is for it to roll a few metres to the middle of the platform.

I agree. Also, roads can tolerate tighter curvature (at cost of traffic speed), which gives greater freedom in where to accommodate the ramps for a new bridge. Light rail structures, with curve and gradient limits closer to road norms, are clearly cheaper than those for heavy rail and likely wouldn't need a 25 tonne axle load capability!

Given the number of branch line reopening proposals that never seen to get very far due to cost, maybe it's time to look at what can be done with level crossings, maybe some kind of full redesign to make them as safe as possible, so they can be used on smaller branch lines to reduce the costs of bridges.

The crossing area must be illuminated automatically on a train's approach to ensure the driver can observe the road at all times of day. The obstacle detection, local monitoring and full barriers should make such an installation just as safe if not more so for pedestrians, cyclists, and motorists as a typical light rail intersection with standard traffic lights alone.

Another option is to adopt some tramway principles, probably line of sight but also a powerful hazard brake in case any unforeseen obstruction appears (such as a road user who appears to stop but then suddenly moves forward). This is most likely what any future extension to the tram-train lines in South Wales will do.

 

[MarkyT]

Yes I agree in the case of the railway being on an embankment. Another one was near Selly Oak a few years back.

That's usually where a new road is punched through an existing rail embankment. That may be as part of a level crossing removal but would be unlikely to be in the immediate vicinity of the old crossing.

The vertical curve is very obvious from the platform so I'm pretty sure it starts north of Fleet Street. I'm not sure if there's still a formal restriction on the gradient in a station (unless it's somewhere trains are left unattended) but with the steep gradients each way going out of City Thameslink the worst that could happen if a train was left there unbraked is for it to roll a few metres to the middle of the platform.

I recall a 1:100 limit for a new platform if there's no scheduled reversing there, otherwise 1:400. I don't know if that's current. I remember being surprised the Bodmin and Wenford Railway was permitted to build Colsloggett Halt on a very steep gradient, but that's in no way comparable to Thameslink!

Another option is to adopt some tramway principles, probably line of sight but also a powerful hazard brake in case any unforeseen obstruction appears (such as a road user who appears to stop but then suddenly moves forward). This is most likely what any future extension to the tram-train lines in South Wales will do.

For example magnetic track brakes as discussed in the Talerddig thread to provide 'assured stopping', even in cases of local adhesion problems (assuming the equipment is functional and used). FLIRTs definitely have that technology fitted in Germany and Switzerland.

 

[Bald Rick]

Hi,
I'd agree with edwin_m that a road over bridge is cheaper in most situations.

The Uckfield bypass is one where it's not so obvious though. Immediately to the north of the old rail formation is the River Uck. The road crosses the river on an embankment and bridge. Any road over bridge of the rail formation would need a new, higher, river crossing as well. The waterway is more of a big stream so it's not a huge engineering challenge but it'd add cost & complexity.
The choice would be a new road alignment, probably a little to the west with a new rail & river bridge rather than reworking the existing river bridge.

The rail formation is quite straight and it may be far less disruptive to build up rail embankments and have a short rail over bridge.

Urban legend has it that East Sussex County Council committed to pay the costs of this bridge as park of the Uckfield bypass planning process. But mysteriously no-one has seen any documentary evidence of this...

If Uckfield - Lewes is ever re-opened this detail will need some decent consideration & modelling to work out the cost / difficulty balance.
Cheers,
Mr Toad

That initial study was done in the 2007 feasibility, and the answer was a new road bridge.

I honestly don't see why new full width barrier crossings can't be considered.

It can be considered, but in almost every case the whole life cost of a new bridge will be less than the whole life cost of a Level crossing over the 60 year appraisal period. And it significantly reduces the safety and performance risk.

Plus a mandatory ban of minimum 1 year plus a punitive fine on conviction, and 100% prosecution of miscreants. That approach basically stopped drink driving back in the day

A quick look at any court reports in local papers will show that drink driving has anything but stopped.

 

[TPO]

@Bald Rick : yeah, that's why I said "back in the day" with drink driving. It was effective in the days when there were rather more traffic police about, and hence there was a good chance of being caught. Times have indeed changed though and I'm aware it's on the rise again. Whereas cameras are there all the time, and modern technology can connect electronically without needing someone to change the film....

TPO

 

[Snex]

Surely it's just easier to do a road bridge aswell, since you can just build the new bridge off site, then redirect the traffic across it. If you done a rail bridge, then it'll involve road closures or contraflows which will just delay it and push up the cost since there's nowhere to direct the road traffic since the embankment is needed to build the rail bridge in the first place.

The new bridge at Blyth and the Allerdale Bridge on the A1, both in the North East, are two examples of that, and ignoring the fact the new Allerdale Bridge has serious issues, they both opened without any real issue for existing traffic.

Obviously if the railway is already on an embankment, then things are different.