Dawlish avoiding route?

[321362]

If sea levels keep rising, most of Dawlish's seafront will be underwater within 100 years or so. The current line simply isn't possible to keep around super-long-term.

Does the Meldon viaduct need replacing as per NR assessment or can it be repaired to original standards via those who still construct in the same methods?

Meldon Viaduct definitely needs replacing if this ever happens.

 

[Topological]

There is some excellent straw clutching here.

A tunnel will be cheaper than reopening the line via Okehampton.

The fast services do not call at the stations on the Dawlish stretch anyway, so the service would not be lower than now if only the slows went that way. However, because it is only the slows the line could be singled.

IF there is a need for Meldon quarry output then it can run via the current route. There are plenty of paths overnight, or it would just be slotted through as any freight train is now.

Any case for reopening Okehampton has to rely on traffic generated by the stations on that line. That is, there is no case.

== Doublepost prevention - post automatically merged: 28 Jan 2026 ==

That's not what I said at all. I said it has a usefulness in its own right in providing services to places along its route, and its ability to be used as a diversionary route is an additional advantage on top of that.

If as few as 2 or 3 trains per hour incurs an excessive maintenance burden then there is something more fundamental amiss. Are IETs really that punishing to the track? If so they need fixing themselves.

Exactly. With a rotten service, so nobody wants to use it, and then gets closed after some storm damage and we lose it altogether. So it is not safe to construct something which would allow that to happen.

== Doublepost prevention - post automatically merged: 28 Jan 2026 ==

I think the time will come when Dawlish will be under threat anyway, rising sea levels, a warmer wetter climate, the geography of the area soft unstable sandstone, will make cutting tunnels very difficult and costs will soon rise, there is going to be no cheap options, but much of tthis is mute, a reform government won't invest in railways,

Most likily outcome would be the line being severed and the rest possibly run down then closed with car ownership being the chosen alternative

This could be true but it would seem to be a 99% chance that a tunnel route can be identified sufficiently far inland from Dawlish itself. The direct route from Newton Abbott to Exeter is a long way inland from Dawlish.

I would say that a car-based solution would get the nod over IETs going via Okehampton. That is even if they manage to find a business case to open a connection for local trains and the very occasional stock movement between the Plymouth-Penzance network and the rest of the UK.

== Doublepost prevention - post automatically merged: 28 Jan 2026 ==

The reality, the line will get patched up again, and the can 'Dawlish avoider' will get kicked down the road yet again, after all it been kicked down the road for 90 years.

The best we can hope for, re-opening beyond Okehampton will be partly justified as an alternative route, and when the route via Dawlish is unuseable there will be a very basic service via Okehampton to maintain connectivity. The route via Okehampton will not be an attractive alternative, slow with reversals, but it will be touted as a win.

There will be no new construction in the forseeable future, what will happen 15 or 20 years into the future if sea levels continue to rise and storms become more frequent and severe I dont know, but once the principle of a reduced service during and after bad weather becomes established it will probably limp along with more frequent closures for years.

Very true.

However, because Okehampton is not the best soluton. Even when it gets to solution stage it will be something that involves minimal new line to connect the parts of the existing route.

 

[tbtc]

The great thing about spending a billion or more on recreating the route through Dartmoor is that you could rely on it for diversions

Except for the inconvenient fact that... when Dawlish has been properly breached for the first time in about a decade, the Okehampton route is also closed because of the bad weather

(And a line over the moors from Okehampton to Tavistock is presumably going to be even more vulnerable than the Okehampton to Exeter section already is...)

Multiple lines closed 27 Jan in South West and central South due to flooding

Heavy rain from storm Chandra on already saturated ground has flooded multiple rail lines, which are currently closed Castle Cary - Taunton Exeter - Barnstaple Exeter - Oakhampton Exeter - Taunton Salisbury - Romsey Eastleigh - Fareham Par - Newquay Liskeard - Looe Brockenhurst - Bournemouth...

www.railforums.co.uk

There's also the assumption that trains would be diverted that way if Dawlish closed. Which seems as optimistic as the way that people claim an Uckfield line could be used for London-Brighton diversions if the BML were to be closed. Which sounds nice. Except that there's already a two track alternative line for Brighton passengers, but TOCs don't bother diverting along the Arun Valley when the Haywards Heath line is closed

We've finally had the first S&C diversions of WCML services in over a decade (after many years of RRBs up the M6) and I think it's gone to some people's heads. But, fear not, every time there's a leaf on the line somewhere, they'll be dusting down their 1950s track atlas to see whether they can crowbar in a suggestion to reopen some rural route

(A tunnel from Exeter following the rough alignment of part of the A38 would have some merit, you'd finally be able to provide a competitive rail time to Plymouth, so it would bring large benefits every day, rather than just being "useful" occasionally, but it doesn't replicate some abandoned alignment so it's not as exciting to certain folk)

 

[FGWHST43009]

The great thing about spending a billion or more on recreating the route through Dartmoor is that you could rely on it for diversions

Except for the inconvenient fact that... when Dawlish has been properly breached for the first time in about a decade, the Okehampton route is also closed because of the bad weather

(And a line over the moors from Okehampton to Tavistock is presumably going to be even more vulnerable than the Okehampton to Exeter section already is...)

Eventually a new line is going to be needed long-term. With both routes vulnerable to poor weather, it makes sense long-term to have at least two routes, one existing and one new. It's not desirable financially, but a new inland line is likely to need less maintenance than a weather dependent one. A new line would also enable local services to run more frequently on both lines with long-distance journey times reduced as well, especially if extended to Plymouth. Either spend more money now and have a reliable connection when Dawlish is severed again or be forced to spend more later to rebuild Dawlish. If money wasn't an issue I would just totally convert the Dawlish route into a massive bridge offset from the cliffs. But by the time you've done that just build the new inland high-speed line. The Okehampton route could be reopened to via Tavistock to reinstate the local connection there but given that line is also vulnerable a proper main line will be needed. The new line would be electrified and include connections for trains to serve Newton Abbot.

For service patterns, 2tph Plymouth to Paddington, one coming from Penzance and running fast to Exeter and the other stopping at Newton Abbot. Alternatively the Newton Abbot caller could originate from Paignton. 2tph regional from Exeter to Plymouth via Dawlish, Newton Abbot and Totnes along with the 2tph from Exeter to Paignton, with one Exeter-Plymouth regional extended to Penzance. Connect the regionals with Barnstaple/Okehampton/Exmouth services. Extend Cardiff-Exeter regional to Plymouth calling at Newton Abbot. Plymouth-Edinburgh will run fast from Plymouth to Exeter, while Bristol-Manchester could be extended to Plymouth calling at Newton Abbot. Coordinate timetables so Totnes regional trains connect with longer-distance trains at Newton Abbot.

 

[stevieinselby]

But surely a tunnel between Exeter and Newton Abbott would miss out the urban areas or Dawlish and Teignmouth.

Dawlish and Teignmouth are small fry compared with Newton Abbot and Torbay – Dawlish and Teignmouth have a population of about 30k, compared with Newton Abbot, Torquay and Paignton having about 140k.
While missing out Dawlish and Teignmouth would be a sub-optimal solution, if the coastal route becomes completely non-viable then there may not be an option to continue to serve them but it's still going to be essential to maintain the connection between Exeter and Newton Abbot. There could potentially still be some kind of rail service to Starcross and Dawlish Warren if needed.

Is leaving Torbay / Dawlish to Plymouth temporarily unserved a risk too expensive not to take? Newton Abbot to Exeter Central / St Davids / St Thomas is 100,000 journeys per year, to Paddington 66,000, and to Plymouth 71,000, so it isn't an unimportant flow for Newton Abbot.

On the other hand, from Dawlish it's 96,000 per year to Exeter compared to 11,000 to Plymouth, so maybe all that needs to be done is to ensure that trains can run Plymouth – Newton Abbot shuttles (or Penzance – Newton Abbot, or Plymouth – Paignton?) if Dawlish is out of use.

Not sure what you're suggesting here – that trains run between Paignton and Newton Abbot, and between Newton Abbot and Plymouth? If it's a temporary measure during short-term closures at Dawlish then sure, that's viable and is a better option that having road replacement running all the way to Plymouth – but it wouldn't be suitable as a permanent measure.

 

[SynthD]

Yes, this is something to bear in mind, particularly if the railway decided that coastal routes were too difficult under the guise of climate change.

Fairbourne, Wales is a better example of that, but there's no inland former route to attach wishes to, or judge for non-tidal flooding.

 

[Sir Felix Pole]

The great thing about spending a billion or more on recreating the route through Dartmoor is that you could rely on it for diversions

Except for the inconvenient fact that... when Dawlish has been properly breached for the first time in about a decade, the Okehampton route is also closed because of the bad weather

(And a line over the moors from Okehampton to Tavistock is presumably going to be even more vulnerable than the Okehampton to Exeter section already is...)

Point of order - the Okehampton route doesn't go through Dartmoor - it goes around it which is why it required 58 miles between Exeter and Plymouth! The only moorland section is about 4 miles from Meldon to Bridestowe and even that is on the very edge of the National Park - at Sourton it is paralleled by the dual-carriageway A30 so it isn't exactly a barren wilderness. Lydford to Tavistock is entirely in a river valley.

If 'money was no object' (copyright David Cameron 2014) you would build a Spanish type LAV roughly paralleling the A38 with a tunnel under Haldon Hill and some spectacular viaducts to cross the river valleys. To serve Newton Abbot and Torbay you would build a spur to the mothballed Heathfield branch.

The South Devon Railway has been unsatisfactory ever since it was built in 1842-45 - steep gradients, tight curvature, vulnerable coastal section, 60mph speed limit and indirect. Exeter to Plymouth is 44 miles by road; 52 miles by rail. We will, however, continue to reinforce failure until it becomes unsustainable after the inevitable catastrophic breach.

 

[A S Leib]

Not sure what you're suggesting here – that trains run between Paignton and Newton Abbot, and between Newton Abbot and Plymouth? If it's a temporary measure during short-term closures at Dawlish then sure, that's viable and is a better option that having road replacement running all the way to Plymouth – but it wouldn't be suitable as a permanent measure.

Yes; I mean to ensure that services west of Newton Abbot can run independently, but not to do so in ordinary circumstances.

 

[ABB125]

The two long-term options that actually solve the problem (and would be done in most other European countries) are either something like this:
(straight line between the edge of the Exeter conurbation and Newton Abbott)

Or this:
(straight line between the edge of the Exeter conurbation and the edge of the Plymouth conurbation, roughly following the A38, with an optional link to the Heathfield branch as suggested by @Sir Felix Pole)

The alignment of the former Teign Valley Railway could potentially be used in the Exeter suburbs, if the trackbed is still extant, to allow the new route to take a more direct route south, but this would then require a complex interface with the A30/A377 road junction).

 

[Topological]

IF faster journeys to Plymouth were the aim then clearly the second option is better.

However, the first makes sense. Short new alignment, still serves Newton Abbott and Totnes with London trains and keeps the option of leaving the coast line as a loop for local trains (so maintaining their current service levels).

 

[ABB125]

IF faster journeys to Plymouth were the aim then clearly the second option is better.

However, the first makes sense. Short new alignment, still serves Newton Abbott and Totnes with London trains and keeps the option of leaving the coast line as a loop for local trains (so maintaining their current service levels).

Alternatively, you could build the second option but only as far as the Heathfield branch, which would be suitably upgraded. Whilst this would be a slightly longer route than the first option, it leaves open the possibility of completing the route at a later date (pie in the sky, I know! )

Realistically, neither option will happen, this being Britain...

 

[Topological]

Alternatively, you could build the second option but only as far as the Heathfield branch, which would be suitably upgraded. Whilst this would be a slightly longer route than the first option, it leaves open the possibility of completing the route at a later date (pie in the sky, I know! )

Realistically, neither option will happen, this being Britain...

Good point, but a bit of a long route having gone to all the trouble of pointing at Plymouth.

I would not be so negative. I doubt this week will have made anyone think too seriously, but once NPR is underway there could be enough political push in these Tory v Lib Dem seats.

 

[MarkyT]

This was my 2014 suggestion:

 

[Topological]

This was my 2014 suggestion:
View attachment 196870

This map raises a different question. Would you want a station for Chudleigh?

Noting that Chudleigh is not massive, but it is a good location near the A38.

 

[Bevan Price]

Sooner or later, the entire coatal section between Exeter & Teignmouth is likely to become permanently threatened by the sea and if Plymouth is to remain served by rail, a new or reinstated route will be required. The problem is will any government be willing to pay the huge cost ?

To me, the best solution would be a totally new route between Exeter & Plymouth, with much tunneling to avoid both the coast and the steep Devon banks; there would be a spur from this line to Newton Abbot & Torbay. Fully reinstating and upgrading the Okehampton line would be the next best solution.

And don't forget, at some time, parts of the Exeter - Exmouth line may also need to be moved further inland.

 

[MarkyT]

This map raises a different question. Would you want a station for Chudleigh?

Noting that Chudleigh is not massive, but it is a good location near the A38.

I guess it'd be a popular local railhead with a journey time of under 10 mins to Exeter. People from Bovey Tracey and Kingsteignton and the wider area might find it more attractive than Newton Abbot. It could become a new 'Torbay Junction' if the new line was extended further along the A38 towards Plymouth in a later project, with a junction just west of Chudleigh.

 

[SynthD]

The alignment of the former Teign Valley Railway could potentially be used in the Exeter suburbs

This was my 2014 suggestion:

Where do you dive down, is it after city basin junction and before Marsh Barton road, or after that and before Marsh Green road west?

 

[MarkyT]

Where do you dive down, is it after city basin junction and before Marsh Barton road, or after that and before Marsh Green road west?

Likely over Marsh Barton Rd at embankment level, then options beyond. Marsh Green Rd W might be reconstructed with a bridge and ramps over a new rail alignment, with the roundabout replaced by revised junctions either side. The railway continuing elevated (yellow) over the junction complex could avoid much of that. I'd imagine it landing to the north of Alphington Road (blue) in the fields dropping down steeply (2.5%) to dip under the A30 roundabout after passing over Alphin Brook. That should avoid going under sea level, which would be the case if going under Alphin Brook.

 

[The Planner]

So what are the scientific predictions in terms of the coast being inundated by rising sea levels? 20, 30, 50, 100 years? Because if its the latter two, its just sitting on a risk register.

 

[zwk500]

This was my 2014 suggestion:
View attachment 196870

OUt of interest, what does that vertical profile look like if adjusted for the maximum gradient of freight permitted in the standards of 1:100? Or at that point do you just build a 'Devon Base Tunnel' and flatten it out completely.

 

[DevonshireRail]

In reverse order:

3 is no good. Tremendous amount of brand new construction required and the article thinks it might take 10 to 15 years, so multiply by three, even if they do still have all the original GWR surveys etc (if I understand the article correctly).

2 is no good either because by the time you've rebuilt the branch line in a form that would actually be useful you've effectively done 3 anyway, as the GWR recognised.

This leaves 1, which has several advantages...

You're not building something which has no function other than as a Dawlish avoider. Instead you're restoring rail services to Tavistock, providing Tavistock and Okehampton with services to Exeter/Plymouth that aren't maximally indirect, and so on, with the option of using the route as an alternative to Dawlish in times of need.

It's also a route which plainly (at least so I would hope) isn't suitable as a Dawlish replacement. So it minimises the temptation to do something horrible like not repairing the sea wall and just having a couple of withering stub-end branches to Dawlish Warren and Teignmouth.

AIUI there is very little obstruction on the route and the main problem is getting someone to realise that you can repair Meldon viaduct using the same techniques that the original builders would have used to be compatible with it being listed.

It was a double track main line in the first place, so it's much more suitable to be one again than the Teign Valley branch would be.

The gradients, although more prolonged, are less severe than Hemerdon or Rattery, so it is potentially more useful as a freight route than any variation on the other options (including more extreme variations that also avoid Dainton; they can't avoid the others).

Would a "useful Teign Valley Line" be that similar to option 3? Don't get me wrong, the route would have to be altered. The area around Longdown especially because of A the Perridge Tunnel, B the steep gradients, and C, the fact that the route makes a large westerly incursion in this area that massively reduces it's directness. I'm mostly attracted to the fact that the Teign Valley Line could enhance Torbay and NA service frequencies, and that perhaps a third of the track milage is still intact. I think a station around Heathfield/Bovey Tracey/Chudleigh would also be useful and may receive a patronage on par with some Riviera Line stations.

Perhaps I'm massively underestimating the cost of double tracking the route. Great point re the Southern route not being suitable as a Dawlish replacement though.

 

[FGWHST43009]

The two long-term options that actually solve the problem (and would be done in most other European countries) are either something like this:
View attachment 196866 (straight line between the edge of the Exeter conurbation and Newton Abbott)

Or this:
View attachment 196867 (straight line between the edge of the Exeter conurbation and the edge of the Plymouth conurbation, roughly following the A38, with an optional link to the Heathfield branch as suggested by @Sir Felix Pole)

The alignment of the former Teign Valley Railway could potentially be used in the Exeter suburbs, if the trackbed is still extant, to allow the new route to take a more direct route south, but this would then require a complex interface with the A30/A377 road junction).

Second option would be best for capacity. Totnes does lose long-distance services but running a frequent regional service to connect with fast trains at Newton Abbot with a western connection from new line to Newton Abbot. Or run a second London-Plymouth as an extension of the semi-fast to Exeter via Newton Abbot and Totnes while the fast goes non-stop from Plymouth to Exeter. The XC from Plymouth to Edinburgh would also run non-stop from Plymouth to Exeter

 

[35B]

Would a "useful Teign Valley Line" be that similar to option 3? Don't get me wrong, the route would have to be altered. The area around Longdown especially because of A the Perridge Tunnel, B the steep gradients, and C, the fact that the route makes a large westerly incursion in this area that massively reduces it's directness. I'm mostly attracted to the fact that the Teign Valley Line could enhance Torbay and NA service frequencies, and that perhaps a third of the track milage is still intact. I think a station around Heathfield/Bovey Tracey/Chudleigh would also be useful and may receive a patronage on par with some Riviera Line stations.

Perhaps I'm massively underestimating the cost of double tracking the route. Great point re the Southern route not being suitable as a Dawlish replacement though.

It's less double-tracking the route, and more that the alignment was a floodprone corkscrew rural byway that was tightly speed constrained.

By the time you've sorted that lot out, you might as well go with a clean sheet of paper and engineer something that will do what's required.

 

[zwk500]

Likely over Marsh Barton Rd at embankment level, then options beyond. Marsh Green Rd W might be reconstructed with a bridge and ramps over a new rail alignment, with the roundabout replaced by revised junctions either side. The railway continuing elevated (yellow) over the junction complex could avoid much of that. I'd imagine it landing to the north of Alphington Road (blue) in the fields dropping down steeply (2.5%) to dip under the A30 roundabout after passing over Alphin Brook. That should avoid going under sea level, which would be the case if going under Alphin Brook.

The length of your blue line is 427m (give or take) according to Google Maps. In that time, you've got to go from 5.03 clearance above the highway, +tolerance, +bridge deck; to 3.99m (or 4.165m if OLE-ready), + tolerance, +tunnel clearnaces, +structure depth, +distance from structure to highway. So that's 9.02-9.195m +extras, modified by the height difference between the A30 junction and the Alphington Road junction, and not including vertical curves.

1 in 40 for 427m allows a height change of 10.675m. 1.5m is not a lot of tolerance to play with - does the relative height of the two road junctions help or hinder?

Oh and Alphin Brook may have it's own clearance requirements to allow for flood levels. I personally think you'd be better off taking it over the A30 then coming down on a 1:80 or 1:100 gradient after getting out of the city.

 

[Topological]

The length of your blue line is 427m (give or take) according to Google Maps. In that time, you've got to go from 5.03 clearance above the highway, +tolerance, +bridge deck; to 3.99m (or 4.165m if OLE-ready), + tolerance, +tunnel clearnaces, +structure depth, +distance from structure to highway. So that's 9.02-9.195m +extras, modified by the height difference between the A30 junction and the Alphington Road junction, and not including vertical curves.

1 in 40 for 427m allows a height change of 10.675m. 1.5m is not a lot of tolerance to play with - does the relative height of the two road junctions help or hinder?

Oh and Alphin Brook may have it's own clearance requirements to allow for flood levels. I personally think you'd be better off taking it over the A30 then coming down on a 1:80 or 1:100 gradient after getting out of the city.

Going over the A30 is possibly easier to build too, since you would presumably push the bridge over in a weekend road closure rather than needing to dig out the road.

Although that is floodplain land there is also a lot of development going on around the area and you may again actually get a station on the West side as part of a new development there. I would suggest that developer contributions might be an option once the line is identified.

I appreciate that this is now Chudleigh and Exeter West, but realistically that would be fewer stations than the current situation and modern BEMUs should do OK with acceleration.

I know everyone would like a high speed line, but given that it then joins into a slow route at Newton Abbott, this is not about high speed connectivity.

 

[mike57]

So what are the scientific predictions in terms of the coast being inundated by rising sea levels? 20, 30, 50, 100 years?

Whatever scientists say, I think the real answer is they do not know. The warming climate certainly seems to have become more obvious over the last 10 years or so. I think the bigger issue will be the increase in damaging wind storms, both in ultimate strength, and frequency. It will only take one 'perfect storm' or a series of storms to destroy the railway to the point where reconstruction is no longer viable.

 

[Railwaysceptic]

I know everyone would like a high speed line, but given that it then joins into a slow route at Newton Abbott, this is not about high speed connectivity.

This is why post 75 offers the most constructive suggestion.

 

[Topological]

This is why post 75 offers the most constructive suggestion.

Post 75 is about a new high speed route to Plymouth?

That is not the most constructive suggestion, it is purely a pipedream.

 

[HSTEd]

I know everyone would like a high speed line, but given that it then joins into a slow route at Newton Abbott, this is not about high speed connectivity.

Without journey time savings you will have serious trouble making a case though.

Doesn't seem to be much point designing the route for less than 249kph (although TSIs no longer matter), after all the bulk of the rolling stock that would be using it is capable of 200kph plus.

 

[Topological]

Without journey time savings you will have serious trouble making a case though.

Doesn't seem to be much point designing the route for less than 249kph (although TSIs no longer matter), after all the bulk of the rolling stock that would be using it is capable of 200kph plus.

Surely these Exeter to Newton Abbott lines save time over via Dawlish.

I just cannot see you are going to get a business case for all the extra mileage of new line to reach Plymouth and after cutting off Torbay and Totnes.