How should the Crewkerne issue be resolved both short and long term?

[yorksrob]

You couldn’t get enough water this way. You’d need over 1000 tonnes of water, per night, per mile of route, and it has to be sprayed at a slow rate to soak in, otherwise it will wash ballast and the formation away, block drains, flood neighbours, etc.

That is a lot.

 

[Grecian 1998]

The latter, without the dwell times due to the passing loops and fewer station stops (only calling at Honiton and Yeovil between Exeter and Salisbury would cut 10 station stops, so about 30 minutes faster), that's likely to be slightly quicker than 3 hours.

Axminster is the busiest intermediate station on the line followed by Honiton and then Gillingham. All are significantly busier than Yeovil Junction which seems to be more important operationally than anything else. Axminster is also one of the few stations which usually has a sizeable number of passengers travelling in both directions (Honiton westward is focused on Exeter, Crewkerne eastwards focused on Salisbury and London).

Axminster and Gillingham are fairly insignificant towns in their own right but are major railheads for a large area (Axminster being the railhead for the Jurassic Coast), in the same way that Totnes is smaller than Ivybridge, Teignmouth or Dawlish, but has a much more frequent Paddington service than any of those do due to its catchment area.


In terms of the answer to the question, the realistic short-term answer is probably going to be 'muddle through until it rains a bit'. The West of England has had frequent disruption in recent years with infrastructure problems, SWR strikes and RMT / ASLEF strikes, the 159s showing signs of aging leading to short forming etc. and yet the trains have become busier to the point that 3 coaches will usually lose time due to overcrowding. Passengers do seem to come back swiftly as soon as everything reverts to normal, despite the persistent poor timekeeping.

In the longer term, the best answer is probably to hope that the Lib Dems hold most of their seats along / near the line in the next election and enter a coalition with Labour, then use their position as coalition partner to push for funding for some form of redoubling and resilience work. I don't think this will happen, but I can't see any prospect of getting funding for improvements before then. It worked for the Borders railway...

 

[The Ham]

Axminster is the busiest intermediate station on the line followed by Honiton and then Gillingham. All are significantly busier than Yeovil Junction which seems to be more important operationally than anything else. Axminster is also one of the few stations which usually has a sizeable number of passengers travelling in both directions (Honiton westward is focused on Exeter, Crewkerne eastwards focused on Salisbury and London).

Axminster and Gillingham are fairly insignificant towns in their own right but are major railheads for a large area (Axminster being the railhead for the Jurassic Coast), in the same way that Totnes is smaller than Ivybridge, Teignmouth or Dawlish, but has a much more frequent Paddington service than any of those do due to its catchment area.


In terms of the answer to the question, the realistic short-term answer is probably going to be 'muddle through until it rains a bit'. The West of England has had frequent disruption in recent years with infrastructure problems, SWR strikes and RMT / ASLEF strikes, the 159s showing signs of aging leading to short forming etc. and yet the trains have become busier to the point that 3 coaches will usually lose time due to overcrowding. Passengers do seem to come back swiftly as soon as everything reverts to normal, despite the persistent poor timekeeping.

In the longer term, the best answer is probably to hope that the Lib Dems hold most of their seats along / near the line in the next election and enter a coalition with Labour, then use their position as coalition partner to push for funding for some form of redoubling and resilience work. I don't think this will happen, but I can't see any prospect of getting funding for improvements before then. It worked for the Borders railway...

It's possible that the fat service could still serve 2 maybe even 3 of the 10 stops which is cut, due to the stops needed to allow trains to pass each other where the track switches between single and double.

It's also possible more double track could mean that line speeds could be improved at the same time, however at the very least BEMU's could accelerate faster, either or both of which could see small individual improvements, which help the headline figure quite significantly.

The suggestion was to indicate (using a fairly rough model) the story of headline figure which could be achieved, not necessarily designing an actual timetable.

The actual pattern of stops for the express would require investigation. For example (and I'm not suggesting anything about any stations, just giving an example) the reason that a station is busy may be because it's a local hub attracting travel to/from other local stations, whilst a less well used stop may be more focused on travel to/from further flung places.

Likewise you could have the express services running to given stations at a rate of 0.5tph (for example calling at Honiton one hour and Yeovil the other), as that would still be providing an uplift in frequency over what is currently the case.

It's also worth noting that Honiton may not need the express services calling there (at least not every service) as it's also likely to see an uplift in frequency due to Devon Metro proposals.

 

[Annetts key]

The answer is simple, if you have modern EMUs, they will have far better acceleration and shorter braking distances. Therefore any lost time due to these TSRs can be made up, so the published timetable can be used. Just install OLE along the whole route...

 

[Bald Rick]

The answer is simple, if you have modern EMUs, they will have far better acceleration and shorter braking distances. Therefore any lost time due to these TSRs can be made up, so the published timetable can be used. Just install OLE along the whole route...

Or have them carry their electricity with them…

 

[Kite159]

Short term the timetable should be adjusted to have the trains passing at Gillingham rather than Tisbury, and then drop the Exeter - Waterloo services into the XX:47 paths from Salisbury with a semi fast service running from Salisbury to Waterloo in the existing XX;21 path. Most of the services change crews at Salisbury, although some adjustments would be needed in terms of PNB if the staff is working one of the tighter turns. Advantage is that it gets rid of the not to stop orders, disadvantage is longer journey times (and the need for anybody from Salisbury - Exeter to change at Basingstoke for Clapham.

Heading towards Exeter the paths can stay the same, although this idea would probably need more units.

 

[BerkshireRails]

I doubt that the kit currently exists to do that in a controlled manner with the reasonably high volumes that would be needed.

You could have the weed killing trains doing the spraying of water on the earthworks.

 

[Annetts key]

You could have the weed killing trains doing the spraying of water on the earthworks.

They don't even get the ground wet

 

[Bald Rick]

You could have the weed killing trains doing the spraying of water on the earthworks.

It would be nowehere near enough

 

[Snow1964]

Deleted, moved to other Crewekerne thread

 

[weeklycommuter]

Could you install a watering system to ensure the embankment doesn't dry out again?

Its probably cheaper than most other engineering options. Network Rail could probably pay for a desalination plant and still have it be cheaper than extensive structural changes on the railway.

Or run a nightly train spraying water on the embankment during dry spells.

You couldn’t get enough water this way. You’d need over 1000 tonnes of water, per night, per mile of route, and it has to be sprayed at a slow rate to soak in, otherwise it will wash ballast and the formation away, block drains, flood neighbours, etc.

The Salisbury to Exeter Rail User Group Newsletter No 23 24th July 2025 at the following link suggests on page 4 that cutting back the vegetation more would help.

SERUG - Publications

SERUG Publications SERUG Newsletter No 26 26 June 2026 SERUG Newsletter No 25 February 2026 SERUG Newsletter No 24 11th November 2025 SERUG Newsletter No 23 24th July 2025 SERUG Newsletter No 22 with 2024 AGM details SERUG Newsletter No 21 with General Election update SERUG Newsletter No 21

sites.google.com

The practice of not cutting back sufficient foliage from the trackside, thus allowing plants and trees to suck water from the soil, is likely to be a contributor to the dry embankments.

 

[BerkshireRails]

For It's also worth noting that Honiton may not need the express services calling there (at least not every service) as it's also likely to see an uplift in frequency due to Devon Metro proposals.

As Honiton would be part of Devon Metro wouldn't it make sense to have a higher frequency interchange with devon metro at key stations like Exeter St Davids and Central, as well as Axminster or Honiton where the Devon Metro services will terminate.

Network Rail seems to have fixed the Crewkerne non stopping issue by cutting frequency between Yeovil and Exeter to one train every other hour with an increase journey time of an hour and an hourly service to Yeovil with an increased journey time of a 25 minutes. West of England Service reduction

 

[The Ham]

As Honiton would be part of Devon Metro wouldn't it make sense to have a higher frequency interchange with devon metro at key stations like Exeter St Davids and Central, as well as Axminster or Honiton where the Devon Metro services will terminate.

Network Rail seems to have fixed the Crewkerne non stopping issue by cutting frequency between Yeovil and Exeter to one train every other hour with an increase journey time of an hour and an hourly service to Yeovil with an increased journey time of a 25 minutes. West of England Service reduction

It depends on the frequency.

Even without the express services Honiton retains it's current service as well as gains the service to/from Exeter so would still have a higher frequency than it currently does.

I did also suggest that some of the express services could also call at Honiton. Even 1 express service every 3 hours would still be a sufficient improvement.

 

[BerkshireRails]

It depends on the frequency.

Even without the express services Honiton retains it's current service as well as gains the service to/from Exeter so would still have a higher frequency than it currently does.

I did also suggest that some of the express services could also call at Honiton. Even 1 express service every 3 hours would still be a sufficient improvement.

That could cause some confusion. A better option would be to have a Honiton express stopper on a clock face hourly or half-hourly pattern to tie in with the half-hourly Devon metro services. Having one train every 3 hours stopping at Honiton would be less valuable and less attractive than an hourly call at Honiton. An example of when different calling patterns cause confusion is on the Elizabeth Line between Harrow & Harlington and London Paddington.

 

[The Ham]

That could cause some confusion. A better option would be to have a Honiton express stopper on a clock face hourly or half-hourly pattern to tie in with the half-hourly Devon metro services. Having one train every 3 hours stopping at Honiton would be less valuable and less attractive than an hourly call at Honiton. An example of when different calling patterns cause confusion is on the Elizabeth Line between Harrow & Harlington and London Paddington.

Whilst an hourly express service would be better, it depends on what the demand is to go towards London.

As I've said the existing service will still exist, providing an hourly service.

Let's say the Devon Metro runs at xx:15 and xx:45 with the existing Waterloo service at xx:30 but then a new express service added at xx:00 but only at 05:00, 08:00, 11:00, 14:00, 17:00 and 20:00 (so providing 1.333tph to/from London).

The thing is sometimes the choice isn't a 3 hour frequency or an hourly service but it might be a 3 hour frequency or not being served by the express service.

If the latter was the case then it's far more valuable and attractive to have some express services than none. If cos of the further is the case than clearly an hourly service is preferable.

 

[Kite159]

Best to leave the express services to the faster GWR route to London, rather than messing around with skip stopping on the West of England.

 

[The Ham]

Best to leave the express services to the faster GWR route to London, rather than messing around with skip stopping on the West of England.

Is there more capacity for GWR to run more services from Paddington?

If not, then there may well come a point when their trains are at capacity, it's into that scenario that redoubling the WofE Line and adding extra faster services to the line.

As I've said, the suggestion would be to retain the current service, so the skip stop service would be in addition to that.

 

[BerkshireRails]

Is there more capacity for GWR to run more services from Paddington?

If not, then there may well come a point when their trains are at capacity, it's into that scenario that redoubling the WofE Line and adding extra faster services to the line.

As I've said, the suggestion would be to retain the current service, so the skip stop service would be in addition to that.

There isn't any more capacity in London Paddington or London Waterloo without major infrastructure upgrades in the peaks, as there is already 20tph, and it is designed for 22tph.

You realise that if you need more capacity, a platform extension could be an alternative way of providing extra capacity to London to permit trains longer than 6 cars West of Salisbury.

Whilst an hourly express service would be better, it depends on the demand to go towards London.

As I've said the existing service will still exist, providing an hourly service.

Let's say the Devon Metro runs at xx:15 and xx:45 with the existing Waterloo service at xx:30 but then a new express service added at xx:00 but only at 05:00, 08:00, 11:00, 14:00, 17:00 and 20:00 (so providing 1.333tph to/from London).

The thing is sometimes the choice isn't a 3 hour frequency or an hourly service but it might be a 3 hour frequency or not being served by the express service.

If the latter was the case then it's far more valuable and attractive to have some express services than none. If cos of the further is the case than clearly an hourly service is preferable.

The existing services are already express services. The only way I can see two trains per hour running to Exeter from London via WofE is by extension of the Salisbury stoppers. By this fact, you're better off having an hourly express and an hourly intercity if that is how you want to go, so you have a nice, attractive clock face hourly service pattern rather than a 3-hourly service pattern.


Also, there is a 15-minute gap between services between Honiton and Exeter, which isn't practical for a stopper and express style service. What will probably happen is the train will leave Honiton at XX:22 & XX:52 for the Express and XX:28 & XX:58 for the Stopper. By the time you get to Exeter Central, the express would be at XX:02 & XX:32 or XX:10 & XX:40, depending on if the removal of the station call saves 2 or 4 minutes, as you have four station calls, the express would no longer call at Feniton, Whimple, Cranbrook & Pinhoe.

There could be justification to extend the Devon Metro services to Yeovil Penn Mill to provide a Yeovil Junction to Yeovil Penn Mill half-hourly connection, which would then also call at Crewkerne & Axminster between Honiton and Yeovil Junction. A half-hourly service with a good timed connection at Yeovil Junction could makecompensate the loss of the direct Crewkerne to London service.

P.S. I feel like this thread really should be a part of the Salisbury to Exeter improvements thread as anything you do to Crewkerne is interwoven with Salisbury to Exeter improvement thread.

 

[The Ham]

The existing services are already express services. The only way I can see two trains per hour running to Exeter from London via WofE is by extension of the Salisbury stoppers. By this fact, you're better off having an hourly express and an hourly intercity if that is how you want to go, so you have a nice, attractive clock face hourly service pattern rather than a 3-hourly service pattern.

The three hour service is only at Honiton, it runs every hour, it just doesn't serve the same stations (some key ones, like Salisbury) every hour.

 

[The exile]

You couldn’t get enough water this way. You’d need over 1000 tonnes of water, per night, per mile of route, and it has to be sprayed at a slow rate to soak in, otherwise it will wash ballast and the formation away, block drains, flood neighbours, etc.

Does normal rainfall really equate to 1000t per (mile x width of embankment base) per night?

 

[Bald Rick]

Does normal rainfall really equate to 1000t per (mile x width of embankment base) per night?

No, but then this is a recovery from a long spell of little to no rainfall.

Also, it’s not just the embankment base width you have to worry about, as water has a habit of running off and evaporating. You need to consider a much wider strip.

But to match average rainfall in, say Wiltshire (typical average is 2-3mm per day over the year), you’d need about an average of about 1-200t a day per mile, and then hope it doesn’t rain soon thereafter.

 

[The exile]

No, but then this is a recovery from a long spell of no rainfall.

Ok - I thought it was being suggested as a long term solution, where I reckon on about 150t/mile, assuming an embankment base width of 25metres (which may be massively out)

 

[Annetts key]

The long term solution would be to replace the embankments and the underlying substrate with a more suitable material.

Unfortunately, when the original railways were built, often, embankments were built using whatever material the railway had available at the time. So that could be spoil from cuttings or tunnels, or a mix of ash, coke and other material.

However, the cost of doing this would be extremely expensive and far exceeds the funds available. So, unless there is sufficient political will, the best that can be hoped for, is that some embankment widening and strengthening work is done. Which will also be expensive, but cheaper than complete replacement. But, it could be many years before any such work could start, because simply put, money to pay for it would need to be found.

 

[HSTEd]

No, but then this is a recovery from a long spell of little to no rainfall.

Also, it’s not just the embankment base width you have to worry about, as water has a habit of running off and evaporating. You need to consider a much wider strip.

But to match average rainfall in, say Wiltshire (typical average is 2-3mm per day over the year), you’d need about an average of about 1-200t a day per mile, and then hope it doesn’t rain soon thereafter.

Water that ran off the embankment could likely be recovered and dumped back on top.

In any case, its a pretty small amount in terms of industrial use of non potable water.

 

[Snow1964]

Had a weird conversation with an elderly (ex railway) gentleman who was clearly schooled in the post war find a solution generation.

Actually not weird, more thinking out of the box, but as this is a speculative solution on Crewkerne area problems, going to pitch them here, for discussion.

1) Long term got to cut the vegetation, or rebuild the sub base so there is less clay (which dries out) near the ballast.
2) Irrigating is bit bonkers to artificially keep it damp
3) Accept the speed restrictions might happen every 3 years with dry summers and add more passing loops, because delays vary the crossing points
4) I think he would rather see it restored to a quality double track, but we all know that's unlikely

The conversation then moved on to what about signalling these extra loops
5) West Highland style hydraulic points, but they require low speed and still need signals so need proper points/turnouts
6) do you need fixed signals, could you bring forward cab signals / ECTS

Then I muddled the conversation by saying they use it in mid Wales to Barmouth etc on 158s, which are about to be replaced by 197s

He then said, could you reform some class 159 trains into fixed 6car sets using a Welsh 158 cab at outer ends, or simply switch the ECTS panels to 159s.

I have no idea if it could be done, but afterwards I thought you could have a subset of 6car sets for Salisbury-Exeter section, and couple on standard units at Salisbury for Waterloo section.

What I have no idea about (I know nothing about signalling costs, apart from not cheap) is if adding 2 or 3 conventionally signalled extra passing loops, is cheaper than an ECTS solution, bearing in mind only a very limited fleet ever pass along that line, and there is an outline plan to fit locos with ECTS anyway.

My gut feeling is no modern manager has the skill to successfully justify why financially and practically these are better options to action now, than the paralysis let's just do nothing until autumn, and hope it doesn't happen again in next few decades. In other words the risk of repeat with its negative financial implications has worse cost-benefit number that upgrading now to provide a solution.

 

[Bald Rick]

Autumn on this route is much more easily (and cheaply) solved by installing Double Variable Rate Sanders on to the units, or more likely, buying new units with that technology installed.

Clay banks is a more tricky issue, however the engineers have learned a lot from the last dry spell and there has been a definite improvement this summer in comparison. Although not on this line, this time, I imagine the lessons learned elsewhere will be used here next time.

 

[Annetts key]

Had a weird conversation with an elderly (ex railway) gentleman who was clearly schooled in the post war find a solution generation.

Actually not weird, more thinking out of the box, but as this is a speculative solution on Crewkerne area problems, going to pitch them here, for discussion.

1) Long term got to cut the vegetation, or rebuild the sub base so there is less clay (which dries out) near the ballast.
2) Irrigating is bit bonkers to artificially keep it damp
3) Accept the speed restrictions might happen every 3 years with dry summers and add more passing loops, because delays vary the crossing points
4) I think he would rather see it restored to a quality double track, but we all know that's unlikely

The conversation then moved on to what about signalling these extra loops
5) West Highland style hydraulic points, but they require low speed and still need signals so need proper points/turnouts
6) do you need fixed signals, could you bring forward cab signals / ECTS

Then I muddled the conversation by saying they use it in mid Wales to Barmouth etc on 158s, which are about to be replaced by 197s

He then said, could you reform some class 159 trains into fixed 6car sets using a Welsh 158 cab at outer ends, or simply switch the ECTS panels to 159s.

I have no idea if it could be done, but afterwards I thought you could have a subset of 6car sets for Salisbury-Exeter section, and couple on standard units at Salisbury for Waterloo section.

What I have no idea about (I know nothing about signalling costs, apart from not cheap) is if adding 2 or 3 conventionally signalled extra passing loops, is cheaper than an ECTS solution, bearing in mind only a very limited fleet ever pass along that line, and there is an outline plan to fit locos with ECTS anyway.

My gut feeling is no modern manager has the skill to successfully justify why financially and practically these are better options to action now, than the paralysis let's just do nothing until autumn, and hope it doesn't happen again in next few decades. In other words the risk of repeat with its negative financial implications has worse cost-benefit number that upgrading now to provide a solution.

To be honest, if the money was available, the long term solution to the unstable / clay embankments is to build a new formation next to the existing line of route. Then once the (new) line is open, the old unstable / clay embankments can be bulldozed.

During this work, the line can be doubled where needed.

But as the money is not available, this is just a dream...

ETCS/ERTMS will not happen quickly, but yes, that will be the future signalling system. However bringing that forward ahead of the long term plan would cost far more than alterations to the existing signalling system.

In the mean time, it is possible that the existing signalling system could be altered to extend existing loops / double track sections without being excessively expensive.

And adding one or two new loops / double track sections is also possible. But the costs do very much depend on the actual requirements and the type of the existing signalling system.

For example, how much spare communications capacity is there with the existing system? Running a lot of new cables for long distances or providing lots of new equipment (to maintain the existing signalling functions as well as for the new signals, new points and their control systems) eats up money.

Alas, IMHO, under Network Rail, signalling alterations appear (to me) to cost more compared to how BR went about it. But, standards have changed. There are far more procedures and restrictions and "red tape" (some of which has nothing to do with safety).

 

[Brubulus]

I

To be honest, if the money was available, the long term solution to the unstable / clay embankments is to build a new formation next to the existing line of route. Then once the (new) line is open, the old unstable / clay embankments can be bulldozed.

During this work, the line can be doubled where needed.

But as the money is not available, this is just a dream...

ETCS/ERTMS will not happen quickly, but yes, that will be the future signalling system. However bringing that forward ahead of the long term plan would cost far more than alterations to the existing signalling system.

In the mean time, it is possible that the existing signalling system could be altered to extend existing loops / double track sections without being excessively expensive.

And adding one or two new loops / double track sections is also possible. But the costs do very much depend on the actual requirements and the type of the existing signalling system.

For example, how much spare communications capacity is there with the existing system? Running a lot of new cables for long distances or providing lots of new equipment (to maintain the existing signalling functions as well as for the new signals, new points and their control systems) eats up money.

Alas, IMHO, under Network Rail, signalling alterations appear (to me) to cost more compared to how BR went about it. But, standards have changed. There are far more procedures and restrictions and "red tape" (some of which has nothing to do with safety).

I think the solutions you propose are quite extreme. The most likely solution is mechanical reinforcement and stabilization of the embankments involved. This will obviously require a program of closures and expense, but is cheaper than the other options.

 

[pwatkin]

So, we're a long way from July now, and the embankments must be sufficiently very wet by now. Trains are still running every 2 hours in the evenings from Exeter instead of every hour. Does anyone know what the Engineering justification of the continuing restriction is please?

 

[brad465]

So, we're a long way from July now, and the embankments must be sufficiently very wet by now. Trains are still running every 2 hours in the evenings from Exeter instead of every hour. Does anyone know what the Engineering justification of the continuing restriction is please?

Pretty sure it's returning to normal timetable this weekend? The long time I believe was exacerbated by unit and staff shortages.