Riviera Line improvements

[MarkyT]

The terminal bay layout does solve that issue but if trains are wanting to regularly work in and out of Goodrington you'd really rather have the loop if you can somehow keep it for operational reasons.

Agreed, for a half hourly local service extension to Goodrington the single line stub should be short and only contain the terminus platform.

Ah, I didn't realise it was that tight even for a very short ROL or even PROL.

The minimum ROL length under modern standards is ~46m (50yds). The overlap/clearance IBJs might move a little closer to the points, but you'd still lose about a coach length of standage in both platforms if the signals were moved. That ~15m lost at resignalling was also critical for standage.

That's an idea. It's an extra couple of SEU in the costs but preserves the flexibility of both platforms without imposing the overlap timeout restriction (although the overlap timeout may not be that critical if the departure to Goodrington can always be signalled first.

The problem is the conflict compounds further delay in the event of late running. If a down arrival is a few minutes late, you might hold an opposing movement, but say you can't let the late runner into the yard, because it's unexpectedly full due to an earlier failure. So you must let the empty out first but that must wait for the timeout before it can even start moving towards the other platform. Then that slow movement must complete before the original late runner can continue on to Goodrington. Then you end up checking the next down arrival following close behind, as the whole sequence has taken so long.

The intermediate signals could be placed to stand a 5 car IET clear of the north crossing. They could be used to admit a second train to the platform under a warning aspect rather than a call on with the forward ROL occupied by a short train (permitted in platforms).

 

[devon_belle]

Electrification feels like an obvious big win, as well as more bi-di and the second crossover at Paignton. The GWR was planning OHLE west of Taunton in the 1930s to tackle the banks. It would also speed up stoppers through better acceleration, possibly even making room for an Exminster station stop without a dynamic loop. I don't know how big of an issue freight is, but having to split trains over the banks doesn't feel like an efficient use of paths.

 

[brad465]

Electrification feels like an obvious big win, as well as more bi-di and the second crossover at Paignton. The GWR was planning OHLE west of Taunton in the 1930s to tackle the banks. It would also speed up stoppers through better acceleration, possibly even making room for an Exminster station stop without a dynamic loop. I don't know how big of an issue freight is, but having to split trains over the banks doesn't feel like an efficient use of paths.

With the rise of battery potential, I don't see full electrification happening, especially as the rough sea wall can be left unelectrified. Electrification would likely cover Exeter northwards, then Newton Abbott-Plymouth to cover the Devon Banks, possibly also Paignton-Torquay to assist climbing out of the former.

 

[sharpener]

What about the line from Newton (well, Aller Jct) onwards to Plymouth? That section is twisty, undulating and slower than that between Exeter and Newton. Is there any benefit to new faster alignments (inc longer tunnels) between Aller and Totnes, and between Totnes and Ivybridge (or even Plympton) (abandoning the old alignments)? Is that way too expensive? Can the existing alignment be sped up cost-effectively with electrification and perhaps smoothing a few corners?

How helpful would tilting 802s (like the Pendolinos on the WCML) be here to deal with the curves?

Curvature definitely restricts speed in many parts of Devon and Cornwall. Dainton is 55mph on both ramps, with curves down to ~360m radius. Much of Totnes to Plymouth is 60mph with curves down to ~400m. The sea wall is also constrained by curves.

Thanks for that. The line from N Abbot to Totnes goes up a tributary of the Teign, through Dainton tunnel at the watershed and then comes down a trib. of the Dart so not a lot of scope for alternative alignments, the A381 in comparison is terrible! I can't see a route where a longer tunnel would help as ironically that bit is fairly straight.

The subsequent question was would tilting trains help. I am guessing in principle, yes, now the gradients are not a problem.

I am less familiar with the onward stretch to Plymouth but looking at the map the only question that occurs to me is why does the line loop a mile to the North to make the crossing over the R Yealm?

ETA it seems to be bc it is following (approx) the 100m contour - though with a fair amount of cut and fill to avoid the worst of the wiggles. Presumably the extra mileage was better than losing height.

 

[MarkyT]

Thanks for that. The line from N Abbot to Totnes goes up a tributary of the Teign, through Dainton tunnel at the watershed and then comes down a trib. of the Dart so not a lot of scope for alternative alignments, the A381 in comparison is terrible! I can't see a route where a longer tunnel would help as ironically that bit is fairly straight.

The subsequent question was would tilting trains help. I am guessing in principle, yes, now the gradients are not a problem.

I am less familiar with the onward stretch to Plymouth but looking at the map the only question that occurs to me is why does the line loop a mile to the North to make the crossing over the R Yealm?

ETA it seems to be bc it is following (approx) the 100m contour - though with a fair amount of cut and fill to avoid the worst of the wiggles. Presumably the extra mileage was better than losing height.

View attachment 210265

I asked Google AI a series of questions about the performance of IETs on steep banks. I include one answer in full as it's fascinating. I've not checked the figures, but the conclusions seem reasonable:
Me:

On diesel, what is the time and distance to reach 55mph on a rising gradient of 1:40?

Google:

Under pure diesel power, a UK Hitachi Intercity Express Train (IET) starting from rest on a continuous 1:40 rising gradient will never reach 55 mph because it hits its absolute physical balancing speed at approximately 51 to 52 mph. A gradient of 1:40 (such as the infamous steep sections in Devon and Cornwall) exerts an immense gravitational resistance (0.245m/s^2) against the vehicle. When combined with the lower power-to-weight ratio of the underfloor MTU diesel engines in diesel mode, the train runs out of spare tractive effort to push beyond 52 mph.


Performance Limits under Diesel Power on a 1:40 Grade
Because it cannot reach 55 mph, the performance figures up to its maximum ceiling are broken down below:
To reach 30 mph (48 km/h): Time: ~33 seconds, Distance: ~240 metres
To reach 45 mph (72 km/h): Time: ~75 seconds, Distance: ~780 metres
To reach 50 mph (80 km/h): Time: ~145 seconds, Distance: ~1,650 metres (approx. 1 mile)
To reach 52 mph (84 km/h): Time: Indefinite (Acceleration curves completely flatten out) Distance: Becomes its maximum balancing speed

Factors That Alter This Limit
Class 802 vs Class 800:
The Class 802 variants feature higher-rated diesel engines (940 hp vs the original 750 hp on the Class 800). If operating a 5-car Class 802 with all engines functioning at maximum output, the train can just scrape past 55 mph on a 1:40 slope, but it requires over 3 to 4 miles of continuous climbing and several minutes to settle at a balance point of roughly 56–57 mph.

Momentum: In real-world operations (such as conquering Hemerdon Bank), IET drivers do not start from a dead stop at the base of the hill. They hit the gradient at the full line speed (up to 100 mph) and use the train's massive kinetic energy to carry them up and over the summit before the speed decays down to the diesel balancing limit.

So it seems there's little point in speeds above 55mph on Dainton. A non-stop can keep going at 60mph through Newton Abbot station and on to Stoneycombe where the 55 starts. I expect most trains starting from a stand at Newton Abbot can also make that balancing speed before the foot of the bank. I guess this is what astonished me on that first Voyager run. The train accelerated like a rocket out of Newton Abbot and the gradient seemed not to slow it at all once it reached a steady speed, unlike any previous traction experienced. The engines were clearly running at full tilt though.

Although the Devon ascents are gruelling, they work because they're not too long, so the fatigue of extended continuous high output is limited and systems have time to throttle back and cool down on the adjacent easier sections.

The 'Plateau' section between Rattery and Hemerdon banks is not flat but has more typical railway gradients throughout, rarely steeper than 1:100. This section is where speed improvements might make a difference as the expresses have a long non-stop segment and could physically reach higher speeds with their installed power, but curvature imposes the limit here. The line hugs the south fringe of Dartmoor following high contours, and any straightening would be very difficult and expensive, requiring long cutoff tunnels or viaducts.

Here's a ~10km 'base tunnel' route from South Brent to Yondertown. You'd still have to climb Rattery and Hemerdon, both fairly straight already, but the really twisty bits are avoided. With expresses diverted, perhaps local trains might make a few more stops on the old route.

 

[sharpener]

I asked Google AI a series of questions about the performance of IETs on steep banks.

Thanks for that, interesting.

AI certainly had its uses, I asked it for a gradient longitudinal section and this dating from 1899 is what it found for me. Shows how spikey the summits are esp. Dainton.

This Wikipedia entry confirms my OH's experience upthread about ailing HSTs having to skip Totnes to take a run at the banks. It was obviously a well rehearsed process and TOT pax were detrained at Newton Abbot to catch the following local train.

 

[MarkyT]

I asked Google about the electrification of Newton Abbot-Plymouth. It estimated 3 to 5 minutes might be saved by electrification alone, as speed is still constrained so much by curvature. I asked it to compare that with the notional straighter cutoff illustrated in my previous post, combined with modest improvements to existing lines on either side. That produced an overall saving of 8.5-11.5 minutes. The full answer:

Electrification of the line between Newton Abbot and Plymouth using 25kV OHLE would significantly enhance the performance of Great Western Railway's Hitachi IET (Class 802) trains, unlocking up to 4,480 kW of electric power compared to their 2,800 kW diesel limit. This massive power boost would eliminate speed drop-offs on the severe 1-in-36 gradients of the South Devon Banks (Dainton, Rattery, and Hemerdon). [1, 2, 3]
The estimated journey time savings for expresses serving Newton Abbot, Totnes, and Plymouth under the two infrastructure scenarios are detailed below.

Scenario 1: Existing Track Alignment and Layout Remains
Because the current alignment is severely limited by sharp geographic curves, maximum speeds are structurally capped between 40 mph and 75 mph. Performance gains come strictly from rapid acceleration and maintaining line speed uphill. [1, 2]

• • Newton Abbot to Totnes (~9 miles): 1 to 2 minutes saved.
    • Rapid acceleration out of Newton Abbot.
    • Train maintains a flat 40–50 mph line speed up the steep Dainton Bank.
  • Totnes to Plymouth (~23 miles): 2 to 3 minutes saved.
    • High-power ascent up Rattery and Hemerdon banks.
    • Eliminated "diesel lag" when recovering speed after tight curves. [1]
  • Total Section Saving: 3 to 5 minutes.

Scenario 2: New 10km South Brent to Yondertown Cutoff & Upgrades
This scenario completely transforms the route. It replaces a highly winding, low-speed section with a 125 mph straight alignment (largely in tunnel) and raises surrounding sections to 100 mph. [1]

• • Newton Abbot to Totnes: 1.5 to 2.5 minutes saved.
    • Modest track easing allows faster approaches.
    • Higher acceleration out of station stops.
  • Totnes to Plymouth: 7 to 9 minutes saved.
    • Cutoff Line Benefit: Bypassing the old winding route at 125 mph cuts ~4 minutes alone.
    • 100 mph Upgrades: High electric power allows IETs to quickly reach and hold 100 mph on the improved approaches.
    • Hemerdon Bank's steep descent/ascent impact is minimized by the modern alignment. [1]
  • Total Section Saving: 8.5 to 11.5 minutes.

Summary Comparison Table

Journey Segment	Current Typical Time	Scenario 1 (Electric, Old Track)	Scenario 2 (Electric + Cutoff)
Newton Abbot – Totnes	~11–13 mins	9–11 mins (~1.5m saving)	8.5–10.5 mins (~2m saving)
Totnes – Plymouth	~23–26 mins	21–23 mins (~2.5m saving)	15–18 mins (~8m saving)
Total Route	~34–39 mins	~30–34 mins	~23.5–28.5 mins