Excessive amounts of speed restrictions for trams

[K.o.R]

I spent the weekend in Manchester and got to ride the Metrolink for the first time in many years.

What I found very strange was the huge number of speed restrictions. It seemed like anything other than dead straight heavy-rail track, there was a speed restriction down to 40, 30 or even 20mph, when I'm pretty sure the normal trains back in the day would have happily bumbled along at 50mph or even more.

It's not like there's no cant on the track at all. I can understand some of the restrictions when street running, but even then there's a fair few places where it seems like they could have put a larger curve in rather than have the tram crawl along at 5mph.

 

[AndrewE]

It's what you get when much the permanent way is shared street-running, and the new owners/people designing it have the safety-at-all-costs mentality, even in places where it is on old heavy rail alignments.
It's the reason that I oppose conversions: places like Oldham get stuck with very slow journeys with such frequent stops that a bus might beat it to the city centre. I wouldn't bother to go to Altrincham now, I did go when it was a train trip.

 

[Taunton]

Edinburgh the same, any curve seems to need a 10mph or less restriction. At street-running road junctions it's a considerable waste of road capacity as the tram crawls round. Worse is the lengthy final section out to Edinburgh airport, through open fields but which follows the old field boundaries through a series of right-angle turns, each of which requires slowing to walking pace, tediously long because the considerable tram length has to be fully clear before the power goes back on.

It can't be tram profile wheels, as old heritage trams were never like this. Incidentally, I noticed exactly the same on the street running tram in Dubai, same vehicle type.

 

[Tetchytyke]

places like Oldham get stuck with very slow journeys with such frequent stops that a bus might beat it to the city centre

Yep, the tram is way slower than walking through those terrifying underpasses to Mumps Station, waiting up to 30 minutes for a train, sitting on the train for 30 minutes then walking from Victoria at the other end.

As for the OP's question as to why trams go slowly around tight curves, do the words "Croydon" and "Tramlink" ring any bells?

 

[K.o.R]

Yep, the tram is way slower than walking through those terrifying underpasses to Mumps Station, waiting up to 30 minutes for a train, sitting on the train for 30 minutes then walking from Victoria at the other end.

As for the OP's question as to why trams go slowly around tight curves, do the words "Croydon" and "Tramlink" ring any bells?

Yes, but that was an example of an extremely tight curve that couldn't have any cant assistance due to it being right on top of another junction, and couldn't be any more gentle because of obstruction, and an extreme overspeed incident.

The occurrences I am talking about are either already slow due to being on-street sections, or overly-conservative speed limits on heavy rail sections that are nowhere near comparable to the Croydon example.

 

[edwin_m]

Yes, but that was an example of an extremely tight curve that couldn't have any cant assistance due to it being right on top of another junction, and couldn't be any more gentle because of obstruction, and an extreme overspeed incident.

The occurrences I am talking about are either already slow due to being on-street sections, or overly-conservative speed limits on heavy rail sections that are nowhere near comparable to the Croydon example.

Rather surprisingly the RAIB report doesn't record what the cant was on the curve at Sandilands, at least that I can see on a quick search. It does note that computer modelling of the derailment used the design value for cant on the curve, which rather implies that it wasn't zero, and I certainly recall seeing cant on the similar curve at Coombe Road. It's far enough that the junction wouldn't constrain the amount of cant, but on very tight tramway curves the amount of cant possible is much less than would be implied by the cant-curvature rules on heavy rail. This is true even if off-street, and arises from geometric limitations on simultaneous rotation and twist of the tram running gear, and probably also from the excessive cant deficiency should a tram stop on the curve.

Perhaps you could give us some examples of places on Metrolink where you consider the alignment to be unnecessarily curved or speeds to be unnecessarily slow? Some of us might then be able to respond with possible specific reasons for this (there usually is one, though some might not agree with it).

 

[K.o.R]

I can't really remember specific areas; the branches I travelled on were Altrincham, East Didsbury and Bury, all sections which are basically heavy rail with large sweeping curves that nevertheless have speed restrictions.

There's the exit from Deansgate-Castlefield which is 5mph on a sharp but not overly-so curve, and some instances on the Airport branch where the track went across a road junction on a curve that was fairly big but could probably still have been bigger.

 

[507 001]

In segregated areas, 9 times out of 10 any speed restrictions have nothing to do with track geometry or wheel profiles. It’s because most of Metrolink runs on line of sight these days.

The Altrincham line is 50mph for the vast majority of the distance from old Trafford to Timperley, apart from a short stretch approaching Stretford This is where the otherwise straight and flat route goes round a left hand curve that you can’t really see around.

The Bury line used to be 50mph all the way from collyhurst tunnel to Bury. The need for more capacity (because the line is so overcrowded) resulted in it needing to be converted to LoS, and so it had to slow down because of its twisty, hilly nature.

The Rochdale line is actually not that much slower than it was pre conversion. Ok, you’ve lost the semi-fasts but you’ve gained connectivity into the town centres and Manchester City centre. The fact that ridership is so much higher than it was pre conversion suggests we’re doing something right.

The curves mentioned on the airport line for example (and there are similar on the Ashton and Rochdale street running areas), where we cross roads are actually minimum radius curves for an M5000. There are a couple of other minimum radius curves on phase 1 and 2, but the regulations for speeding these had changed (and the geometry isn’t exactly the same in terms of cant etc) which means that these are usually 8mph, and often 5mph unfortunately.

Are some of the restrictions a little overzealous? Probably. Did I prefer driving under TCB? Yes, undoubtedly.

It is what it is though, and to provide the capacity needed there’s not a huge amount we can do.

 

[the sniper]

I wonder if we could see ERTMS/ETCS (or something similar/lighter) fitted on some systems, particularly the likes of the Bury line? I feel a little like line of sight driving off street is only a Sandilands scale incident away from coming under scrutiny, regardless of the restrictions that are inherent with it.

 

[507 001]

I wonder if we could see ERTMS/ETCS (or something similar/lighter) fitted on some systems, particularly the likes of the Bury line? I feel a little like line of sight driving off street is only a Sandilands scale incident away from coming under scrutiny, regardless of the restrictions that are inherent with it.

I doubt it. Certainly not ETCS.

The rumour is that TfGM are looking at options for an in cab signalling system to replace TMS, but by all accounts it would still be a predominantly line of sight based system.

 

[Roast Veg]

I wonder if we could see ERTMS/ETCS (or something similar/lighter) fitted on some systems, particularly the likes of the Bury line? I feel a little like line of sight driving off street is only a Sandilands scale incident away from coming under scrutiny, regardless of the restrictions that are inherent with it.

In another thread the use of Tesla Autopilot for London Overground was discussed. Several posters, including myself, believe it would be more appropriate as a scheme for trams.

 

[507 001]

In another thread the use of Tesla Autopilot for London Overground was discussed. Several posters, including myself, believe it would be more appropriate as a scheme for trams.

No thanks.

 

[WatcherZero]

Since one has just driven into a plane, doubtful that Tesla's system is anywhere near ready, some of the other auto manufacturers systems have leapfrogged them and are considered more reliable.

 

[Starmill]

I doubt it. Certainly not ETCS.

The rumour is that TfGM are looking at options for an in cab signalling system to replace TMS, but by all accounts it would still be a predominantly line of sight based system.

Communications based train control is about the only option on the market for something like this at the moment. The reality is that is simply far too expensive for a system of this nature and would be a 'challenge' to interface to the line of site sections.

As has been ably explained upthread the lowering of the speed increased capacity significantly. It means trams can now run much closer together than before the signalling was updated (except for the short section under Network Rail control).

The very tight curves are a total nightmare. They were done partly to minimise on land take (i.e. a cheap solution) and partly to try to move the trams out of conflicts with general traffic. However, the 5 mile / hour restrictions are so painful that I always think it'd have been easier to just put up with slightly more traffic. The curves being so tight also costs more in wheelset maintenance. The good thing is that this lesson was learnt a bit, and the Trafford Park line was amended at a late stage to remove some of the curves. I don't know if it has any 5 curves still though.

== Doublepost prevention - post automatically merged: 24 Apr 2022 ==

In another thread the use of Tesla Autopilot for London Overground was discussed. Several posters, including myself, believe it would be more appropriate as a scheme for trams.

I confess I initially thought this was an April Fool.

 

[K.o.R]

In segregated areas, 9 times out of 10 any speed restrictions have nothing to do with track geometry or wheel profiles. It’s because most of Metrolink runs on line of sight these days.

The Altrincham line is 50mph for the vast majority of the distance from old Trafford to Timperley, apart from a short stretch approaching Stretford This is where the otherwise straight and flat route goes round a left hand curve that you can’t really see around.

The Bury line used to be 50mph all the way from collyhurst tunnel to Bury. The need for more capacity (because the line is so overcrowded) resulted in it needing to be converted to LoS, and so it had to slow down because of its twisty, hilly nature.

The Rochdale line is actually not that much slower than it was pre conversion. Ok, you’ve lost the semi-fasts but you’ve gained connectivity into the town centres and Manchester City centre. The fact that ridership is so much higher than it was pre conversion suggests we’re doing something right.

The curves mentioned on the airport line for example (and there are similar on the Ashton and Rochdale street running areas), where we cross roads are actually minimum radius curves for an M5000. There are a couple of other minimum radius curves on phase 1 and 2, but the regulations for speeding these had changed (and the geometry isn’t exactly the same in terms of cant etc) which means that these are usually 8mph, and often 5mph unfortunately.

Are some of the restrictions a little overzealous? Probably. Did I prefer driving under TCB? Yes, undoubtedly.

It is what it is though, and to provide the capacity needed there’s not a huge amount we can do.

Thank you for the very detailed explanation!

 

[507 001]

Communications based train control is about the only option on the market for something like this at the moment. The reality is that is simply far too expensive for a system of this nature and would be a 'challenge' to interface to the line of site sections.

As has been ably explained upthread the lowering of the speed increased capacity significantly. It means trams can now run much closer together than before the signalling was updated (except for the short section under Network Rail control).

The very tight curves are a total nightmare. They were done partly to minimise on land take (i.e. a cheap solution) and partly to try to move the trams out of conflicts with general traffic. However, the 5 mile / hour restrictions are so painful that I always think it'd have been easier to just put up with slightly more traffic. The curves being so tight also costs more in wheelset maintenance. The good thing is that this lesson was learnt a bit, and the Trafford Park line was amended at a late stage to remove some of the curves. I don't know if it has any 5 curves still though.

== Doublepost prevention - post automatically merged: 24 Apr 2022 ==


I confess I initially thought this was an April Fool.

Bombardier did do various levels of CBTC, including a version designed for street running, although I confess I have no idea if Alstom are continuing with this.

The lowest line speed on the TPL is 10mph. As I said in my previous post, we have other 25m radius curves that the M5000s are quite happy going round at 10* (high st to Market st, south langworthy rd into Langworthy OB) but the testing regime changed between phase 2 and 3 which meant lower speeds on the phase 3 curves. It wouldn’t have mattered about them being a cheap solution if they had been tested as they would have been during phase 1 & 2.

*well, as happy as an M5000 can get.

 

[geoffk]

I used to use Metrolink a lot. "Line of sight" running will be the reason in many cases but also facing points (at either a junction or crossover) tend to have a severe speed restriction, whichever route is taken. I believe that is because, unlike railway points, they are not locked so could move under the tram. Examples are the facing crossover at Shaw and places where single track becomes double - approaching Timperley from Altrincham and Newbold from Rochdale.

 

[507 001]

I used to use Metrolink a lot. "Line of sight" running will be the reason in many cases but also facing points (at either a junction or crossover) tend to have a severe speed restriction, whichever route is taken. I believe that is because, unlike railway points, they are not locked so could move under the tram. Examples are the facing crossover at Shaw and places where single track becomes double - approaching Timperley from Altrincham and Newbold from Rochdale.

They are locked, but the absence of any type of train protection system under line of sight means that the risk of a collision is higher.

 

[geoffk]

They are locked, but the absence of any type of train protection system under line of sight means that the risk of a collision is higher.

OK thanks.

 

[edwin_m]

They are locked, but the absence of any type of train protection system under line of sight means that the risk of a collision is higher.

However, I think TMS which does these functions has a lower safety integrity level than railway signalling, so a "wrong side failure" is less unlikely.

 

[507 001]

However, I think TMS which does these functions has a lower safety integrity level than railway signalling, so a "wrong side failure" is less unlikely.

Not necessarily. It operates differently depending on where it is. So if it’s in a street running area then yes, you’re mostly quite correct. However points are still locked and proven to be so, they just don’t interlock with the signals.

In segregated areas, then it’s a different story, especially in the last couple of years where it’s been beefed up considerably. It is still designed to fail right side, and a wrong side failure in a segregated section would be taken very seriously.

 

[edwin_m]

Not necessarily. It operates differently depending on where it is. So if it’s in a street running area then yes, you’re mostly quite correct. However points are still locked and proven to be so, they just don’t interlock with the signals.

In segregated areas, then it’s a different story, especially in the last couple of years where it’s been beefed up considerably. It is still designed to fail right side, and a wrong side failure in a segregated section would be taken very seriously.

Is it a high-integrity signal interlocking (Safety Integrity Level 4) or an industrial grade programmable logic controller (SIL 2 I think) like on the street sections?

 

[507 001]

Is it a high-integrity signal interlocking (Safety Integrity Level 4) or an industrial grade programmable logic controller (SIL 2 I think) like on the street sections?

Couldn’t tell you if I’m honest, but it does work in a completely different way to the street running sections.