When will Sydney track speeds get back up to 130-160km/h

[pway_master]

Hi all (new member here),

Recently got access to pre-historic (circa 2003) route diagrams courtesy of an old thumb drive containing the SRA Track Atlas......I was disappointed to find track speeds much lower today than they were 20 years ago (examples of East Hills & St Mary's shown below).

Left (circa 2003), Right (circa 2023) Glenfield (Down 160 is now 115)

Left (circa 2003), Right (circa 2023) East Hills (Up 160 is now 115, Up 140 is now 115)

Left (circa 2003), Right (circa 2023) Penrith (Down Main 130 in now 115)

Left (circa 2003), Right (circa 2023) St Marys (Down Main 160 in now 115)


Why?, after more than 15-20 years TfNSW/Syd Trains still haven't returned the only decent stretches of track back to 160km/h. This has puzzled me to no-end, considering the amount of money spent on resignalling, track renewals etc... that we still have the blanket 115km/h limits posted around Sydney.

With the NIF's coming online there will be much more rolling stock capable of operating many services to the prior speeds (capped at 130), which imho is an inherent network capacity that should not be overlooked.

Just my two cents.

Interested to hear any stories or answer as to why this en-masse slowing of the Sydney Trains network happened, & whether or not anything will ever be done about it.

[Transtopic]

Hi all (new member here),

Recently got access to pre-historic (circa 2003) route diagrams courtesy of an old thumb drive containing the SRA Track Atlas......I was disappointed to find track speeds much lower today than they were 20 years ago (examples of East Hills & St Mary's shown below).

Left (circa 2003), Right (circa 2023) Glenfield (Down 160 is now 115)

Left (circa 2003), Right (circa 2023) East Hills (Up 160 is now 115, Up 140 is now 115)

Left (circa 2003), Right (circa 2023) Penrith (Down Main 130 in now 115)

Left (circa 2003), Right (circa 2023) St Marys (Down Main 160 in now 115)


Why?, after more than 15-20 years TfNSW/Syd Trains still haven't returned the only decent stretches of track back to 160km/h. This has puzzled me to no-end, considering the amount of money spent on resignalling, track renewals etc... that we still have the blanket 115km/h limits posted around Sydney.

With the NIF's coming online there will be much more rolling stock capable of operating many services to the prior speeds (capped at 130), which imho is an inherent network capacity that should not be overlooked.

Just my two cents.

Interested to hear any stories or answer as to why this en-masse slowing of the Sydney Trains network happened, & whether or not anything will ever be done about it.

This is a bugbear of mine as well. The previous speed limits of up to 160km/h on the then CityRail network clearly demonstrates that the track standards at the time were capable of those speeds and that doesn't even take into account track upgrades since. I can recall travelling on the XPT to Bathurst when it must have been travelling at 160km/h on the Western Line to Penrith.

The reduction to a blanket 115km/h in the Sydney metro region seems to have arisen from the Waterfall accident inquiry and the limitations of the current signalling technology, rather than the track itself which is capable of higher speeds. The introduction of digital signalling, combined with ATO, should allow for the higher speed limits to be reinstated. The latest rolling stock, including Millenniums, Oscars and Waratahs have a maximum design speed limit of 130km/h and the NIF and new regional fleet are designed for 160km/h.

[tonyp]

This is really only about long distance expresses - and is a justifiable discussion in that regard. For urban services with regular stops, average speed (determined mostly by acceleration/deceleration and stop dwells) is far more significant. Distances between station stops will determine the maximum line speed that the train can reach before it has to start decelerating again. No stopping suburban train is going to reach 130, let alone 160 km/h. Evidence of this is seen in the metro operation where significantly quicker journey times are achieved (comparing on the basis of same distance/ number of stops) with a maximum speed of only 100 km/h.

Even for long distance trains, what is a faster speed across the Sydney basin going to achieve? A few minutes, before the train runs onto the speed-restricting snake-like alignments beyond the Cumberland Plain and the northern and southern plateaux?

[Randomness]

A few minutes also makes a difference if disruption occurs and stops are skipped to try getting things back on time. The feeling of speed also influences public transport usage as even if a route is slower than if using a car, a user may feel more elated to not be stuck in traffic.

The UK usually does 160km with 3 aspect signalling and 200 with 4 aspect. With our 4 aspects in Sydney it should be entirely safe to run at these speeds.

[tonyp]

A few minutes also makes a difference if disruption occurs and stops are skipped to try getting things back on time. The feeling of speed also influences public transport usage as even if a route is slower than if using a car, a user may feel more elated to not be stuck in traffic.

The UK usually does 160km with 3 aspect signalling and 200 with 4 aspect. With our 4 aspects in Sydney it should be entirely safe to run at these speeds.

The feeling of speed is muted in a modern, sealed air conditioned train. If you want a feeling of speed, you used to get that in the old interurban U boats which felt like they were doing 500 km/h from the amount of noise! There's limited opportunity in Sydney for "catching up" anyway - between Blacktown and Penrith and parts of the Campbelltown route - until one gets stuck behind another train.

[BAMBAM]

All suburban trains follows the lowest speed sign. So technically, trains never went 160. Only trains that do are either intercity or regional trains.

[pway_master]

All of the points raised are valid & make sense. I understand that projects are completed to the standard of the day. I'm hoping this thread will point out to more people that increasing some speeds is not all that far-fetched (however no current projects appear to address this eg. TfNSW MTMS doesn't mention track speeds AFAIK.)

All suburban trains follows the lowest speed sign. So technically, trains never went 160. Only trains that do are either intercity or regional trains.

This was certainly the case prior, however the latest NSG604 speed sign rule appears to allow the 4th generation EMU's to operate to the white speed boards. Meaning the the speeds that were lowered (after 2003) are now limiting services arbitrarily in regions that were > 115km/h prior.

Latest NSG604 speed sign rule (Railsafe website, 2024)

A few minutes also makes a difference if disruption occurs and stops are skipped to try getting things back on time. The feeling of speed also influences public transport usage as even if a route is slower than if using a car, a user may feel more elated to not be stuck in traffic.

The UK usually does 160km with 3 aspect signalling and 200 with 4 aspect. With our 4 aspects in Sydney it should be entirely safe to run at these speeds.

This is where my bugbear now lies with regards to the East Hills corridor which was the subject of significant investment *(eg. providing the outer tracks between Revesby & Wolli Creek). A majority of T8 city services trundle along at 110-115km/h non-stop on tracks that were initially specified as at least 130km/h (K2RQ project specification). This failure of the project to deliver these speeds is something that should be pointed out.

Compared with driving in the peak hours the T8 line eg. could provide a much more attractive journey which is faster than the M5. Also, once the legacy sets (K & V set) are removed from service, there is no reason why all services should not operate at 130km/h max where practical (any contrary opinions would be welcome for discussion). I'm sceptical that the ATP/ETCS project in Sydney would bother to look at improving speeds unless they are explicitly specified included into the project outcomes.

Hopefully we will be provided with some insight in the near future. fyi. I spoke to a friend in the pway design world who is eager to get speeds back up, but is lacking the support from projects to get any improvements funded.

[Glen]

Simple answer, it's become a very conservative railway because that's easier.

[tonyp]

The difference between 115 and 130 km/h might add up to something on a long, straight country line, but between Revesby and Sydenham, which is the longest (16 km) express section on that line, it would achieve no more than 1 to 2 minutes. Another thing to consider is the effect of manual driving. In my observation, Sydney drivers typically do a lot of on and offing while driving, rather than holding line speed exactly. ATP/ETCS might address this, like the metro which can hold line speed precisely.

As I'm always saying, too much is made of maximum speed in urban transit. The most important issue in shortening journey time is average speed and this is determined by acceleration/deceleration, ability to reach line speed quickly and hold to it, dwell time and number of stops. Unless you're talking about 160 km/h, which can pretty-much never be reached in an urban operation, maximum speed is inconsequential.

[Glen]

Sydney train drivers do so much offing because if they didn't they'd get to each station even earlier.

All of the components you mentioned used to be better in Sydney before the network took a giant valium two decades ago, for reasons no one has ever been able to explain logically.

[BAMBAM]

This was certainly the case prior, however the latest NSG604 speed sign rule appears to allow the 4th generation EMU's to operate to the white speed boards. Meaning the the speeds that were lowered (after 2003) are now limiting services arbitrarily in regions that were > 115km/h prior.

While the rule does say that, There was a general instruction to ALL Sydney Trains Drivers (Suburban) that they are to operate to the lower speed board, so if there’s only Yellow and White speed board, they’ll be operating to yellow speed board. If there was Yellow, Blue and white, they’ll be operating to Blue speed boards.

[boxythingy]

The trains travelling at the near maximum allowable speeds would be non-revenue services anyway. They don't make it easy for Customers to benefit from high speeds that easily in Sydney to make the overall journey from the suburbs into the CBD faster.

How familiar are you with the actual old timetables? Ex Macarthur services for example used to stop at Padstow and Riverwood as the standard all stops pattern for most off-peak services and the Sydenham express services stopped at both Beverly Hills and Kingsgrove instead of St Peters. The current timetable last revised as a major timetable change is deplorable with cross platform changes at Revesby to access the stops between Padstow to Turrella, with every second off peak ex Macarthur arriving onto Platform 1 service timed for the all stops ex Revesby service to depart from Platform 3 within 1min or at the same time

[tonyp]

The trains travelling at the near maximum allowable speeds would be non-revenue services anyway. They don't make it easy for Customers to benefit from high speeds that easily in Sydney to make the overall journey from the suburbs into the CBD faster.

Except that higher maximum speeds in an urban commuter context don't actually benefit journey time. Higher average speeds do that. Look at the metro - striking differences in journey times compared to suburban, yet limited to 100 km/h. It's very hard to shake this myth, it has a hold on some people's minds!

[pway_master]

While the rule does say that, There was a general instruction to ALL Sydney Trains Drivers (Suburban) that they are to operate to the lower speed board, so if there’s only Yellow and White speed board, they’ll be operating to yellow speed board. If there was Yellow, Blue and white, they’ll be operating to Blue speed boards.

Yep, a driver friend has confirmed this is & sent me a screenshot that an instruction was issued to all Sydney Trains Drivers.
.

Now, with the amalgamation of NSWTrainLink drivers into Sydney Trains coming up this year...... will the Oscars & NIF's become limited to 115km/h as well, I suspect there is a chance this may occur .

The trains travelling at the near maximum allowable speeds would be non-revenue services anyway. They don't make it easy for Customers to benefit from high speeds that easily in Sydney to make the overall journey from the suburbs into the CBD faster.

Fair point, however I did catch an packed express peak service yesterday between Sydenham & Glenfield which only stopped at Holsworthy. In the 20 or so minutes non-stop we hit 115km/h just after Bexley & kept that rate until Holsworthy. Now if the existing track, signalling, rolling stock & overheads are capable of 130km/h (as evidenced by the old diagrams) what is the valid excuse for not allowing my train to run at 130km/h (apart from ex-Railcorp prejudices). This is apart from trying to achieve any direct time gains, rather just trying to discern why 115km/h is deemed acceptable but 130km/h (or above) is not?

[boxythingy]

Fair point, however I did catch an packed express peak service yesterday between Sydenham & Glenfield which only stopped at Holsworthy. In the 20 or so minutes non-stop we hit 115km/h just after Bexley & kept that rate until Holsworthy. Now if the existing track, signalling, rolling stock & overheads are capable of 130km/h (as evidenced by the old diagrams) what is the valid excuse for not allowing my train to run at 130km/h (apart from ex-Railcorp prejudices). This is apart from trying to achieve any direct time gains, rather just trying to discern why 115km/h is deemed acceptable but 130km/h (or above) is not?

I often catch trains on this line, and yes it looks like current timetabling means 115km/h or so can be achieved but if you were to look at some of the modified timetables on weekends during trackwork, you'll find there is so much dwell time (arr/dep) built into the timetable at stops like Glenfield, Sydneham etc. During delays, you can expect the trains to be moving below 40km/h along the same stretch of rail.

Except that higher maximum speeds in an urban commuter context don't actually benefit journey time. Higher average speeds do that. Look at the metro - striking differences in journey times compared to suburban, yet limited to 100 km/h. It's very hard to shake this myth, it has a hold on some people's minds!

Which brings us to this point which is exactly valid, a train or any vehicle travelling at the maximum speed limit isn't the only determining factor. The train can hit 200km/h like a hoon going down the street on a quiet Sunday afternoon for a few secs and then just run at 20km/h on other sections.

It really needs to be an efficient combination of the right mix of stopping patterns based on regular travel patterns, picking up passengers at the most deserving stops without preferencing certain "major" stops to preventing sudden influxes and overcrowding.

I am glad some trains, but not enough ex Macarthur trains for example skip Revesby, even though it is clearly served with plenty of empty running all stops trains.

Perhaps make some limited stopping patterns on that section of line e.g. Revesby straight to Kingsgrove then all stops to the Airport, or Revesby then all stops to Narwee and then express to all stops through the Airport, or whatever makes sense where the points are allowing trains to make extended 1-2min stops at most to overtake etc. and able to pick up an above average speed again to make up for the "lost" time. This is what some overseas railway systems seem to be capable of doing or executed with so much more elegance.

[Linto63]

I am glad some trains, but not enough ex Macarthur trains for example skip Revesby, even though it is clearly served with plenty of empty running all stops trains.

You do realise that the express services call at Revesby to allow passengers from west of there to change there for stations to the east. Without the Revesby stop to get from Panania to Padstow would require travelling all the way to Wolli Creek to change, a 27 kilometre round trip for stations that are 3 kilometres apart.

[Transtopic]

The thing that is overlooked in a comparison of the average speeds of the DD and metro trains is that the Waratahs for example have a design specification for acceleration/deceleration of 1.0 m/s/s and the Alstom metro trains of 1.1 m/s/s and 1.3 m/s/s for emergency braking. The Waratahs have been geared down to 0.8 m/s/s but can be upgraded to their original design specification. Putting aside dwell times, the Waratahs are capable of a much higher performance in an all stations stopping pattern than suggested and even more so with longer distance express services with their higher maximum speed. The performance will be further enhanced with the digital signalling and ATO upgrade which will no longer rely on driver operation other than starting the train from the station, monitoring the system and taking over manual control in an emergency.

[boxythingy]

The performance will be further enhanced with the digital signalling and ATO upgrade which will no longer rely on driver operation other than starting the train from the station, monitoring the system and taking over manual control in an emergency.

Is this increased performance expected as part of the signalling upgrades in the Airport Tunnel under the More Trains, More Services program? Sick and tired of travelling on of the slow and sluggish all stops services all the way back Macarthur way as early as from 7pm Sundays.

[tonyp]

Is this increased performance expected as part of the signalling upgrades in the Airport Tunnel under the More Trains, More Services program? Sick and tired of travelling on of the slow and sluggish all stops services all the way back Macarthur way as early as from 7pm Sundays.

If this elevation to the standard of metro running performance is somehow achieved (which they haven't managed to on the Paris RER A double-deck operation, but of course we know we're more competent than the French), they will also have to complete the station dwells within 25-30 seconds to totally attain a similar standard. That should be an interesting exercise with two-door double deck trains, human guards and no PSDs. Don't hold your breath.

[Glen]

The performance will be further enhanced with the digital signalling and ATO upgrade which will no longer rely on driver operation other than starting the train from the station, monitoring the system and taking over manual control in an emergency.

Is this increased performance expected as part of the signalling upgrades in the Airport Tunnel under the More Trains, More Services program? Sick and tired of travelling on of the slow and sluggish all stops services all the way back Macarthur way as early as from 7pm Sundays.

The all stations weekend evening service is simply a way of saving crew hours.

[Glen]

If this elevation to the standard of metro running performance is somehow achieved (which they haven't managed to on the Paris RER A double-deck operation, but of course we know we're more competent than the French)

I haven't looked up the Paris RER Line A for several years but when last I did they had a pretty impressive throughput on their timetable.

Mind you that's with 10 car DD's and three doors per car.

How's it going in practice?

[tonyp]

I haven't looked up the Paris RER Line A for several years but when last I did they had a pretty impressive throughput on their timetable.

Mind you that's with 10 car DD's and three doors per car.

How's it going in practice?

The throughput of RER A is great, because they have so many people to move (at least 300 million a year, which is like the entire Sydney Trains system!), but the 2 minute headways don't last too long before they drop back a little for breathers and/or miss a couple of stops to catch up. It's a tough gig and they don't manage it with ease, but the French railways are very capable.

However, it's the journey time we're discussing and in that regard, the Paris operation is closer to Sydney Trains than Sydney Metro. Indeed it shows that automatic operation won't speed up a heavily-used double deck service that much.

I've done comparisons in the past on the basis of similar pretty straight and level track profiles, same distance, same or similar number of stops. Here are a couple of the results. Bear in mind that the Sydney suburban examples typically include long express sections, whereas the Sydney Metro and RER A are all stops. So the Sydney suburban services have differing stopping patterns, but I've chosen examples that have the same number of intermediate stops as the other two systems.

58 km

Sydney suburban Emu Plains- Central: 21 stops, 75 minutes, average speed 46 km/h.

Paris RER A Marne-La-Vallee - St-Germaine: 22 stops, 72 minutes, 49 km/h.

Sydney Metro: Tallawong-Marrickville: 20 stops, 62 minutes, 57 km/h.

63 km

Sydney suburban Emu Plains- North Sydney: 23 stops, 86 minutes, 44 km/h.

Paris RER A Boissy-St-Leger - Cergy-Le-Haut: 24 stops, 80 minutes, 47 km/h.

Sydney Metro Tallawong-Canterbury: 23 stops, 68 minutes, 58 km/h.

So RER A's strength is its ability to move large patronage at close frequency, but with a pretty ordinary journey time. RER A represents what Sydney's suburban operation strives to be with more automation and better signalling, yet it doesn't achieve much improvement in journey time. This suggests to me that this is an inherent weakness of double deck trains, something that an average blind freddy and his dog has known for years but somehow doesn't find its way past the rose-coloured glasses of double deck enthusiasts. Double deckers have their role on long distance semi-express services where they don't have to stop so much. For stopping services, they're not so good.

We've struck the right balance with Sydney Metro - much higher capacity, capable of 2 minute headways and with an average speed at least some 10 km/h higher than that of the other two operations. By the way, this also provides an insight into the other question here - will higher speed limits improve journey time? The metro is theoretically the tortoise here, limited to 100 km/h, compared to the hares on suburban at 115 and RER A at 120. Isn't there an old fable about the hare and the tortoise? As I keep saying like a broken record, it's average speed that's significant.

[Transtopic]

The performance will be further enhanced with the digital signalling and ATO upgrade which will no longer rely on driver operation other than starting the train from the station, monitoring the system and taking over manual control in an emergency.

Is this increased performance expected as part of the signalling upgrades in the Airport Tunnel under the More Trains, More Services program? Sick and tired of travelling on of the slow and sluggish all stops services all the way back Macarthur way as early as from 7pm Sundays.

The current signalling upgrade to the Airport Line to 18 tph is based on the existing signalling technology. There may not be any immediate benefit to journey times until the digital signalling and ATO upgrades have been completed.

Siemens Mobility has already been awarded contracts for the first stage of the digital signalling and ATO upgrade on T4, including a new Traffic Management System, which will be integrated with existing Sydney Trains systems. The clearly stated objectives of the contracts are to increase frequencies to 24 tph per line as well as to reduce journey times. How the reduction in journey times is achieved is irrelevant. as it would be a combination of automation, better acceleration/deceleration and closer headways with the digital signalling, resulting in higher average speeds, which tonyp is always prattling on about.

[Glen]

I'm always fascinated by why we don't simply just drive trains as fast as we used to.

That would speed them up!

[tonyp]

Siemens Mobility has already been awarded contracts for the first stage of the digital signalling and ATO upgrade on T4, including a new Traffic Management System, which will be integrated with existing Sydney Trains systems. The clearly stated objectives of the contracts are to increase frequencies to 24 tph per line as well as to reduce journey times. How the reduction in journey times is achieved is irrelevant. as it would be a combination of automation, better acceleration/deceleration and closer headways with the digital signalling, resulting in higher average speeds, which tonyp is always prattling on about.

There is definite possibility of improvement in suburban journey times at least back to what they used to be on the system, as outlined by Glen many times, and doubtless that would meet the objectives of the contract on its own. I would see those journey times as being not dissimilar to those on Paris RER A, which itself performs journey times similar to what our suburban system used to perform in its heyday.

However, to have any hope at all of bridging the large uplift to metro journey times, you'd not only have to improve acceleration/deceleration, but also bring dwell times strictly under 25-30 seconds. Indeed you'd probably have to fully automate with little human intervention and fit PSDs to the platforms.... and then replace the fleet with 3 door single deck trains. And then you'd have to separate the lines ....

Don't abandon all hope though. The Perth system achieves Sydney Metro-level of journey times with a human driver. There's obviously vastly less demand on that system than Sydney's but, even so, they've found the need already to upgrade from 2 door to 3 door single deckers, which suggests that their 2 door cars can struggle with dwell times in peaks in Perth (let alone the situation in Sydney). They were actually considering 4 door cars but I guess they struck a balance with seating in the end. They're also about to bring some of their line separation to a close with two new lines feeding into an existing legacy line with no additional tracks. So it will be interesting to see what effects these changes will bring to performance.

There's all these railway-operation clues around for the taking and you really should look at them, rather than just have blind faith that the existing human-crewed double deckers can somehow achieve all these stellar objectives as they are. Their design and technology is simply against them. Wrong sort of train for the job.