More Trains, More Services | Signalling Upgrades

[Newcastle Flyer]

Fe fi fo fum, does thou smell ye something?

[matthewg]

The project to replace the ageing mechanical train stop system with ETCS Level 1 Limited Supervision based system has suffered serious 'feature creep', Level 2 with radio block has been trialled between Wolli Creek and Hurstville already.

Now it looks like we are diving into Level 3 with moving block and possible automatic operation of all trains.

If the goals of the project keep moving, it will never make it to wide deployment!

[grog]

I think we will see L2, not L3. ATO can be overlaid in a future phase. From the signalling strategy it looks like in cab signalling to start. They can get close to moving block headway with smaller fixed blocks anyway.

[moa999]

Some form of limited ATO to allow for use of PSDs as we move to common door position fleet on at least some lines would appear sensible.

Agree full ATO would be way too disruptive and costly (Epping to Chatswood multiplied a number of times)

[grog]

We will get Grade of Automation 2 (GoA2) which has a human closing the doors and starting the train away from the station but the train driving itself to the next stop. Similar to the current situation on London Underground lines with upgraded signalling.

[Glen]

There would have been a lot less people travelling in 1938.

A large part of the rail network was still being serviced by steam hauled trains 80 years ago so the department of railways the DRTT and the government of the time had more electric trains to run to the areas that were electrified

I'm sorry I don't quite understand the link you are making?

By 1930 most of the (then) urban area of the Sydney network was electrified, as follows:

1926 – Illawarra line to Royal National Park.
1926 – Bankstown line.
1927 – North Shore line.
1929 – Western line to Parramatta.
1929 – Northern line (Strathfield to Hornsby).
1929 – South line to Liverpool.

Source: https://en.wikipedia.org/wiki/Railways_ ... rification

[Glen]

Remember that was SD. Current trains are DD so 24tph is pretty damn good given a good proportion of the headway is dwell time at city stations.

Yes but if dwell time is the problem, why are we upgrading the signalling to run more trains between stations ?

Surely the issue is running trains closer together at the stations (the point at which they spend the longest time), which the system was originally capable of doing.

Once upon a time it used to be common to see trains close-up on each other at stations, or for the next train to enter the station before the preceding train had fully cleared the platform, by use of the intermediate trips.

What happened to all of this?

---------------------------------------------------------------

I recommend reading of the May 2018 edition of ARHS 'Australian Railway History' which contains a 1936 report on the running of the Sydney electric system.

A highlight is a picture of an actual signalling disc from 15/10/1936 showing 36 trains departing St James (at that stage a dead-end terminus) in the one hour from 5.00pm to 6.00pm, made up of 30 trains from the double-track shunting neck and 6 trains from the 'siding' (which later became the future track between Circular Quay and St James).

The disc records two instances of 4 trains in a row at headways of 80 seconds or less.

I'm pretty sure they would not have been pumping so many trains through if they didn't need them.


The article also mentioned that prior to electrification, in 1924 Sydney Terminal handled 955 steam trains every weekday, far surpassing the next busiest dead-end station in the world which at the time was Boston with 580 trains per day.

In peak hours, Sydney's suburban steam trains used to reverse in as little as 3 minutes, using a drop-on engine from the other end, just mind-blowing by comparison with today's slow standards of doing anything.

[boeing]

^ So what is the difference, then to now? Is it a safety thing?

[Aurora]

As I said guys, safety and minimising risk takes precedence over running trains next to each other to increase services.

[Glen]

As I said guys, safety and minimising risk takes precedence over running trains next to each other to increase services.

So what were the safety problems?

[grog]

At the moment trains have a slow dwell and a large separation. The result is a minimum headway of 3 minutes. By reducing the separation by 30 seconds you can reduce the headway without reducing dwell, and as such achieve 2.5 minute headway with double decker rolling stock.

You could also reduce headway by reducing dwell, but that is not what is proposed.

[tonyp]

I imagine the single deck Sydney electric stock also had much shorter dwell times than the double deck stock. They must have been timing the dwells in years back (they did that sort of thing pretty fastidiously back them) but I haven't been through enough of the engineering papers to see the results. I recall personally that it didn't take long to exchange a crowd in a carriage.

[boronia]

Apart from less overall passenger numbers, people were probably more "disciplined" in the etiquette of boarding sensibly.

[Frosty]

At the moment trains have a slow dwell and a large separation. The result is a minimum headway of 3 minutes. By reducing the separation by 30 seconds you can reduce the headway without reducing dwell, and as such achieve 2.5 minute headway with double decker rolling stock.

You could also reduce headway by reducing dwell, but that is not what is proposed.

Would the dwell time at Bondi Junction have to be reduced with a 2.5 minute headway since trains currently have 5 mins to turnaround ?

[grog]

Probably, although you would probably be able to clear a train to re-enter the platform slightly quicker. They may need to use Martin Place to turn back some services if 24tph becomes too much for Bondi Junction - they would likely have to give up on checking the trains for passengers though at the risk that people might have to take a trip into the siding and back out again 10 minutes later if they aren't paying attention to train destination.

[Glen]

Probably, although you would probably be able to clear a train to re-enter the platform slightly quicker.

Bondi Junction is a scissors crossover so you would not need to shunt.

On a 2½ minute frequency each platform would have a departure every 5 minutes so ideally a 3 minute turnaround and 2 minutes with no train on each platform.

Every second arrival and departure would involve a conflicting move over the scissors crossover (about ½ minute margin) - every other pair would not, so this sort of pattern:

Platform 1
arrive 9.00 9.05 9.10
depart 9.03 9.08 9.13

Platform 2
arrive 9.02½ 9.07½ 9.12½
depart 9.05½ 9.10½ 9.15½


Keep in mind that Bondi Junction now, is like St James was, except that at St James every train had to shunt.

Refer to earlier post about 30 trains per hour running out of the St James dead end. Imagine how close those conflicting moves were, and they were programmed automatically. Still very impressive!

[Glen]

At the moment trains have a slow dwell and a large separation. The result is a minimum headway of 3 minutes. By reducing the separation by 30 seconds you can reduce the headway without reducing dwell, and as such achieve 2.5 minute headway with double decker rolling stock.

Yes I get that bit, but what I don't get is how trains could run 80 seconds apart in 1936 (so obviously much closer than we currently schedule) yet they cannot run this close now?

Nobody has been able to tell me what was so unsafe about the signalling system installed when the underground was opened, and used for decades afterwards. (Hint .... probably nothing).

[boronia]

This photo was taken 5 years ago, and show how close trains were apart at middle of the day on a week day. There wern't any significant delays at the time

DSC00651 (Small).JPG (124.76 KiB) Viewed 5847 times

[tonyp]

Nobody has been able to tell me what was so unsafe about the signalling system installed when the underground was opened, and used for decades afterwards. (Hint .... probably nothing).

Does it coincide with when the seagulls arrived in management and the old talent was pushed out in the last couple of decades or so?

It's worth bearing in mind that the single deck system yielding that performance carried up to about 250 million people per annum, a level of patronage not reached again until the 2000s.

[Linto63]

Nobody has been able to tell me what was so unsafe about the signalling system installed when the underground was opened, and used for decades afterwards. (Hint .... probably nothing).

Probably wasn't unsafe, but there was no such thing as a health & safety regime. If the odd train ran into the back of another every now and again, was more of a case of 'these things happen'. Now there is hell to pay, and rightly so. The Granville disaster in 1977 was probably a major catalyst for change.

And its not just health & safety, but procedures are required to be in place by professional indemnity insurers, and risk assessments need to be done. Gone are the days of drivers being able to override the rules & regulations at their discretion.

This photo was taken 5 years ago, and show how close trains were apart at middle of the day on a week day. There wern't any significant delays at the time.

Quite a common scene either from both directions at Central, so obviously the tripcocks are configured to accommodate. But when operating this close, trains are crawling.

[Glen]

This photo was taken 5 years ago, and show how close trains were apart at middle of the day on a week day.

Yep, that's what close-up signalling and speed activated trips can achieve.

In theory of course, the train behind could still erroneously drive into the train in front, but only at a very slow speed.

It's worth bearing in mind that the single deck system yielding that performance carried up to about 250 million people per annum, a level of patronage not reached again until the 2000s.

I don't have the details to confirm this, but I'm thinking much of the loss could have been from off-peak, nights and weekends, when cars became more readily available. The introduction of TV in 1956 also lead to a drop in evening travel.

Conversely peak hour travel to the CBD probably continued to grow, otherwise (one would think) they would not have introduced double-deck trains in the first place.

[boronia]

This photo was taken 5 years ago, and show how close trains were apart at middle of the day on a week day.

Yep, that's what close-up signalling and speed activated trips can achieve.

In theory of course, the train behind could still erroneously drive into the train in front, but only at a very slow speed.

But wouldn't there be a trip at that point, and the driver would have to reset the system before proceeding, deliberately rather than erroneously?

[Glen]

Yep, that's what close-up signalling and speed activated trips can achieve.

In theory of course, the train behind could still erroneously drive into the train in front, but only at a very slow speed.

But wouldn't there be a trip at that point, and the driver would have to reset the system before proceeding, deliberately rather than erroneously?

Yes that's what I meant.

If driver error or foul play was involved, the trip would have knocked out the air, and after replenishing, the train could gather very little momentum before it reached the train in front.

There is a good description of trips here:

https://en.wikipedia.org/wiki/Train_stop

[grog]

At the moment trains have a slow dwell and a large separation. The result is a minimum headway of 3 minutes. By reducing the separation by 30 seconds you can reduce the headway without reducing dwell, and as such achieve 2.5 minute headway with double decker rolling stock.

Yes I get that bit, but what I don't get is how trains could run 80 seconds apart in 1936 (so obviously much closer than we currently schedule) yet they cannot run this close now?

I'd say a combination of slow speeds, tight signal spacing, relay drivers, single deck rolling stock with low dwell times and manually operated doors allowing the guard to blow the whistle and leave the station without waiting for all passengers to board and doors to close etc.

[Glen]

Nobody has been able to tell me what was so unsafe about the signalling system installed when the underground was opened, and used for decades afterwards. (Hint .... probably nothing).

Probably wasn't unsafe, but there was no such thing as a health & safety regime. If the odd train ran into the back of another every now and again, was more of a case of 'these things happen'.

Did that really used to happen or are you just surmising?

Gone are the days of drivers being able to override the rules & regulations at their discretion.

Not sure what you mean by that?

However if you happen to be referring to trains passing signals at stop, see for example:
https://railsafe.org.au/__data/assets/p ... t-STOP.PDF