http://www.theage.com.au/victoria/doubters-fear-signal-upgrade-wont-work-on-melbournes-old-and-complex-rail-network-20140913-10gdrf.html
I made the comments earlier in the year about the costs etc
Here's the story to prove it...
Cost blowout is coming!

http://www.theage.com.au/victoria/doubters-fear-signal-upgrade-wont-work-on-melbournes-old-and-complex-rail-network-20140913-10gdrf.html
I made the comments earlier in the year about the costs etc
Here's the story to prove it...
Cost blowout is coming!

There'll be a reduction in scope for the HCS implementation, mark my words. Probably Dandenong-Cranbourne only, as there's no V/Line or freight services to deal with that way. There was a reason they wanted to do a HCS trial on the Sandringham line first.

Similarly Imade the same point earlier this year.
The issue is not about implementing a new high capacity signaling system.  There is no doubt such a system is neededthat can offer a capacity and efficiency network right across the network).  The prime issue and the oneconcerning highly experienced people in the Department of Transport and elsewhere is the type of system being proposed.   In simple terms the system being proposed byMetro Trains and their consortia is only used on dedicated “Metro” systems that use one type of train.   To the best ofmany peoples knowledge, this system has NOT been used on a mixed traffic railway of multiple train types and configurations.

The systemthat others would like to see is the European ECTMS (the techno guys will correct me on the right acronym).  
Basically it is a PROVEN European wide system used for mixed trafficnetworks that have all the same characteristics as the Dandenong Corridor, (junctions, level crossings, freight trains, regional passenger trains, metropolitan passenger trains etc).   The system isalso used in other locations as well.

In asimplistic sense the comparison with Myki is very relevant where there was enormous customization undertaken and considerable impacts on the customer.   The proposed system will dramatically impactfreight and passenger operations.  Only asmall fleet of V/Line V’Locity trains will be fitted so that impacts fleet utilization and hence the recent clearance tests of V”locity sets on the Avon River Bride.  I haven’t been updated onhow it affects freight but I’m told there will be both technical and operating restrictions involved.

In contrastthe system preferred by others is scaleable so it can be progressively installed across the network driving long term benefits for everyone.

(Sad to sayit seems to be a Victorian mindset here yet again like singling the Bendigo Line, not spending just 1% more on gauge convertible sleepers for RFR, maintenance and other projects for future proofing, Myki, etc)

If this ever works, and we get 30 trains an hour, we might as well close all the roads that have level crossings. A warning time of 20 or 25 seconds, plus time for the train to pass, plus time for the warning system to stop means that no cars will have a cat in hell's chance of getting across.

If this ever works, and we get 30 trains an hour, we might as well close all the roads that have level crossings. A warning time of 20 or 25 seconds, plus time for the train to pass, plus time for the warning system to stop means that no cars will have a cat in hell's chance of getting across.

Managment of the metro system already know and understand this problem, I think you will find the installation of the system will be on condition the level crossings are all to be grade separated.

The source of this is an interview with senior metro managment on ABC local Melbourne 774khz sometime ago. He stated the level crossings was one of the most serious restrictions on increasing the rail traffic, due to grig lock being caused on the roads by closed level crossing barriers. He also stated the Melbourne network had over 200 level crossings where as the Sydney network only had 4.

woodford

Managment of the metro system already know and understand this problem, I think you will find the installation of the system will be on condition the level crossings are all to be grade separated.

The source of this is an interview with senior metro managment on ABC local Melbourne 774khz sometime ago. He stated the level crossings was one of the most serious restrictions on increasing the rail traffic, due to grig lock being caused on the roads by closed level crossing barriers. He also stated the Melbourne network had over 200 level crossings where as the Sydney network only had 4.

Sydney had a progressive program from the 1960's through to the 70's to remove nearly all the metro level crossings - it worked really well. No such program here.

Reading the original Age article, there seems to be some doubts about the ability of the new signalling system to actually deliver what it is supposed to. V/line and express trains will still have to crawl.  Wouldn't a better option have been to proceed with the tri- or quad- tracking of those pinch points? Or would that make too much sense?

Similarly Imade the same point earlier this year.
The issue is not about implementing a new high capacity signaling system. There is no doubt such a system is neededthat can offer a capacity and efficiency network right across the network). The prime issue and the oneconcerning highly experienced people in the Department of Transport and elsewhere is the type of system being proposed. In simple terms the system being proposed byMetro Trains and their consortia is only used on dedicated “Metro” systems that use one type of train. To the best ofmany peoples knowledge, this system has NOT been used on a mixed traffic railway of multiple train types and configurations.

The systemthat others would like to see is the European ECTMS (the techno guys will correct me on the right acronym).
Basically it is a PROVEN European wide system used for mixed trafficnetworks that have all the same characteristics as the Dandenong Corridor, (junctions, level crossings, freight trains, regional passenger trains, metropolitan passenger trains etc). The system isalso used in other locations as well.

Problem is that the systems that exist at the moment don't do what is wanted in Melbourne.

What we want is ETCS Level 3. The problem is that this doesn't exist as proven product yet - indeed last time I looked it didn't even exist as a proposed standard yet. This leaves a choice of ETCS Level 1 or 2, or a proprietary CBTC system. The first two won't deliver the capacity improvement. The third has a high risk of vendor lock-in.

ETCS (European Train Control System) is an international standard, vendor kit will interoperate, and it has been applied to all sorts of trains. Three big ticks. Problem is that it won't deliver any capacity improvements at the moment. ETCS comes in three levels. Level 1 is a bolt on to the existing signalling system, which makes it 'cheap'. However, this also means that there is no increase in capacity - really the only benefit is a reduction in SPADs. Indeed, unless additional kit is added, ETCS Level 1 is likely to cause a reduction in capacity. ETCS Level 2 uses mobile telephones (GSM-R) to communicate instructions to the trains, but still uses the existing track circuits. This, at least, means no reduction in capacity, but doesn't mean an increase in capacity as the existing track circuits are used. It also requires significant additional kit - probably replacement of much of the remaining relay based signalling with modern computer based interlockings. And it requires significant mobile telephone bandwidth as each train essentially has a continuous phone call open to the central computer. For both levels, it would be possible to resignal the lines with new short track circuits to get additional capacity, but this would be expensive...

CBTC (Communications Based Train Control) uses radio to communicate both the position of the train to the central control and the instructions to the train. It does increase line capacity (proven). ETCS Level 3 will be a CBTC, but it doesn't exist yet as proven product. This leaves a slew of proprietary products. The fundamental problem with all of these is that if we choose one of them, we will be locked into the product and the vendor forevermore (or at least until we spend another squillion dollars to replace it). And when ETCS Level 3 comes down the path, the proprietary systems will be dead products. Most of the current products are designed specifically for metro style systems and adapting it for a mix of trains would be a financial risk. (Incidentally, the ARTC system (ATMS) is a CBTC, but is not designed for metro systems.) To give you an idea of the risks, one of the leading CBTC systems was chosen by London Underground a couple of years ago for installation on the District and Circle lines. After a year or so of development, LU and the vendor jointly decided to can the project as the product wasn't a good fit.

ETCS, at the moment, is really about interoperability (particularly in Europe) and improving safety (particularly on high speed/high
density lines). Most existing CBTC products are about increasing capacity on metros. Your choice. Choose wisely.

2 questions.

i. In relation to a signaling system, what is the difference between one type of train and another?
My understanding is the signaling is moved from trackside to in the cab. (Possibly an oversimplification) wouldn't it be a case of handing out boxes to every metro/freight train that comes to town?

ii. Can't remember where I saw it, but freight train heading to WA being signalled in the cab with authorities given well in advance.
Is this an Australian system, is it HCS and (depending on the answer to q1)  could it be adapted to other networks.
I.e. if was specifically built here for one condition, modify it to suit another.
(So we're at least locked in to and Aus bssed product)

Sydney had a progressive program from the 1960's through to the 70's to remove nearly all the metro level crossings - it worked really well.

The difference is fairly noticeable. I can't ever recall seeing any level crossings whilst driving or training around Sydney. Does anyone know where these 4 crossings are?



When I was briefly in Clayton, the traffic queues around the level crossings were horrendous at certain times of the day.

IN the interview I have already refered to the metro management stated more services on the Dandenong line could by used currently, but to do so would mean the level crossings would be closed for too much time resulting in road traffic kaos, he actauly did have the times on hand. I cannot now remember what these were but it was something in the order of 40% of the time closed. For this reason more services could not be used  via Dandenong till something like 11 level crossings were graded separated. He said it was critical the grade separtions are done NOW.

My impression from discussions on the capacity of the network that the level crossing issue is VERY seriuosly underestimated, this manager said it would be almost useless to do any further capacity improvements with the existing 200 odd level crossings in place as the level crossings would effectively block all road traffic.

In effect what he said was around 20 billion needs to be spent on level crossing up grades BEFORE the network capacity could be increased. It appears the road lobby including a good portion of Railpage think road and rail exist in isolation any intergration not being an issue THIS IS NOT CORRECT.

Note: The proposed line to the airport does not suffer much from this as there are now only (I think) a single level crossing involved and this would not be difficult to remove.

woodford

2 questions.

i. In relation to a signaling system, what is the difference between one type of train and another?
My understanding is the signaling is moved from trackside to in the cab. (Possibly an oversimplification) wouldn't it be a case of handing out boxes to every metro/freight train that comes to town?

ii. Can't remember where I saw it, but freight train heading to WA being signalled in the cab with authorities given well in advance.
Is this an Australian system, is it HCS and (depending on the answer to q1) could it be adapted to other networks.
I.e. if was specifically built here for one condition, modify it to suit another.
(So we're at least locked in to and Aus bssed product)

Question 1: The braking curve. That is, the plot of speed versus time once a brake application has been made. There are standard braking curves for EMUs, loco hauled passenger trains, and various types of freight trains. They are very different, hence the distance to stop from different speeds varies significantly.

In this context, the fear is that the proprietary CBTC systems can only support one braking curve - a metro one - and will require modification to support multiple curves. Modifications means financial risk.

Question 2: You are probably thinking of ARTC's ATMS (Advanced Train Management System), which, as I have already said, is a CBTC. Apparently works well, but has the same problem - would require modification and hence financial risk. From a government point of view, the risk of ATMS would be even higher than purchasing from the big vendors. It's still proprietary, and has been developed for a single operator. If that operator ever lost interest you'd be stuck with all of the development and maintenance costs.

IN the interview I have already refered to the metro management stated more services on the Dandenong line could by used currently, but to do so would mean the level crossings would be closed for too much time resulting in road traffic kaos, he actauly did have the times on hand. I cannot now remember what these were but it was something in the order of 40% of the time closed. For this reason more services could not be used via Dandenong till something like 11 level crossings were graded separated. He said it was critical the grade separtions are done NOW.

My impression from discussions on the capacity of the network that the level crossing issue is VERY seriuosly underestimated, this manager said it would be almost useless to do any further capacity improvements with the existing 200 odd level crossings in place as the level crossings would effectively block all road traffic.

In effect what he said was around 20 billion needs to be spent on level crossing up grades BEFORE the network capacity could be increased. It appears the road lobby including a good portion of Railpage think road and rail exist in isolation any intergration not being an issue THIS IS NOT CORRECT.

Note: The proposed line to the airport does not suffer much from this as there are now only (I think) a single level crossing involved and this would not be difficult to remove.

woodford

I would generally agree with the Metro manager - it would be difficult to substantially increase train services on, say, the Dandenong line without significant political backlash. Which would make it politically unpalatable. However, the recent improvement to a 10 minute service did *not* cause any significant local political heat. Similarly, the 5 minute peak service on the Frankston line is not causing significant heat between Glenhuntly and Cheltenham. So I would say there would be room for service improvements on most lines without removing level crossings. I would suspect the Metro manager is applying a pressure on the government to get something they want.

I've also always been amused by quoting closure times - in this case quoted at 40%. At most traffic lights each road is closed for through traffic for *more* than 50% (when red overlaps and turning cycles are taken into account). Indeed the north-south roads between Caulfield and Oakleigh cross Dandenong Rd shortly after the rail crossings, and the north-south roads are closed for traffic for much more than 50% of the time at these traffic lights. The problems with level crossings are caused by other factors than simply the percentage time closed - which may indicate that there are other ways of reducing congestion than simply grade separation.

With the exception of an access road into Tottenham yard, there are now no level crossings between Melbourne and Albion, and there have never been any level crossings on the Sunshine - Broadmeadows line.

If we just grade separated all the level crossings on the Dandenong line, I can't see why we even need High Capacity Signalling. The current system copes with a train every 2 minutes, the only reason we can't run trains every 2 minutes is because of the crossings. Not the capacity of the signalling. But spending $2.5b on flashy new signalling does sound better, doesn't it? *rolls eyes*

If we just grade separated all the level crossings on the Dandenong line, I can't see why we even need High Capacity Signalling. The current system copes with a train every 2 minutes, the only reason we can't run trains every 2 minutes is because of the crossings. Not the capacity of the signalling. But spending $2.5b on flashy new signalling does sound better, doesn't it? *rolls eyes*

Moving Block Signalling works better than fixed block when things go wrong - delayed trains can move closer together when they're going slowly. There's also the possibility of course for Moving Block Signalling to be upgraded into full Automatic Train Operation in the future - it would use pretty much all of the infrastructure already installed.

Will Moving Block signaling allow speed limits to be raised to 115 km/h between Oakleigh and Dandenong? There's a few gaps between stations where trains would be able to exceed 80 km/h which would allow travel times to decrease without skipping stations.

Will Moving Block signaling allow speed limits to be raised to 115 km/h between Oakleigh and Dandenong? There's a few gaps between stations where trains would be able to exceed 80 km/h which would allow travel times to decrease without skipping stations.

I believe that it would allow for a higher average speed (based on a now abandoned UK plan to do this), but I don't think it would allow for a higher speed limit. That would have be done via track upgrade.

A little bit off topic, but regarding speeds, I recall many years ago, a level crossing accident somewhere east of Dandenong resulted in country trains ( don't think it was Vline back then) being replaced by buses for part of the journey. During the afternoon peak, a three car Hitachi was commandeered to run from the City to Dandenong, as a replacement for one country train, where passengers were transferred to buses.  I boarded at Caulfield and we ran express to Dandenong. (no Clayton stop in those days.) I decided to time the trip and was amazed that it only took 12 and a half minutes! That's an average of slightly over 90Km/h. In those days I think that max speeds were Caulfield to Oakleigh 80 Km/h and Oakleigh to Dandenong 100Km/h. Full marks to the driver.

LancedDendrite wrote:
There's also the possibility of course for Moving Block Signalling to be upgraded into full Automatic Train Operation in the future - it would use pretty much all of the infrastructure already installed.

And that's a good thing?! Because why??

And that's a good thing?! Because why??

I didn't say anything about the merits of ATO (or its pitfalls). But if you insist...


Pros:

• Allows you to run high frequency 24 hour services without needing to pay huge penalty rates


• Doesn't need new drivers to be trained up in anticipation of increasing service frequencies or receiving new rollingstock


• No driver changeovers needed


• Can go up and down the same section of track many times without suffering from fatigue and complacency issues

Cons:

• Requires lots of expensive signalling infrastructure to be installed on each line it is used on (although as I said, Moving Block Signalling gets you pretty close to ATO)


• Is generally unable to handle level crossings


• Can't really deal with door opening/closing


• Can't deal with sick or criminal passengers onboard, nor help people with wheelchairs to board (although that shouldn't really be what a driver is there for anyway)

That comprehensive enough?

The Victoria Line in London manages to schedule 34 trains per hour (roughly 1 train every 105 seconds) at peak times using ATO with drivers to operate the doors. Would a similar system work on the Dandenong Line assuming all crossings are emilinated, all trains are of the same type and V/Line and freight aren't using the line during peak hours?

That would be much better, as the driver could also take manual control should the automation fail.

So the new signalling will go through the Caulfield loop as well? If it was just beyond Caulfield I don't see what the point would be. If so Frankston trains will no longer use the loop?

I probably do understand much about HS however can explainhow this would integrate with current signaling systems at places like Caulfield and Southern Cross

I probably do understand much about HS however can explainhow this would integrate with current signaling systems at places like Caulfield and Southern Cross

If I'm interpreting your poorly-phrased question correctly, you're asking about how High Capacity Signalling integrates with existing signalling systems?

If that's the case, then it pretty simple - when HCS is enabled on a particular section of railway line, only HCS-equipped trains are allowed through by protecting the signal to enter that section using some form of authentication/authorisation - radio transponders are the usual method.

Alternatively, fixed-block signalling can still exist alongside HCS, with HCS-enabled trains being able to enter fixed-block sections that have HCS equipment installed and are already occupied by another HCS-equipped train (i.e passing the signal at 'danger') but not in any other case. This is not a good compromise however and should really only be used in situations where you can't implement HCS on every train that will pass through the section.
For an example, on the Dandenong corridor you have mixed traffic - V/Line Passenger trains to/from Gippsland, the regular Maryvale freight train and Metro suburban services. It is impractical to install HCS equipment on all of the V/Line rollingstock that use the line (N class locos, Sprinters and Vlocitys) and would require HCS-installed locos to lead freight services. An even greater task is installing HCS equipment on the Metro rollingstock that use the Dandenong corridor, as only 25 next-generation trains would be purchased for the line. It's a rather large task to perform just to trial HCS.

This is one of the many reasons why the initial roll-out of HCS was going to be on more isolated lines on the suburban system, like Sandringham or Cranbourne.