Some brief transport stuff from this week

A post in an occasional series wrapping up a few brief transporty things from the last week or two.

The new train design

This might be the least crowded train I’ve ever caught. That’s because it’s a pretend train, a mock-up of a carriage and a half, somewhere in a warehouse in outer-suburban Melbourne. I got to see it last week on behalf of PTUA — we’ve been included in stakeholder consultations this year on the design.

New train mock-up

It looks pretty good, and has more places standees can hold on than the current Siemens and Comeng fleet, but could do with more still.

There’s photos of the mock-up over on the PTUA web site — take a look (and please consider joining if you’re not already a member — the PTUA’s work is only possible thanks to member subscriptions).

Busway knocked back

A few weeks ago The Age reported on Transdev’s plan for a busway from Doncaster to CBD.

  • Dedicated bus lanes along middle of Eastern Freeway (in the median originally designed for rail), with stations at interchanges, including pedestrian access from overpasses
  • Busway would continue along Hoddle Street, Victoria Parade and Lonsdale Street, to a new terminus underneath Southern Cross Station
  • Double-articulated buses with doors on both sides to allow centre platform stops along Hoddle Street in a centre median
  • Every 3 minutes in peak, every 5-6 minutes off-peak
  • $500 million build cost
  • Transdev wanted it to run as a PPP for 30 years, effectively locking them in as the operator for that time
  • Off-board payment with Myki readers on platform stops, to speed up dwell times

It would have been cheaper/more achievable than Doncaster rail, remembering that a lot of benefits of Doncaster rail would be gained by first doing the cheap easy bit: rail to Bulleen, and feeding all the buses into there.

The plan has officially been knocked back.

The question is: can the problems of greater capacity (to cope with crowding) and speed (to encourage more people out of cars) be resolved another way?

Better traffic priority along Hoddle Street, Victoria Parade and Lonsdale Street is the key: both bus lanes where missing, and traffic light priority.

More articulated buses would help with capacity. There seem to have a handful now, but not many.

Skyrail under construction near Murrumbeena station

Can Skyrail carry freight?

I’ve been asked about this twice this week alone, once online, once in the barber shop this morning.

Can the Skyrail (under construction from Caulfield to Dandenong) handle freight and V/Line trains? The rumour that it can’t persists.

It’s not an entirely silly question. Freight trains in particular can be heavier than passenger trains, and the diesel locomotives used for freight and long distance V/Line services to Bairnsdale are heavy beasts.

The answer is an emphatic yes, they will run on the Skyrail — just as they run on the 1970s era viaduct between Flinders Street and Spencer Street stations.

Here’s the official answer from the Level Crossing Removal Authority:

WILL YOU CONTINUE TO RUN DIESEL TRAINS ON THE OLD TRACKS UNDERNEATH THE NEW RAIL LINE?

The new elevated structure will be designed to safely carry both Metro passenger trains and diesel freight trains. Just as passenger and freight trains share tracks currently, they would continue to share tracks in the elevated design. The tracks underneath the elevated structure will be removed to create new community spaces.

It’s fascinating that this rumour continues to do the rounds.

And it’s certainly not helped that this completely discredited Railpage article from five months ago has never been corrected.

By the way, now that construction is in full swing, the photo above, and the one below show just how close the elevated rail will be to some people’s homes/gardens. It’s not hard to see why some residents aren’t too happy about it.

Dandenong line capacity boost undersold

On the weekend I was chatting about the Dandenong skyrail proposal to rellies at their house near Hughesdale station. They are keeping an open mind; they are not immediately under the rail line, and they can see the obvious benefit from getting the level crossings removed quickly, but are concerned about some aspects, which it sounds like are not being adequately explained.

For instance: condition and maintenance of any parkland created by the elevated rail. The Level Crossing Removal Authority (LXRA) seems to be forming the view that they would fund at least ten years of maintenance.

And of course after media attention last week a lot of local chatter is about the future need to expand the line to four tracks. (Three tracks isn’t nearly as useful as four.)

It’s worth noting that the skyrail project, along with other measures, buy a lot of extra capacity for the line.

How much? I’ve been trying to work it out, but official information is really scarce. The official figure thrown around is 42%, but this doesn’t seem to take into account all the upgrades.

Dandenong line at Richmond

Currently

Looking at what might well be the busiest hour, inbound AM peak 7:31-8:30am at Caulfield, there are 15 Metro services and 2 V/Line services scheduled, or 17 trains per hour in all.

The current signalling allows for a train every 3 minutes, or 20 trains per hour. The widely used practice is to use 80% of the theoretical capacity (to allow for short delays), which means the line is using just over its practical capacity. No wonder delays so easily occur.

Each train’s desired load standard is 798 (133 per carriage), but seat modifications currently underway are meant to increase that to 900 (150 per carriage). Remember, this is NOT a capacity figure.

New signalling

Firstly, the signalling (with conventional technology) to allow a train every 144 seconds (eg 25 trains per hour), in line with a long term plan that was kicked-off with the Westall upgrade project some years ago.

Staying within the 80% of theoretical capacity (so it runs more smoothly than now) therefore gives you 20 trains per hour — 18 Metro and 2 V/Line.

Why not high-capacity (in-cab) signalling? They want to prove it first on the Sandringham South Morang line — not unreasonable given the complexities of shared running between Metro, V/Line and freight. What they are saying is that any upgrades in the mean time will be future compatible with in-cab signalling.

Removing the crossings

Secondly, you couldn’t just do the signalling upgrade without removing the crossings, because it would lock up the road system across the south-eastern suburbs, delaying motorists, pedestrians, cyclists and bus users alike. So grade separation helps unlock that track capacity.

Dandenong line, 6pm

High capacity trains

The third major point is the trains to run on the line. The new High Capacity Metro Trains (HCMTs) will run on the Dandenong line in the next few years. As detailed in this extensive post from last year, they’ll start as 7 car sets with a load standard of 1100 people (157 per car).

This is achieved by having continuous usable space and, I expect, fewer seats. Hopefully a balance will be found between carrying capacity and ensuring those travelling the longest distances can get a seat.

Later on, when the Metro rail tunnel opens, the plan is for the HCMTs to expand to 10 car sets — a load standard of 1570. (Crush load around 2000.)

Longer trains obviously involves extending platforms — something not possible in the City Loop, but included in the Metro rail tunnel and the rebuilt skyrail stations. It will also obviously be necessary on all the other stations out to Cranbourne and Pakenham, as well as Sunbury (as the rail tunnel will connect them).

Put it all together – what do you get?

Somewhere in the mix would need to be duplication of the Cranbourne line, and flying junctions (overpasses) at Dandenong so trains don’t get delayed waiting for each other there.

So, pulling together all the various upgrade figures (which have been gleaned from various government information — it doesn’t seem to all be in one place anywhere), what do we get?

Dandenong line capacity

The government often quotes a 42% capacity boost. As I understand it, that figure includes the level crossing removals and new trains, but excludes the metro rail tunnel. I’m trying to find out how precisely they get that figure, but taking the tunnel into account, it woefully undersells the true capacity boost anyway.

Perhaps these numbers don’t exactly tally with the official ones, but 28,000 per hour is huge. That’s about the equivalent of 12 lanes of traffic. Behold, the power of mass transit!

From trying to explain some of this to my relatives, it’s clear to me that the government/LXRA hasn’t communicated this point very well (if at all).

With all these upgrades, a huge amount of extra capacity is provided, much higher than the quoted 42%, putting off the need for extra tracks for many years.

How long will this last? Well, that depends on the future growth along the corridor, obviously. Capacity can grow even further with high-capacity in-cab signalling, provided it’s okay to have all trains (including V/Line) run at a consistent speed. That should allow enough trains to provide a line out to Rowville.

Hopefully I haven’t messed up these figures. And this is not to say that planning shouldn’t happen now for future track expansion.

But if basically you’re more than doubling peak capacity by 2026 when the metro tunnel opens, that’ll take a while to fill up again.

  • Submissions to the Dandenong Skyrail project close TODAY at 5pm. If you have any thoughts, even just a few words, good or bad, put them in.

What if Bentleigh had got Skyrail?

A few people have asked me about this in the past few months — what if Bentleigh had got “skyrail”?

It’s interesting to consider, though it was never going to happen.

Firstly, the timing was wrong. The Coalition had fully funded the Ormond level crossing for removal in May 2014, with designs already having determined that it would be rail under road. In late-2014 the incoming Labor government spotted an opportunity to piggyback the Mckinnon and Bentleigh crossings onto it, which made logistical sense as well as political sense — Bentleigh being a marginal electorate, they knew to have the multiple crossings removed before the 2018 election would be a plus.

It was only after the Level Crossing Removal Authority came into being that alternative strategies such as elevated rail have been considered.

Secondly, having rail over road may have caused complications at the southern end, where the rail line goes under Brewer Road, necessitating rebuilding of the road overpass.

There are pros and cons with every design. By missing skyrail, we missed out on some good outcomes.

No doubt some locals are relieved rail is going into a trench. But rather than elevated rail and a park outside their back fence, they get an impassable cutting. The jury’s still out on whether noise is worse at ground level from skyrail or trench rail.

Better outcomes from skyrail?

Murray Road pedestrian/cyclist crossinglocal campaigners are continuing to fight for this, but the presence of a storm water pipe means the project team says the only viable solution is an at-grade crossing. Will it happen? Only if the safety audit comes up green and the minister can stomach approving an additional (non-car) level crossing as part of a level crossing removal project. With skyrail it would have been easy.

More pedestrian access — apart from Murray Road, there could have been additional pedestrian/cyclist (and possibly even motorist) crossing points anywhere and everywhere. The most obvious locations are midway along Glen Orme Avenue (Ormond, where the tennis courts access is provided), connecting Foch Street and Leila Road (Ormond), Blair Street and Ward Street (Bentleigh), and connecting the car parks just south of Centre Road. Mind you there aren’t unlimited opportunities, as for most of the length of the rail line in this section, there are houses on at least one side.

More efficient train operation. Apparently it can be a difference of up to 6% of energy consumption, as stations underneath ground level require extra braking and extra power to depart and accelerate out. This will be particularly apparent with the design as planned, where the line will come back up to surface level between stations.

Better stations. With a high proportion of costs of trenched rail going into moving underground services, and bus replacements, if these can be avoided then more can be spent on the stations and other outcomes — this seems to be what’s proposed on the Dandenong line. It’s not really working out cheaper overall, but stations are getting fully-enclosed weather cover, and (at least some; it’s not finalised) escalators — none of the Ormond-Mckinnon-Bentleigh stations will have escalators, although parts of the platforms will be underneath the roads, partly enclosing them from the weather.

Future 4th track without disruption. Supposedly if skyrail goes ahead on the Dandenong line, the 3rd and 4th tracks will be able to be built largely without disrupting the initial two tracks. It’s unclear what will happen if the 4th track ever goes in on the Frankston line. It might require partial closure again to widen the cuttings.

Pictured: Nicholson Street: October 2015 vs December 2015 after tree clearing
Nicholson St, near Mckinnon station, October 2015 (from Google Maps)
Nicholson St, near Mckinnon station, during level crossing removal works

Saving more trees. While a special effort has been made to remove and store the palm trees for re-planting later, almost every other bit of flora in the corridor has been cleared away. From some angles it resembles a moonscape.

As this Dandenong line FAQ notes: One of the benefits of the elevated rail design is the ability to retain trees and vegetation close to the rail corridor. By elevating the rail line, we minimise our impact to the root systems of trees, and are able to retain a significant amount of trees within the rail corridor.

Parkland underneath. Glen Eira council is forever reminding us that they have less green space than anywhere else in the state. Some of what little green space there is in Bentleigh will be taken by car parking, to make up for a small loss of spaces at Ormond. Skyrail could have enabled more green space right along the line.

Continuous bike/shared user path. The current project will restore the bike path from Bentleigh to Mckinnon, but there still won’t be a connection from Mckinnon to Ormond — cyclists and walkers have to divert via local streets, because the rail corridor is too narrow. If it were skyrail, a shared user path could easily fit underneath.

A more prominent train system. Some might not consider this important. But out-of-sight, out-of-mind is a concern. At ground level or elevated, the trains are prominent. Hide them away and they are less obvious to people. Does that influence how people think about their transport choices? I don’t know.

Trains will still be visible in their trench of course, if you’re a pedestrian walking by, but less so if driving past. Hopefully good prominent signage and stations will make up for that. (I’d like to see signage include a countdown timer to the next trains, more prominent than the one we once had.)

Better views from the train. OK, no biggie, but it’s more pleasant looking out over your suburb than staring at a plain (or graffiti-covered) wall. It’s also easier to navigate if you can see where you are.

That said, the current plan is for trains to return just about to street level between stations. Roller coaster!

Train trench, Gardiner. Opened in January. Already tagged.
Train trench near Gardiner station

And then of course there’s the disruptions

Road and rail shutdowns. The LXRA claims Dandenong skyrail will involve about 15-20 weekend shutdowns, and two longer shutdowns of 9 and 16 days. Yesterday the government claimed that if it was done as trenches, the Dandenong line would require shutdowns “for 230 days during construction” and that “these closures would be between 30 hours and eight weeks long”.

For the Bentleigh project, it’s shaping up as numerous weekend shutdowns, plus 9 days (January 2016), 10 days (Easter 2016) and the big one: about 35 days (TBC; June-August). That’s a total of about 70-80 days in all, though in some cases (such over the past weekend) they closed the line for the weekend as early as 7pm on Friday.

The list of line closures so far is:

  • Third track closed (thus reducing capacity, and resulting in several cancellations daily) Mon 16/11/2015 until the project is finished in July 2016
  • Last two trains after midnight, Fri 13/11/2015
  • 1am Sat 21/11/2015 to last service Sun 22/11/2015
  • 1am Sat 28/11/2015 to last service Sun 29/11/2015
  • 1am Sat 12/12/2015 to last service Sun 13/12/2015
  • First service Sat 23/1/2016 to last service Sun 31/1/2016
  • 8:45pm Fri 12/2/2016 to last service Sun 14/2/2016
  • 7pm Fri 4/3/2016 to last service Sun 6/3/2016

And expected as the project progresses:

  • 9pm Fri 19/3/2016 to last service Sun 21/3/2016
  • Closure of Ormond and Mckinnon stations 25th March to late July 2016
  • First service Fri 25/3/2016 to last service Sun 3/4/2016
  • Closure of Bentleigh station from 4th June to late July August 2016
  • Expected line closure from 25th June 2016 for about 5 weeks

Each time, they warn of travel time increases by up to 45 minutes. During the January shutdown, lot of people gave up and drove, switched to other train lines (causing crowding elsewhere), or made other arrangements.

Closures don’t just affect train passengers — many involve closing roads, which affects local businesses. There were claims of huge loss of earnings at Burke Road, Gardiner (though it was never the busiest of shopping centres). To counter the problem, the government has organised (no doubt at considerable expense) full colour brochures promoting local shops, sent to every resident in the area, and an accompanying web site. (I notice that the site lists Milsims Games, which moved out of Bentleigh some years ago.)

Would there be fewer shutdowns with skyrail? Parts of the alignment are very tight — perhaps tighter than the section through Carnegie. The north side of Mckinnon station looks like it’s a similar width to the narrowest section between Carnegie and Murrumbeena, but it’s got three tracks in it, not two. And what you don’t want to be doing is acquiring properties if it can be helped. (For the current design it was avoided.)

Hopefully it would have been possible to stagger the construction, building elevated new rail lines above the older ground level lines while they continued to operate.

The key is minimising weekday shutdowns, especially outside school and university holidays. Weekday shutdowns require 100+ buses in operation, as well as numerous support staff — two at every station replacement bus stop, plus many more at the interchanges from first to last service. There are often temporary road markings or modifications, signage, traffic light modifications and traffic monitoring. It’s a very, very expensive undertaking. (We don’t know quite how expensive, but by comparison, recent V/Line bus replacements have cost up to $300,000 per day.)

Risk of unplanned disruptions. So far I’m aware of only one incident: on February 12th, excavation at Bentleigh for the crossing hit a gas pipe, causing a leak, with a sudden road and rail line closure for some hours. As passengers who have been caught up in one know, any unplanned rail closures are messy. Buses often can’t be got to the scene quickly, and are rarely sufficient, especially if the closure extends into peak hours.

Truck movements. Apparently 280,000 cubic metres of earth are to be dug up and moved out. Expect to see thousands of truck movements in the area, particularly during the main shutdown in June/July/August.

E.E.Gunn Reserve partly closed. Parts of this park will be closed for months to temporarily store the earth. One of the benefits of skyrail is that components can be built off-site and shipped in and assembled, like some kind of massive Ikea flatpack — and of course there’s little or no earth to dig out and take away.

Stations shut. As noted above, Mckinnon and Ormond stations are shortly to close for about four five months. Bentleigh will close for a shorter period of about two three months. (Some of this overlaps with the entire line closing for about 5 weeks.)

Bentleigh now vs what if Bentleigh had skyrail? Yeah, my Photoshop skills aren’t up to much.
Bentleigh level crossing
Bentleigh if it got skyrail

Would it have been possible?

A Vicroads contact suggested skyrail might not work over North Road, because of space issues, but a slight re-alignment might have solved that. Certainly the road overpass in Brewer Road in Bentleigh may have caused problems.

Would it have been politically feasible? I wonder if the government could have stomached the risk in the very marginal seat of Bentleigh? They may yet face a similar conundrum further south on the Frankston line.

We might never really know, but had the concept of elevated rail been considered earlier for the Bentleigh area, the potential was there for a project with far fewer disruptions, and some markedly better outcomes for locals.

Under, over? Level crossing removal techniques compared

Ian Woodcock (RMIT) and John Stone (Melbourne Uni) have a new report out comparing level crossing removal methods. The Age has a story on it this morning, and hopefully the full report will be online very soon. UPDATE: Here it is.

I did manage to get a sneak preview, and looks at various case studies around Melbourne, and tries to evaluate aspects such as: connectivity, accessibility, intermodal access (eg interchange), safety, economic development and amenity. The summary table is here:

Summary of level crossing evaluation (from Woodcock and Stone, 2016)
(Click to see the table larger on Flickr)

Method depends on motivation

One of the really interesting points it makes is that the motivations for using various methods have been quite different over the years.

Prior to World War II, it was mostly about improving rail efficiency, and whichever method resulted in the slightest gradient was what got used. So if one looks at the Glen Waverley line, extended from East Malvern and opened in 1930 (and the last big suburban rail project until Regional Rail Link opened in 2015), there are no level crossings, and it’s a mix of rail over and rail under.

After World War II the emphasis was about moving cars, and the cheapest method was road over, such as at Burnley, Oakleigh and Sunshine. The result, the report authors point out, disconnects communities and makes life very difficult for pedestrians and cyclists at street level.

More recently the trend has been trenched rail under road. The high cost of decking mostly prevents additional connections across the tracks or use of land above it. I suspect some people like that it’s out-of-sight, out-of-mind (well, except to train passengers and pedestrians), but the report points out problems with this method — particularly in terms of lack of integration between stations and local neighbourhoods.

As I understand it, putting stations at the bottom of a trench also has poor consequences for rail efficiency. One of the numbers I’ve heard is that it can add 6% to train energy consumption. From this perspective ideally you want stations on a hump, so they can easily brake into stations, and use minimum energy accelerating out of them again.

The report authors have plotted the history of Melbourne grade separations in this fascinating graphic, and you can see the trends over time:

Timeline of grade separations (from Woodcock and Stone, 2016)

As I’ve said many times, all methods have pros and cons.

One lesson here is that one shouldn’t judge current proposals based on what’s been done before. Melbourne has no good past examples of grade separation that was designed to maximise connectivity, economic development and amenity. Most of the past projects have only been targeted at improving rail or road operations, and that’s all they’ve achieved.

Elevated rail such as proposed for the Dandenong line has obvious impacts on those living closest to it, but if designed well it also brings a lot of benefits in terms of overcoming engineering challenges, construction impacts and eventual outcomes such as land use.

The report is well worth reading: Level crossing removals: learning from Melbourne’s experience