Generalised competitive positioning model for USA high speed rail
There have been some recent posts such as Phin’s on high speed rail in the USA on the different blogs, stemming from recent announcements from the Obama administration regarding funds being made available for defined corridors, generally in the higher density parts of the USA, the NE seaboard, California, the Great Lakes, Florida and possibly Texas.
Most commentators seems to be still pessimistic regarding the potential of HSR outside of certain politically favourable areas. Much of this, from transport lobbyists particularly, comes from understanding the HSR does not just succeed by getting people from A to B fast; it requires a whole supporting network of transport infrastructure, hard and soft, working with the ‘headline act’ namely the single HSR route between the major cities. This infrastructure is generally available in Europe and Japan, and consists of urban public transport (fixed rail) of frequency and quality that doesn’t leave the passenger stranded when they get to the downtown terminal. It also consists of supportive pricing, passenger information and other policies that assist the service to be seamless door-to-door, as its main competitor, a combination of road and air, tends to be.
This article in the San Francisco Examiner outlines some of the issues – how to connect cities to their urban rail networks.
I have posted elsewhere that I don’t think the situation is as chronic as some would think. Some of the major routes under consideration eg the North East corridor, the major cities all have urban rail systems and I suspect the key issue will be quality of service and information, not absence of infrastructure. I could be the most savvy New Yorker and know all my lines, yet not know the same depth of detail about Washington DC, and end up getting a taxi (in which case, the airport might be just as convenient as the downtown rail station). And getting this quality of service is a more cultural thing than a case of hard dollars.
We can look beyond the NE and see similar situations elsewhere. California has a mature rail system in San Francisco and an expanding one in LA. The quality of service is good but reach is ordinary. Sacramento, San Jose and San Diego have light rail systems – expanding the reach somewhat although I question how supportive a light rail system can be of a passenger on a time-sensitive journey, beyond a few blocks from the downtown terminal.
I had a think about a conceptual framework for looking at how HSR might be competitive in the USA without having serious hard data to support the proposition. The US aviation market is mature and extensive and I have had some experience with quantitative modelling of it in the past. The road system is also mature and extensive. The rail system is primarily structured around historic metro systems in the oldest cities, light rail (generally reserved track with some street running) in some cities, long distance commuter rail (read Mees A very public solution for a critique of this type of operation) and conventional Amtrak services between the major cities, which outside the NE corridor tend to be tourist-type services unsuited to non-tourist journeys due to slow transit time and lack of frequency.
So without any data that I can be bothered bringing to this discussion I thought of a matrix-type analysis and would welcome people’s comments on how well the different market segments I define below actually fit into the model.
First, I’ve developed a fictitious city-triplet with A-town (pop 2 million) B-borough (pop 1.5 million) and C-ville (pop 250,000) with A and B roughly 600km apart in a relatively dense part of the country, and C about 200km from B.
In common with the Australian experience, a mature air service runs between A and B at high frequency and operated by flag carriers at the prevailing business fare, from the large international airports in both cities. In addition, a turboprop service flies 4 times a day from B to C. The purpose of this service is to provide a good connecting service for A to B passengers, but also for the many other aviation origin-pairs that arrive in B and need an onward connection. This service tends to be otherwise very expensive and is only used by time-poor politicians and poseurs for the B-C journey by itself.
In addition, both A and B have former air force bases that are used by a low cost carrier 5 times a day. The fares are considerably cheaper but all the disadvantages of this model apply in this case.
Further, the flag carrier has started flying a direct service from C to A, each way in the morning and the afternoon. Because this route is a bit less demanded, they sometimes offer specials on this route, only on the direct service, and it can even attract people from outlying parts of B flying to A who don’t mind the drive.
The road network, being mature, has no remarkable characteristics other than noting it would be crazy to drive from C to A through B, even though the cut off freeway from C to A is not direct it is better in peak times than going through B.
The rail network consists of an urban rail service in A and a light rail service in B with a commuter rail service from B to C in peak times, as well as an Amtrak service that runs twice daily stopping at C on its way to elsewhere. A to B is served by a once daily Amtrak service that takes 8 hours for the journey.
From where we sit today we can see that aviation is the only feasible option for business travellers, and low cost air, driving, bus or train are options for tourists or VFR travellers.
I’ve defined some traveller archetypes – and I’ll try to match them according to their travel needs to the modes present and future. All live in A-town.
a
Business traveller at day meeting in B-Borough; downtown origin and destination; taxi to major airports; full fare on flag carrier; all costs paid by employer
b
Business traveller in B-borough for Monday to Friday; suburban location; needs car; self employed
c
Business traveller from A-town visiting C-ville for 2 day seminar; employer pays but watching costs closely; could leave own car at airport;
d
VFR traveller staying in enviros of B-town and C-ville for a few days; prefer central location in B-town
e
VFR traveller visiting friend for weekend on whim, paying for self; suburban location
f
Uni student starting term in C-ville; needing lowest cost option; flexible about start time and mode; suburban location; can get dropped off some of the distance at A and might have minibus connection provided by Uni at B
The matrix of my assessment of how likely the archetype is to use each mode, for that purpose, is shown below:
It is impossible to be definitive in an exercise like this – having generalised this much I’m not giving much concrete to argue about.
Some discussion points:
-HSR and premium aviation are, all things being equal, competitively matched for downtown-downtown travel using taxis at both ends. What aviation gains at higher cruising speed, it loses for check in and the longer taxi journey. Rail is probably ahead on comfort and convenience. Rail’s fares are rarely any better than flying though, at the premium end (check Eurostar if you don’t believe me!). Questions about ’supporting infrastructure’ don’t matter in this market. Hence I’ve marked “a” as definitely travelling on both modes.
-as the business market’s needs diversify, eg to suburban locations or to C-ville, HSR+urban rail could still hold its own, if the supporting infrastructure (urban rail) holds up. Driving self becomes viable after the length of stay increases; the driver has no confidence in the urban transport of B-borough.
-”c” is a very common business traveller type these days, an employee going to a conference which is marginal to the employer, who might say “go, but I can’t bill your travel to anyone so keep it cheap” and as conference venues are often in non-downtown locations, the premium air option is looking unnecessarily expensive. Again the success of HSR+urban rail will depend on how well urban rail holds up its end of the bargain. “c” might find himself being dropped off by his/her spouse at a suburban rail station, prior to boarding a HSR train at the downtown terminal. No taxi to bill to the firm from home.
-the VFR and tourist market is much more price sensitive and the HSR+taxi, like the premium air option, is looking undesirable, especially for those more discretionary options like ‘whim’. However, HSR might have some seats to offload at quiet times, like the low cost air option, and could make a reasonable play for those passengers if the supporting infrastructure eg urban rail is there.
-the uni student living in C-ville will probably need the car and drive themselves. If that is not an option, then probably the cheapest intercity transport either to C-ville, or to B-Borough which is close enough to get some local transport like a Uni-owned minibus. If HSR happens to have cheap seats or student discounts that are targeted at niche low cost markets (for example, positioning runs late on a Sunday night), might be in the running.
My conclusion: HSR is a suitable mode at both the premium and low cost ends of the spectrum, provided the supporting infrastructure is there. If not, then it will have to survive only on the premium end of travel – downtown to downtown, supported by taxis.
Options including C-ville will not be viable from B-Borough, as the taxi cost will be prohibitive even for the “a” customer, and driving oneself from B-Borough to C-ville is not an option in a hire car as that customer will find the premium flights into C-ville airport, either direct or via connecting turboprop, a better option.
It is therefore crucial that at least at the skeletal level that we can see some evidence of a supporting infrastructure for HSR below the ‘headline act’ of the single high speed main line.
Network effects and granularity.
My fictitious city-triplet is in a relatively dense part of the USA and it can be assumed that all modes have more traffic than the city pair line hauls. The airports, bus terminals and even the rail terminal are hubs for intercity services, and also support a local network of transport as well (taxis, buses and limited urban rail).
In Europe and Japan, where the population is the most dense there is also a thick network of connections between rail (and other modes) linehauls. Network effects multiply the possible journeys into a much better ’soaking up’ of potential transport demand and give the public modes a much better shot against private motoring (which has a dense network of routes in all markets).
HSR, too, depends not just on the ‘headline act’ downtown to downtown journey, nor even just to the suburbs of the respective A-town and B-borough, but also to other combinations of cities, changing at B-borough for C-ville, D, E and other cities in the area. The more of these combinations in the network, the more each linehaul trunk will enjoy in the way of patronage; the more sustainable and effective in the long term.
But Europe and Japan also enjoy better ‘granularity’, not only is the number of intercity destinations vast, but also within each city effective public transport (fixed rail) is able to zoom you in to exactly where you want to go, whether by suburban mainline stations, above ground suburban rail, underground rail or light rail.
The USA will clearly lack this – even in some of the better examples like New York or San Francisco. The BART for example, might well get you to SF HSR terminal from Oakland but will it get you from a specific city block in Oakland, or will you still need a taxi? And if you need a taxi – should you just stay in the taxi to the SF HSR terminal? Or to the airport and be rid of the rail option? If you lose someone from rail for part of the journey, they might discount it for the whole journey.
So the question is – how much ‘granularity’ constitutes the effective support that HSR needs at the urban level.
This is where I do tend to part company with some other commentators – I would suggest down to the 5-10km level would be plenty, at which case a good urban rail system in each of these cities is doing its job. Even in Europe and Japan, the lineup of taxis outside the major TGV or Shinkansen stations is still impressive.
In most of the archetype examples given above, luggage is a factor. This implies any of the following: lack of preference for light rail or urban bus, possibility of being dropped off, parking one’s own car, getting a taxi. So long as HSR holds it own against aviation, it is still in the running in these cases.
Second, there is a price versus time trade-off, but only up to a point. A cheap flight is not cheap if the airport takes all the saving away, or the taxi does. Effective HSR+urban rail is going to have price points and timing points that it must achieve reliably. Urban congestion is not normally a problem for people to get to their nearest railstation, but to get to downtown or the airport could be.
Finally, I’m very sceptical of light rail in the equation. For more than a few blocks it is too slow. It provides the necessary granularity around the downtown areas but unless the service approximates the speed of suburban rail, will end up too slow for the time-constrained passenger.
One comment about the prospects of HSR in Australia. You will notice from my analysis above that it is hard for HSR to deliver more than the ‘headline act” and there is a lack of intercity network effect. It is fine to say that even Sydney to Canberra or Melbourne to Canberra are ideal distances for HSR – it is hard for these cities to act as hubs for other journeys as there are no other in consideration.
And hubbing to conventional options such as Sydney to Newcastle or Moss Vale to Wollongong are not possible while the existing transport on those routes remains so poor. As for local journeys, high profile unreliability of suburban services, and slow journey times, make this undesirable.
It is to imagine a commuter coming home from Canberra to Melbourne on the HST only to crawl their way to Traralgon on the Velo or fight their way onto the all stations to Frankston and bus to Mornington.
So I would rate the urban fixed rail system as at least as much of a disadvantage to Australian HSR as to the American candidate areas, if not more so. It is not the much touted ‘population’ or ‘density’ issue; rather the defects in the existing system.
Thank you for your excellent analysis.
-Andrew from California
If Australia was fighting for funds under the Obama $8 billion then no pair of cities would be liekly to be prioritised for funding.
Melbourne – Sydney and Sydney – Brisbane are both on or above the 600 mile limit that was considered optimal for HSR. All Australian cities are low density so would struggle to meet that criteria too.
The best bets would be that HSR in Australia would be branched out from major cities to their nearby cities. IE> making the Regional Fast Rail a proper HSR between Melbourne, Geelong, Traralgon, Bendigo, Ballarrat and Seymour/Albury. or in Sydney linking to Wollongong,Newcastle,Canberra.
When those links are constructed you might start to consider the benefits of building the last 300km of track between Albury (on the Melbourne network) and Canberra (on the Sydney network) for sh!ts and giggles.
Realistically for these things to work a genuine integration of land use and transport planning needs to be implemented. I.E. Melbourne – Sydney is the most logical route. It needs better mid route attractors to keep the route busy. So pump some resources into making Albury-Wodonga a more important town (less focus on Ballarat, Geelong Bendigo more on the Dong). get the twin cities combined population up to 200,000 to go with 250,000 in Canberra and you’ve got a couple of good 300km markers along the route.
I also take to heart your point re: connecting services within the city. Again land use and urban planning play a role. creating the RRL in victoria to allow Melbourne-Geelong trains to run in under an hour is important. similarly the Melbourne Metro tunnel would free up capacity on the Caufield Group to run true expresses from FSS to Caulfield and Quadding to Dandenong would allow you to run expresses to Dandenong. The thing that makes these sorts of investments more worthwhiel is linking them to better land use planning such that there are enough people living on these transport spines through Transit Oriented Development to make the massive investment (and subsequent improved services) worthwhile. Similarly in the inner city Melbourne would be very well served by turning trams into light rail by having dedicated road space. this would significantly increase reliability of tram services connecting into a hypothetical HSR.
Melbourne-Sydney by rail now is under 600 miles.
600 miles = 960 km (approx.)
Melbourne-Albury is 306.5 km
http://www.vicsig.net/index.php?page=infrastructure&line=North-East%20SG
Sydney-Albury is 642.4 km.
http://en.wikipedia.org/wiki/Albury_railway_station,_New_South_Wales
The current route is 949 km and it there is probably room for straitening.
Riccardo, I don’t necessarily agree that European cities have a denser network of connections than Australian cities. Melbourne and Sydney are much larger, comparatively, than most European cities, so they have denser, more viable “headline acts”, and relatively dense and accessible inner cities of comparable population to the denser European cities. Australian regional networks look sparser (and are), but the scale is vastly different. What European network caters well for a very minor regional conglomeration like Traralgon 160km from the nearest hub? Conversely, Australian suburbs are comparably spatially separated, but the actual population is dense when looked at in relation to many European cities and their satellite towns. Or for that matter, many US cities, which have hollowed out cores, and substantial exurban fringe settlement.
Which is not to say the urban transport we have isn’t an impediment; I just think that the level of inter-CBD and fringe traffic that exists on Australian routes is high enough regardless.
Before we get into a discussion of the prospects for Australian HSR though, it is probably worth revisiting the scoping study from 2002.
One of the interesting aspects of this was that it rejected any substantial market share for the Melb-Syd or Syd-Bris routes, despite them being in the 600-1000km area. The reason being that HSR couldn’t reasonably compete with air, time-wise given the assumptions made about the line.
This relates to lachie’s comment: regional city politics are the bane of infrastructure investment in Australia. I am not sure what you meant by “pump resources into Albury-Wodonga” but the viability of projects like that are very limited, politically and economically.
But, the 2002 study still tried to stop at lots of little towns along the way, despite models showing as few as 30 passengers a day at some, and despite the massive cost to travel times that entails. Needless to say it was deemed cost-ineffective to create a HSR regional rail system.
Several assumptions in the report as well that worked significantly against the rail option when making a time comparison. Firstly, a pre-9/11 airport gate waiting time of 20min would now be considered ambitious; it is fair to say the modern door-to-door comparison should be four hours, not the three used in 2002 (rail gate-to-gate of 3 hours). Secondly, the speed within the cities was around 70-80km/h such that a quarter to half of time spent on the Victorian leg of the journey was spent within Melbourne; viable HSR needs some way out of town, which will be one of the major drivers of the cost. Thirdly, as above, the “limited stops” option was 5-6 regional centres, which significantly reduced average running speed (something like 10-15min per stop). Fourthly, detouring via Canberra costs close to an hour.
I’m inclined to think the worst way to plan a HSR is to try and string centres together. A viable Melbourne-Sydney line should go Melbourne to Sydney. It should, I think, stop once, in Albury to connect to other routes. No other town comes close to justifying it. From Albury other lines are viable: an Albury-Canberra-Sydney line that makes the detour; and regional lines that only use the HSR line for sections, much as the Paris-Dijon-Lyon route does.
The question though: how viable is a three hour 800km Melbourne-Sydney line?
Russ I think you hit the nail on the head with three points:
1. Airports now require much more time and hassle after 9/11 and are further inconvenient as they are typically located far from the CBD.
2. HSR needs to be fast inside cities as well as outside them. I have personal experience in Spain alone where the trains go fast through much of the city into the stations. This can be difficult and definitely costly where the infrastructure doesn’t already exist ’cause it would probably require a tunnel or much eminent domain acquisition of property.
3. Only a few stops. In Spain again I rode Madrid-Sevilla. It was a 2.5 hr journey and it stopped TWICE.
Your final question seems similar to the one we face in California. How viable is 2.5 hr SF-LA (FYI a flight is 1 hr airtime)? My answer is: its viable because it doesn’t have the hassles of flying (CBD to CBD, reliable, no airport security, no finding a way to the airport way outside of CBD, the list goes on.)
Cheers,
Andrew from Ca.
Andrew, thanks for responding. SF-LA is interesting because it faces some similar issues to a Melb-Syd or Syd-Bris connection in terms of viability.
By viable, I am sure they are viable in cost terms, if they came free. What I am curious about is:
Firstly, is it viable technically, to travel that distance in that time? Both SF-LA and Melb-Syd have to either skirt the hills via the coast, or traverse them, and that requires engineering works that potentially slow the trains. That is not my area of expertise, however.
Secondly, what is the cost estimate for that kind of route, the likely cost to government, the likely cost of tickets, the opportunity cost of new airports or upgrades to existing flight infrastructure? Those are project specific questions that aren’t easily answered, but ultimately HSR has withered here (and I gather in the US) because the answers haven’t flattered HSR.
saying that melbourne – sydney is under 600 miles by 10 miles means it is still at the top of the range of what the USA has called it ‘viable’ route options.
Russ: the density issue is definietly an issue. the highest density suburbs in Melbourne are lower in density than the average density of Berlin.
Spain is twice the area of Victoria but has 8 times the population. 4 times more dense. this is an issue on two fronts. 1. more dense = more people within catchment of train system. 2. more taxpayers to pay for a more compact system.
my suggestion of ‘pumping money into Albury-Wodonga’ is that, instead of say giving grants to Bendigo University for expansion give them to Albury University. Instead of giving tax breaks to locate a manufacturing plant for Melbournes metro trains in Ballarat put them in Wodonga. etc. doesn’t have to be earth shattering just a subtle shift in emphasis of the regional centres heirarchy to provide a mid point trip attractor (especially if as you say Canberra adds 1 horu to the journey).
I’d still build a canberra connection but with a bypass loop so that you might run every third train through canberra or similar as the Government traffic is significant.
but having said all that 900km of track even at $10 million per kilometre is probably breaking the bank of profitability and government funding. in reality it would be $20 million per kilometre almost certainly kills it.
I definitely think we’d be better served spending that $20 billion improving Melbourne, Sydney, Brisbane Metro systems. and spending maybe $2 billion on upgrading the intercity track (removing bottlenecks etc so that conventional trains on existing track coudl run at 150km/h). you’d then wait 20-30 years for another 1 millino people to live in both Sydney adn Melbourne hopefully around a denser transport network and spend the $20 billion then to make it a full blown 300km/h HSR route.
Thanks for your responses:
I agree with most points including that it would be better to improve intra-urban transport first.
Viability related to density – obviously it is a factor but not as much for HSR as people think. The very point of building new lines to avoid stopping at lots of little settlements – whether in Spain, California or Australia. Stopping at them slows the train down, and you are not going to slow a 350km/h train between 2 x 3 million size cities to serve a 5000 person settlement like Yass or Junee.
Certainly agree that anti-HSR forces tend to underplay the delays at airports such as check in and security (and the long walk to gates as airports get bigger) and understate the travel time (and rig the analysis in other ways)
If you assumed you could leave Melbourne CBD 1 hour before your flight takes off – you’d be foolish to say the least. Some days, even getting a taxi can be a drama, let alone the traffic delays, the queues to check in or get boarding passes (easier now with online checkin or machines doing it) and of course the security screening.
Then your plane is likely to be late. I say “likely” because in peak times, with Sydney as one of the city pairs, it is “likely”.
And they no longer quote flight times less than 90 minutes.
At the other end, again getting luggage off the plane used to be easy, especially with a ‘priority’ tag. I’ve yet to do it in under 10 minutes these days, I used to get it the moment I got to the carousel in the mid 90s.
So I think 4 hours is realistic CBD to CBD.
But the train MUST speed through suburbs at both ends.
I don’t think 200km/h, 300km/h or 350km/h particularly matters. I know this sounds daft but it isn’t.
Selection of route (very direct or circuitous), number of stops, and ability to hold cruising speed is what matters.
The crow flies distance from Melbourne to Sydney is 750 km. Even the Hume is more direct, going via Jugiong rather than Cootamundra.
A train that left Melbourne CBD and accelerated to 200km/h and held it for the entire journey, and decelerated 5 minutes prior to arrival in Sydney CBD, would do it in 4 hours.
Equally a 350km/h train would still take 4 hours if it stops at 10 stations, crawls to Broady and Campbelltown and goes via Canberra Civic and Albury for no particularly good reason.
And you could spend a lot more on the 350km/h version in terms of rolling stock capability, safety systems, track alignment and so on.
The 200km/h option could be a tilting train on fairly ordinary track, such as has just been laid at Beerburrum (and yes, only 1067mm apart!).
As long as it holds 200km/h for the ENTIRE journey by the most direct route, it would do it.
I worked this out for Mildura on Railpage but no one took the bait. The Velocity, running via a Swan Hill extension line or via Ballarat, quoted journey distance 500km and holding 160km/h for the entire distance- might just get there the same time as an air passenger in a turboprop plane would.
Now THERE’s a service for which there’s no market!
So what would a low-speed HSR with its own route look like?
Since it is hard to imagine a route going directly through the Snowy Mountains at a constant 200km/h, I’ll ask you to imagine one to Albury (I’ve posted on this previously).
The route out of Melbourne would need to be along the Epping or Upfield corridors, not the Broady and definitely not the Albion route.
All the current obstacles – road, river and housing, would need to be cleared, and you’d need soundproofing as well. So a tunnel might be the answer, at a price.
Once you get to Donnybrook or Wallan you then need a realigned route through the range that keeps the speed at 200km/h. Probably at a heavy tilt, and tunnelling through the two peaks at HJ and Broadford.
From Seymour there really are no obstacles, apart from perhaps Glenrowan and you would be sitting on 200km/h easily on the existing route. Overseas railways would probably already have cleared 200km/h running with all occupation crossings closed, a few remaining crossings with incredible safety precautions, and more grade sepping. You would probably want to do what the SG (east) already has, and go through Benalla and Wang on the east line with no station and yard.
Finally, you would encourage the Wodonga bypass and avoid stopping twice in Aldonga – alternatively measure the 1.5 hour timing from the first station at Wodonga west.
A fast journey to Sydney wouldn’t go through Albury though. That would be a separate option.
Now what is 200km/h techology? Well it is the XPT’s UK ancestor, which had it as standard highest speed. It is the Scandinavian express trains such as the X2000 which has visited here already, or the IC3 which nearly did. It is the Japanese predecessors of the QR tilt, which do 200km/h on the 1067mm gauge. Far from bleeding edge tech.
lachie, sure, Berlin has Melbourne’s population, with more than double the average density and a substantial network of interconnected nearby regional centres. So do Paris, London, Madrid and several others,which is why they form the central nodes for the European HSR system. But that is irrelevant to what I said. What I said is that many European cities have density gradients that aren’t substantially better than Melbourne’s, they are just structured differently.
Consider, based on the figures at demographia and wikipedia where req’d, Melbourne has CBD employment of 126,000, a total population of 3.5m and an overall density of 15.5 pers./km2. The 17 most dense municipalities, which make up a contiguous urban area around the CBD had (as of 1999) a population of 738,605, and a density of 32 pers./km2
Now consider the following:
Amsterdam: CBD Employ. 80,722, pop. 1.1m, dens. 26.5, district pop. 2.6m, district dens. 9.8
Munich: CBD Employ. 210,000, pop. 1.7m, dens. 36, district pop. 4.3m, district dens. 2.46
Milan: pop. 1.3m, dens. 71, urban pop. 4.1m, urban dens. 17.5
Marseille: pop. 0.8m, dens. 34, urban pop. 1.2m, urban dens. 11
To repeat, most European HSR destinations have density gradients that are comparable to Melbourne. Although I’ll concede that their p/t often works better for some (often significant) parts of their catchment, they also have substantial village centred populations that are no better off than the average Melbourne suburb.
This is why scale matters when you talk about density. Victoria has 4/5 of its population concentrated within 100km of a single node, and only three centres of even moderate significance that don’t lie between adjacent capital cities. Spain’s density at a national level is irrelevant. What matters is the density within a reasonable distance (around 50-100km) of the HSR nodes. Because Spain has a large dispersed regional population, a Victorian HSR system that serves 90% of the population is going to be a hell of a lot smaller than a similarly effective Spanish system. At 1200km, Spain has already built a larger network than Victoria will ever need.
Build a new HSR station at the old North Melbourne yards/steel link
Use the Bunbury Street tunnel onto the Albion Route
Build a new line along Keilor Park Road then tunnel to Tullamarine Airport.
The new line will travel out via Mickleham.
I believe that a HSR or 200km/h service must utilise Mascot and Tullamarine.
Also what do the Airlines in the USA think of Obamas HSR investments?
Riccardo, you raise an interesting point, which is that there is a tendency to equate HSR in Australia with a $20b mega-project and new lines everywhere, which is strongly at odds with similar long distance projects: the 40 years it has taken to build the Hume Freeway or the 20 years it took for the French to build TGV-Sud. Partly, I guess, because private consortiums have always proposed HSR, and therefore wanted full operational control.
What you describe looks closer to a 40 year project, picking pieces to fix as you go, slicing 5-10 minutes at a time. Which brings me to a question: why limit the train-set to 200km/h? Wouldn’t the minutes saved running 300km/h on the new Albury-Yass section and the more easily improved lines and bypasses be cheaper than the minutes saved making a 150km/h versus 200km/h tunnel through the hills and urban area?
Luznug. Apparantly in California the airlines and airports are supporting high speed rail. Modelling has shown that every major airport in the Calfornia area would be exceeding capacity by something like 2030 and they see this as a way to reduce demand for intrastate travel leaving them able to supply the interstate and international travel markets via air. (sorry I can’t remember where I saw this information for a source otherwise I’d link)
I agree with you Russ on going for the biggest bang for buck. A study could be completed to assess that.
If you could get 300km/h on a 500km stretch and 150km/h on the other 300km (through mountains/in city) you’d be at less than 4 hours travel time for Price $X. compared to 200km/h for all 800km at Price $Y. if $X is cheaper than $Y go with faster peak but less consistent overall speed. You could then over time fix some of those 150km/h sections as the value of doing so improves.
I think it is also fair to say that getting some of these improvements in place now for the conventional train woudl also be of benefit. Let the trains run 120km/h non stop and you’d at least be able to compete with the car traffic. as stage 1. of perhaps a 10 stage process to get to full HSR 300km/h route.
Thanks Russ
I’m not saying that it should be 200km/h – only that we don’t need to fetish the technology (off the shelf is fine), that we shouldn’t be saying that aviation is unbeatable (because it isn’t) and that holding top speed longer makes a bigger difference than going for ridiculously fast capability.
I understand in Europe they will push the boundary of 400km/h but they are talking about including longer and longer journeys, for example, Paris to Barcelona or Berlin which will really extend the reach of HSR into aviation territory. And then, they put their money behind these things as well.
Even then, in Barcelona there is a fight about putting the AVEs under the Sagrada Familia church – which is a fair enough issue. My first reaction is why are they building a dedicated tunnel under the city merely for the AVEs?
But then you realise that could be 7.62 minutes saved in the journey time to Paris. Add this to all the other very expensive tunnels, bypasses, flyovers and so on enroute, you’ll get a journey time to Paris that actually competes with aviation – a feat considered impossible when the original Paris-Lyon line was built. Back then, Lyon was thought to be the feasible distance for beating aviation. We are now talking about somewhere twice as far away.
I saw a diagram of the isochrones from London and Paris at high speed – the reach is getting further and further. I expect a solid London to Frankfurt time of 4 hours when the tunnel from eastern Belgium to Germany is done. London to Frankfurt is a huge business aviation market and HSR could completely trash it if the investment is made.
And routes like that will beat the doubters.
Who’da thought – 1820s flanged wheel on steel rail, 1880s electric motors and power transmission equipment, refined to this level? I look forward to what genuinely better technology can do.
Poorly translated by Google German to English:
The rush hours will be by the commuters in the really long distances for imaginary ICE trains are already closely. In an elaborate thesis on the ICE as a commuter and Vorortzug? “Has been a student at the University of Trier in the renowned traffic expert Heiner Monheim deal with the issue – and, supported by the Westerwald area, a representative survey travelers organized. The result: From 300 passengers in Montabaur, to the morning rush hour there once or evening there exits, gave 288 to every day with the ICE to work commute. The majority worked in Frankfurt in the town or at the airport.
288 people commuting daily on the Frankfurt to Cologne ICE! Won’t pay for the service, but pretty amazing nonetheless.
Thanks Lachie,
What i don’t understand with Ricardos posts is why can’t we use the existing Standard Gauge line for a 200km/h max service between Melb-Syd?
ARTC have released a document on the MELB-BRIS route and hopefully this will upgrade as much of the MELB-COOTA formation as possible so future services can accommodate the prospect of 160/200km/s pass services.
In regards to a serious HST it really should only service the 3 major centers MEL-CAN-SYD.
All the rest can use the medium speed trains:ie 160/200km/h.
PS: any updates on the proposed V/Line service to wagga wagga?
Alright, having read all the comments on this post, I have some views of my own to express on the topic, and especially how it has shifted toward Australian HSR.
First of all, I don’t think that the focus should be on getting a VHSR line from Sydney to Melbourne – I cannot see it being a success in the short to medium term for a number of reasons. What should instead be done is improving the existing infrastructure, like a number of commenters have said – so what would I do?
Simple; I would start off by examining what sort of alignment modifications would be necessary to allow 200km/h top speed operation on the Sydney to Melbourne, Sydney to Canberra and Sydney to Brisbane corridors. Once that has been established, then I would have a simple set of goals established to break that project up into a number of smaller ones – such an approach has been done by roading and it has meant that funding can be easily obtained and the benefits enjoyed (as opposed to rail which appears to have tried for big bang sort of projects).
I’ll use Sydney to Canberra as an example of how my goal system would work. Presently, an Xplorer takes 4 hours 23 minutes from Sydney to Canberra. I would have a system where you work to cut that to 3 hours 45 minutes, then 3 hours 15 minutes and so on and so on until you reach your desired journey time. To reach the goal times, you would obviously look for projects that would speed up the trains, so if you have a 70km/h curve that is slowing things down, you would examine your desired route to see how that curve could be eliminated and your deviation created to fit in line with the desired route. Do that several times, and you will see your journey times slowly drop and passenger counts slowly rise. (BTW, if anyone is wondering why I am in favour of journey times of :15 & :45, that means that with a fifteen minute turnaround, you can have whole or half hourly frequencies).
Of course, you would probably find that you would be strongly competing with turboprop routes before long – and that is a good thing; why on earth should there be a Dash 8 taking up a valuable slot to go from Sydney to Port Macquarie, when that slot could be utilised by a 737 from other parts of the country, or a 320 from New Zealand, or whatever.
The other main thing is that you need frequency for such a scheme to work. Why on earth would I take an Xplorer from Sydney to Canberra when I only have two services a day to pick from, while the airlines probably have far more services? If you managed to get Sydney to Canberra down to 2 hours 45 minutes (and I don’t think it would be particularly difficult – it is only a trip of 330km after all) and had four Xplorers devoted to that corridor, that would be one service every hour and thirty minutes, making it attractive to the passenger and probably dealing a blow to the airlines on that pairing. Of course, it would be different if you were talking Sydney to Melbourne or Sydney to Brisbane, where you could get away with a handful of trains a day.
Any thoughts people on Sydney to Brisbane?
Apart from the rapid growth of SEQ I would have thought the issue here is many more viable stops en route (of course you wouldn’t want to stop at them for the headline act, but you could run other trains to pick these up)
Roughly speaking
Central Coast – 100,000
Newcastle/Lake Mac/Maitland/Port Stephens – 500,000
Taree-Port Mac-Kempsey – 100000
Nambucca Heads-North Coffs – 80,000
Grafton and Yamba – 50,000
Casino-Lismore – 50,000
Ballina-Byron – 40,000
Gold Coast/Tweed – 400,000
A well placed station at each of these and supporting buses would make another 1 million people available en route. Furthermore, some of the intermediate runs would be much more competitive with air than the headline act.
And road transport on this corridor is poor. I suspect the only real hindrance to making this the primary route, rather than Syd-Mel, is the more difficult topography.
Syd to Newcastle being absolute hell, and also some rough patches where the Great Divide comes close to the sea, at Coffs and in the Tweed and Richmond Valleys. There would need to be numerous bridges over estuaries that would not be needed on the Melbourne route.
As I have complained elsewhere, there has been no concerted effort to build up the non-tourist economies of these places, and tourist demand, while important, will not be the driver of the premium fares that this service would require. Seeing the silver lining, at least surplus capacity at odd times should be easily offloadable on the tourist market, unlike Syd Mel (fancy a cheap fare to Wagga?)
Competing with air? Might be a lost cause on the GC due to excellent air services (though these have gone down market and might not be appealing to the business traveller).
Elsewhere en route we have either turbo props (Grafton, Lismore and so on) or limited frequency jet services (Port Mac, Coffs, Ballina) that a more frequent rail service could trump. These regional cities won’t experience ground congestion or checkin delay at their end but will at the Sydney or Brisbane ends.
The Central Coast might find the service more appealing than a drive to Sydney Airport, but are now catered for by a reasonably easy drive to Williamtown and flights to GC, Brisbane (and Melbourne) from there.
Just to show population is less a factor than wealth and status, we can imagine a lesser use of the service by Novocastrians (500,000) than Canberrans (400,000) if lines passed their respective cities. Compare the following stats: average weekly earnings, highest qual attained, % of workforce as managers and professionals, with the national or NSW averages shown. Canberra is higher than both on all scores.
I would be extremely wary of extending any high speed rail network beyond the Sydney-Canberra-Melbourne corridor until the viability of the core route (and stops) were established.
But if one were inclined to head north, there are really only three stops that you would realistically consider:
1) a Newcastle/Lower Hunter stop
2) a mid-north coast stop
3) a NSW/QLD border/Gold Coast stop
If the raison d’etre of high-speed rail in Australia is to be time-competitive with air transport, then you would realistically only stop where you could reach nodes of the domestic air network or connect to an international air corridor. Thus at Newcastle, you could tap city and its hinterland (poorly served by an airport near Port Stephens) and at the Gold Coast tap a major domestic trip generator (and fledgling international air hub).
But personally, I wouldn’t bother heading north of Sydney for at least a decade after the trunk route between Sydney and Melbourne (via Canberra) was constructed and at least paying off its operational costs (if not starting to service the debt from its capital costs).
LS
Thanks LS. Good point about stops. The midway stop would presumably be Coffs.
For all the grief about Sydney Airport, Newcastle has not rushed to catch the overflow. Sure it is 2 hours but Narita can take that long from Central Tokyo by road.
At the end of the day I suppose it is about not believing what you read in the papers; some would have Sydney as bursting at the seams but what I see is the airlines squatting on slots and not being forced by price to ration them efficiently. Even an airline like Rex will horde the regional slots rather than let someone else have them. Australia’s political system being so weak; no-one will confront the rent-seekers and speculators to stop them.
Even I had a naive hope that Albo was tackling them the other day by floating Richmond but then rushed straight to print to say everything was OK and Macquarie could have everything they wanted.
Can someone explain to me why we are so defeatist about mountain ranges? I keep coming back to this issue as we seem to think mountain range=no railway and ignore that Switzerland and Japan are some of the most rail-covered places on earth and also the most mountainous.
I know we do v-shaped valleys (riverine) while Switzerland does u-shaped glacial valleys which are much more rail friendly, but still, they end up with big, long, expensive tunnels, which we seem averse to.
I’m sure the Canberra to Melbourne leg would need to punch a hole through the northern Brindabellas but that must be easier than heading south through the Snowys – and eventually you hit the south west slopes and easier terrain.
Ric I tend to agree that the Sydney to Brisbane route is a better chance than a Melbourne-Sydney route thanks to the population centres.
That route could grow organically as a Brisbane-Gold Coast fast rail and a Newcastle-Sydney fast rail and then be extended to ‘fill the gap’. over time. Could conceiably be staged as a couple of $2-4 billion dollar projects to start. and then a third $2-4 billion project to fill in the gap. Also the two ’seed’ projects can be constructed with single state government support.
where as the Melbourne-Sydney route is a $10 billion project straight up – harder to find funding for that against competing priorities. and comparitively it requires a greater level of up front cross state cooperation.
John-ston, Ric and Lachie,
I disagree the proposition that Sydney-Brisbane is a better prospect for HSR than Sydney-Canberra-Melbourne. Especially if your goal is to take on and win market share from air on the East Coast corridor.
As this 2007 data shows, Sydney-Melbourne is one of the busiest air routes in the world in terms of number of services per week. While this may not be the best measure of demand (how many flights run half full?), this does at least show that there is a market for sufficiently high-speed rail, running a limited stops service between Melbourne and Sydney. Note how in the above link, the top internal European destinations such as Paris-Nice (925km), Paris-Toulouse (674km) and Madrid-Barcelona (621km) are comparable to Sydney-Melbourne and Canberra-Melbourne distances. Rail is roughly competetive in door-to-door travel time and mostly on price as well.
The inclusion of the national capital on a route between Australia’s two largest cities would appear to be a no-brainer, as there would be significant levels of government and government-related business that would flow from air to rail, especially on the Sydney-Canberra corridor.
The hard part would be deciding on where to place the additional stops to capture important inland passenger markets who would currently drive or fly to Sydney/Canberra/Melbourne rather than use a train (if that was something that was desirable). At a bare minimum, one stop at Albury-Wodonga (c. 100,000) would be useful in tapping a significant market of government travel to Canberra (due to the complex of military bases around Albury-Wodonga) as well as Sydney and Melbourne. Rail and coach shuttle services could run from NE Victoria and Southern NSW into the station at Albury-Wodonga to hub passengers from the hinterland.
There’s more to post on this, but I have to go now.
LS
Just to be clear, I’m not saying Syd to Bris is a better option than Syd to Mel -only it shouldn’t be written off either
I outlined what I see as the topographic issues. I don’t regard the breakout of the Sydney basin to the SW up to Wingecarribee as a problem as the freeway manages it comfortably (and indicates which way to do it) and through the Kilmore Gap should also be a reasonable route.
You can get into Canberra from the north OK via the Lake George route but the real pain is getting out of Canberra to the south. Either you swing hard round towards the Brindabellas and try to push through a long tunnel out into the Burrunjuck area – or take more easier ground towards Cooma but then what – you’re trapped unless you either head for Gippsland over the Alps or push through the worst of the mountains towards Albury.
So you can do a Syd to CBR and a Syd to Mel (not via CBR) route relatively easily, the Mel to CBR stretch is difficult. It might be easier to be a bit creative and say compromise on the directness to CBR by building a sort of 3 pointed star – a direct route Syd to Mel via the quickest easyish way, through Yass and Jugiong like the hwy, a second route via Murrumbateman (again, the road points the way) but a direct route via Lake George (again, the road points the way). At the cost of some extra trackage you avoid the really nasty punch through the Brindabellas but still provide a direct rail service from Syd to Mel in reasonable time.
Otherwise I have no other comments on the market. LS is right about tackling the City pairs in patronage order – although you need to add Syd-OOL (GC) and Syd MCY (SC) to the BNE figures to get a true response.
Ricc. Back 30-odd years ago, the National Committee on Railway Engineering (the forerunner of the RTSA which, by the way I encourage everyone to join!) in its Bicentennial High Speed Rail Proposal put forward the concept of a ‘T-Line’ deviation between Goulburn and Yass Junction with a triangular spur to North Canberra. Designed with a 1-in-75 ruling grade and 1200 metre minimum curve radius, it would have cost $127 million in 1981 dollars and allowed 160km/h running which could have cut the Sydney-Canberra journey time to 3 hours.
This was one engineering solution to the Canberra entry/exit problem you’ve raised and it would have followed generally the current highway/railway alignments. I’ve got to find this report in the work library and have a read!
LS