Commentary: The case for High Speed Rail

Posted: November 2, 2015 in Transport, Urban planning
Tags: , , , ,

VIDEO: Shaun Micallef: Australia’s NBN proposals

A High Speed Rail (HSR) network connecting Melbourne, Canberra, Sydney, and Brisbane has been shown to create more benefits than costs while fares would pay the operating costs of trains on such a network. Yet there seems little appetite in the government to build one. Initially, this may appear to be due to the high initial costs of $114bn to build it; but on closer inspection it may be necessary to re-evaluate the way we look at HSR as something for regional Australia rather than just for travel between the capital cities.

Cost benefit analysis suggests that HSR is worth building. The Phase 2 Study found that HSR has a BCR (Benefit Cost Ratio) of 1.1 when a 7% discount rate is used. Thus, the benefits of building HSR are greater than the costs. However, the bulk of these benefits would accrue to the users of HSR in the form of time savings. In 2028 dollars, HSR users would receive $141bn of the total $180bn of benefits that HSR is expected to create. An additional $14bn of benefit go to operator benefits, which help to pay the operating costs and up to $16bn (14%) of infrastructure costs, but this still leaves the government paying for $98bn (86%) of infrastructure spending on HSR. (Sources: HSR Phase 2 Study Executive Summary, pages 42, 9).

This means the government would be subsidising the travel costs of HSR users to the tune of $98bn. This is unlike the NBN, where the initial infrastructure spending is expected to be eventually recouped. The majority of users (65%) would be leisure travellers; however the bulk of the benefits would be realised by business travellers (the remaining 35%), who would receive $93.6bn of the $141bn of benefits. That is because of the higher value attached to their time. This point was raised by Alan Davies in Crikey, who proposed that “if the [time] saving is so valuable to business travellers, they should pay the full cost of constructing the line”. The Phase 2 Study even recognises that these travellers would be willing to pay a higher fare:

“Increasing the cost of fares would increase the financial returns and reduce the funding gap, although doing so would reduce the number of people using the system. Even so, the economic benefits of the program would remain positive.”Source: HSR Phase 2 Study Executive Summary, page 9

Higher fares would have the benefit of reducing the government’s cost below $98bn. However, it would do so by impacting leisure travellers the most, with many choosing not to travel on HSR. Business travellers would mostly still continue to use HSR, but the loss of many leisure travellers would see the total benefit of the project reduced. Although the Phase 2 Study claims the economic benefits in such a situation would still remain positive (i.e. a BCR greater than 1), this may be based on the less conservative 4% discount rate, rather than the more conservative 7% discount rate that is normally applied to transport infrastructure projects. The 1.1 BCR that a 7% discount provides is dangerously close to falling below the 1.0 required for the project to be economically viable. Therefore, as it stands HSR does not appear viable without a $98bn government subsidy, most of which would flow to business travellers who least need government welfare.

An alternative perspective

The Phase 2 Study emphasises that HSR accrues more benefits as time progresses, given the growth in population. If governments work collaboratively and actively to preserve potential HSR corridors then HSR cost increases should be limited. Therefore, HSR becomes more viable as time progresses with benefits growing faster than costs.

Since HSR gains most of its benefits from additional users, one way to increase the viability of HSR is to add additional population to the corridor. This would be much easier to achieve around the regional stations where constraints are much more limited than in the major cities. HSR could act as an enabler, allowing a greater number of people to live and work in regional areas without becoming isolated from those services only available in major cities. The Phase 2 Study’s assumptions of modest population growth in regional towns situated on the HSR route show that this was not considered as part of the feasability for HSR.

In fact, the Phase 2 Study finds that HSR will produce $73.2bn in benefits from intercity travel, more than the $67.5bn in benefits from regional travel (Source: Department of Infrastructure, page 43). This finding that most of the benefits accrue from intercity travel rather than regional travel suggest that not enough is being done to massively develop regional Australia. HSR provides this opportunity which in turn makes HSR more viable.

Proposed East Coast High Speed Rail alignment. Click to enlarge. (Source: Department of Infrastructure, page 17.)

Proposed East Coast High Speed Rail alignment. Click to enlarge. (Source: Department of Infrastructure, page 17.)


This idea was floated by the ABC show Catalyst in its 4 December 2014 episode “Future Cities“, in which Dr Julian Bolleter says:

“So, what we think is really important, as the capital cities grow beyond mid-century, is that we begin to think not so much in terms of mega cities, but mega regions. Essentially, it means chains of smaller cities connected with very good public transport infrastructure. So we could conceive of a mega region running from Brisbane to Sydney through Canberra to Melbourne which is bound together by a high-speed rail link, and those cities will have access to affordable land, and they’ll also be able to be designed from the ground up around the principles of 21st-century sustainability. High-speed rail can travel at about 350km/h, so there’s no city along this mega region that is further than two hours commute on a high-speed train from a capital city.”Source: Dr Julian Bolleter, ABC

Achieving this would require us to rethink how HSR would work in AustraliaShadow Transport Minister Anthony Albanese recently wrote on HSR in which he concluded one well thought out and one not so well thought out point. His statement that “people could live in regional Australia and commute to work in the city” was not well thought out; this is true only to the extent that people can currently commute to work in Sydney by flying into Kingsford-Smith Airport, or any other major city airport. However his point that “companies could establish themselves in the regions, taking advantage of lower costs but comfortable in the knowledge the city was a short train ride away” hits the nail on the head.

In order for HSR to be a success in spurring regional development people need to live, work, and spend leisure time in the same place. Employment opportunities as well as services that are needed on a day to day basis such as health and education would be provided locally. But the existence of HSR provides convenient access to services which are not needed day to day, such as medical specialists or major cultural festivals.

If the 1,700km HSR corridor had a station every 100km or so along major regional cities, and these cities were allowed to grow to 800,000 residents each (as Dr Bolleter suggests in the Catalyst video), then it would be roughly equivalent to a doubling of the existing populations of Sydney and Melbourne combined. Once it becomes prohibitively expensive to retrofit the necessary infrastructure into our growing major cities, it will become cheaper to build it in regional cities even after the cost of HSR is factored in. Australian cities have not reached that point yet, but it remains a question of when rather than if they do reach that point.

  1. Marcus says:

    “People need to live, work, and spend leisure time in the same place” – you have hit the nail on the head there!

    Spending billions of dollars so that people can live in the country and take a subsidised high speed train to work is a folly – but enabling people in ‘second tier’ cities to access specialised services only possible in a city of millions is far more sustainable.

    An existing Australia example is the V/Line long distance rail network beyond the commuter belt – with only three or so a train on each route, it provides a level of service that connects regional cities to Melbourne, while still encouraging a local focus.

  2. Once again, we talk about new-build HSR.

    What about spending a fraction of the cost of an East Coast HSR network and upgrading the existing rail infrastructure. Realignments, tilt train passenger stock, electrification, grade separation and a greater separation of freight and passenger rail would provide a significant improvement over what we have at the moment. As the Victorian RFR shows, it doesn’t have to be 350km/hr to be attractive. 160 – 250 km/hr with electrification would be a game changer – not everyone is a time sensitive traveller that must get there on a 350km/hr service!

  3. Oz says:

    Completely agree that for HSR to work people need to live, work, and spend leisure time in the same place. Too many people have suggested it would allow commuting which is ridiculous as the fares would be too high for most people.

    In countries with HSR where it’s used for commuting eg France & Japan, employers subsidise those fares which is why it occurs. It won’t happen here.

  4. Alex says:

    I broadly agree with the post and the previous comments, but there are a few other points to consider:

    – Using HSR to underpin growth in cities along the corridor is definitely the way to go but the idea that one (let alone several) of these could grow to populations of 800,000 simply isn’t feasible within the next century or so. Having said that, the growth of cities like Albury and Wagga to 100,000 to 200,000 each in this timeframe is a more achievable target which would in turn assist the viability of the HSR.

    – The decision to proceed with a second airport in Sydney could set back the prospect of HSR by several decades. A second airport will increase capacity and bring down the cost of air travel, and in addition the Federal Government is unlikely to fund two mega-billion projects in the same timeframe. Conversely if the HSR had been approved it would have had the potential to defer the need to build an airport first by providing direct competition for the busy Sydney-Melbourne corridor (among others) as well as potentially allowing Canberra airport to in effect become Sydney’s second airport.

    – There is however one potential bonus of a second airport in Western Sydney in relation to HSR. There is already a strong push to consider some sort of high-speed rail corridor from the airport to the Sydney CBD via Liverpool and Parramatta as well as or instead of the SWRL extension. If this was committed to it could also be utilised by the HSR to provide access to the CBD, either directly or via an interchange at the airport, thus removing the need for two dedicated corridors into the CBD for the HSR from the south and north.

    – Given all this the development of a “moderately fast train” between Sydney and Canberra by upgrading the current infrastructure and using new rollingstock to achieve 160-200 kph service appears to be the most realistic option and could as Martin suggested be done for a fraction of the cost of a full HSR. This could use existing lines to reach the CBD or run via the proposed airport link mentioned above. A 2 to 2.5 hour trip to Canberra and similar reduced travel times to intermediate destinations would help spur population growth in cities like Goulburn and provide the basis for a a full-blown HSR in future.

  5. Todd says:

    Can anyone confirm that the most recent study also assumes that airlines will not reduce their fares to compete with the HSR. If it is true, then this is a major floor in the calculations.

    Regarding regional access, I think it would be much cheaper (albeit recurrent, as opposed to capital expenditure) if the Govt just more heavily subsidized air travel to regional centers.

    Finally, if you would allow me some indulgence to draw a line on a map, to cater for future population growth, we should look towards the southwest. If a shorter HSR between say Goulburn and the City could put it within commuting timeframes and expense, then there is plenty of scope in those areas to cater for population growth.

  6. Alex says:

    I agree that the airlines would be likely to discount in the face of competition on major inter-capital routes though it is less clear how they would respond on the regional routes where the business is more marginal even though the fares are very high. I suspect that for some of these road-based transport would be the major competitor.

    As previously stated I also agree that medium-distance and moderately high-speed rail may be the best way to go at least in the short term and you are right that most of the gains in terms of creating population growth outside the major capitals is probably within the closest 200km or so of the corridors connected to them, as demonstrated by the Victorian examples of the rail connections to Ballarat and Bendigo.

    One of the problems in Australia however is that it is hard to get Federal Government involvement in HSR (or MSR) unless it crosses state borders because of historic assumptions that intrastate rail infrastructure is a state responsibility, combined with the probability that the other states would cry poor if the Feds subsidised a new line wholly or largely within NSW. One way around this would be to push the case for Federal involvement in a program to develop moderately fast rail on corridors connected to Melbourne and Brisbane as well as Sydney. This would seem more even-handed than starting with Shdney alone but could stil,provide the foundation for HSR in the longer term.

  7. Anthony says:

    To be viable a HSR relies on been competitive with flying. Sydney to Melbourne will take somewhere between 3 to 4 hours minimum which is lnger than flying. But add stops in mid size cities along the route and the travel time blows out significantly. High speed trains can’t go straight to maximum speed so stopping and starting at each town will significantly increase the travel time making it uncompetitive (it would also be difficult to run non stop services on the same line due to these slower trains).
    A better option would be for high speed (or even express trains) to run from nearby regional cities into Sydney and Melbourne – without having to connect these cities. High speed trains from Newcastle and Central Coast into Sydney would be more sensible if they provided much a shorter commute and had a significantly lower price tag.

  8. @Anthony – The current HSR Study has express services travelling Sydney to Melbourne in 2 hours 44 minutes (not 3-4 hours), which is faster than air travel on most journeys when measured door to door. It also recommends running a mix of express and local services, which could easily be done with some timetabling and a few passing loops.

    Fast train services as commuter services to Sydney (as is the case in Melbourne with V-Line) could be a big improvement. But this is different to what HSR would aim to be – not a commuter service. They are not mutually exclusive, nor is one a substitute for the other. Though I would imagine that fast commuter trains to Newcastle/Gosford/Wollongong are more achievable in the next 10-20 years than an East Coast HSR network is.

  9. JC says:

    Sydney-Melbourne in 2h44 would only be possible with very high spec lines and trains – and which would be expensive – and even more expensive to build in the flexibility to provide stopping services without slowing up the non-stops. And even then it would only be touch and go v. flying, and would be unlikley to have the patronage to justify the cost.

    I’m not sure that the alternative of an upgrade to the exiting tracks is a goer. The more railfan types that contribute to this blog may have bettr information – but my sense is that the 1850s Sydney-Melbourne line is so far from a standard western european non-high speed line (typically with average speeds of 200 km/h) that we may bas well start from scratch with a new route.

    But new lines Bribane-Sydney-Canberra-Melbourne at 200 km/h standard would still be much cheaper than HST, and at just over an hour for Sydney-Canberra, it would kill off the flight alternative and Sydney-Melboune in 5-ish hours could take a lot of air business for leisure and less time sensitve travellers.

    It would also be easier to run freight at these speeds than full HST – roll-on waggons like those used in the Channel tunnel could save time, money and GHG emissions with minimal disruption to the profits of the haulage industry.

    Another variation on the theme would be to run at half speed at night (say 23.00-6.00), making it even friendlier for freight, sleeper trains and to save fuel.

  10. Glenn says:

    Great post.

    For several years I have been thinking that the way to make a start on an HSR network is to build a regional network based on Sydney, with all lines built to the highest standards so that there would be no problem with subsequently extending the network to Brisbane and Melbourne.

    The first line would be from Sydney to Newcastle with a couple of stops along the way – maybe Hornsby and Central Coast. The second line would be from Sydney to Canberra with stops at Campbelltown/Macarthur, Southern Highlands and Goulburn. The third line would be from Sydney via Campbelltown/Macarthur to Wollongong.

    Some options which should be considered are:

    – having a major station at Strathfield with the lines to Newcastle and Campbelltown/Macarthur running north and south from there (This would allow the current corridors through Sydney to be used to some extent.)

    – the trains being dual voltage (1500V DC and 25kV AC) so that trains continuing on to Central from Strathfield can run over the existing tracks until volumes justifies building a new dedicated line (This is the way it is done with the TGV in France.)

    – the line from Strathfield to Campbelltown/Macarthur going via Badgery’s Creek (A slight detour admittedly and taking the route away from the existing corridor, but with obvious benefits in terms of connecting the network to the new airport. It may also be possible to run shuttles between Badgery’s Creek and Central.)

    – some trains coming in from Newcastle continuing direct onto Badgery’s Creek and south (Canberra or Wollongong) without going to Central – and the reverse for some trains coming from the south

    – running both express and all stations trains with the number of tracks opening out to four at the stations where express trains don’t stop thereby allowing express trains to pass while an all stations train is pulled up at the station (This is the way it is done on Japan’s Shinkansen and would allow there to be more stations along the route without significantly slowing the express trains.)

  11. Alex says:

    An interesting article on the route selection principles and process for HS2 in the UK – the proposed HSR line from London to the north:

  12. Alex says:

    As JC suggests the existing lines are so far below spec that upgrading them to say 160 to 200kph standard may not be economically viable, but I’m not so sure about his argument that if we have to start from scratch then building 200kph standard track is that much cheaper than building HSR. This comment in the article about the planned HS2 in the UK which I linked to in my previous post suggests otherwise:

    “We had to study strategic alternatives, so we asked what would be the effect if High Speed 2 were built as a conventional railway, another 125 mile an hour railway? And the answer was we’d never build it because the benefits would only be about half but the costs would be 90% because we’d still be building the tunnels, the bridges, the viaducts, the overhead line, the track and the signalling. In cost terms, the difference between a conventional line and a high speed line is less than 10% for new line but we have double the benefits, so it’s not a difficult decision.”

  13. JC says:

    @Alex.Fair enough – I saw and noted that as well and I’m happy to accept that this may be the case – it would be interesting to get a good picture of what drives the cost diffeence between a 200km/h railway and a 350km/h railway if you are building from scratch. Maximum gradients and minimum curve radii presumably have an effect (but by how much?) as well as signalling, trains, length of passing loops and (I suspect a major factor) line safety and security – much higher standards for keeping people, animals etc out.

    There is also a diversity question – HST might be fine/justifiable if you can justify using it exclusiveluy for intercity passenger traffic (as for London-Paris, Tokyo-Osaka, etc. But 200km/h would be easier to mix in freight, commuter trains etc.

  14. Alex says:

    @JC – you raise a valid point about the versatility of 200km/h to serve multiple uses including freight and suburban commuter trains. This doesn’t appear to be an issue with HS2 in the UK because most of the existing track appears to be to 160-200km/h standard and the problem is more one of demand and the capacity of the existing infrastructure.

    In Australia if a 350km/h system is ever built I think we can assume that most freight would stay on the existing albeit run-down network while almost all passenger rail traffic outside the current commuter belt around the major cities would switch to the new system. Unlike the UK where there are commuter trains almost everywhere, most commuting takes place within 100km or so of Sydney Melbourne and Brisbane. There would be some conflicts within the metro areas of these cities but either a 350 or a 200km/h system could use conventional tracks in these sections on an interim basis, albeit with some impact on travel times.

    Over about 100km outside these cities commuting is virtually non-existent. In fact, while the emphasis should be on linking major centres along the corridor rather than commuting, HSR could support the development of some new commuting patterns from centres within the 100-200km band, for example Goulburn and Mittagong. This traffic might not all be one way however, nor necessarily focussed on the Sydney CBD. There could be commuting traffic for example in both directions between Goulburn and the Menangle Park development recently announced by the State government in South West Sydne, or between Goulburn and Canberra.

  15. PMeilak says:

    When the HSR report came out a few years ago now I found it the most depressing document that was designed to make HSR fail. (it succeeded in its design very well)
    As expressed above I would like to see a realistic report on electrifying intercity rail and improving the speed to 200km per hour. Its not that hard as the British have done it for the last 40 years.
    Passenger trains at 200 km/hr and freight at 120 km/hr would make a serious dent into road freight on the Hume. (good/bad for the economy needs to be discussed as a whole not special interest groups). Lets hope for a real report on 200 km/hr electrified rail between Brisbane, Sydney and Melbourne so the discussion can begin.

  16. JC says:

    @Alex.You are probably right – a 350km/h system would not carry freight – and there would continue to be negligible freight on the major SE corridors as the “existing albeit run-down network” would not be able to compete with road. For this model to deliver any real difference we would need a $multi-billion HST as well as quite a few more $billions to upgrade the existing netywork. I don’t think Australia can afford both.

  17. yetihunter84 says:

    Low-range HSR (say, 200km/h) is remarkably practical – you could do it while still re-using much of the existing corridor. 200km/h is much more forgiving with curve radius – you only need 1000m for tilting trains, or 1800km for TGV-style trains. Compare this to the 5km you need for 350km/h (or the 7km specified in the 2013 report just in case we ever needed to get up to 400km/h!!) and you begin to see why the cost was so phenomenal.

    The other great benefit to 200km/h is that diesel-electrics are still practical at those speeds. That saves you several million dollars a kilometre in not having to electrify the thing.

    As Martin says, 200km/h would still be a game-changer. The XPT averages under 80kph on most routes. If an express Sydney-Canberra pretty-fast train could average close to 200km/h, we’d be talking about 90 minutes, not 2hrs+. And that’s a whole lot better than either car or plane or the existing 4 hour 30 minute VST.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s