This blog has discussed at length the fate of High Speed Rail in Australia. One series of posts aimed to cover this blog’s own proposal – and a radical one at that – for Melbourne to Sydney HSR, but that only got from Melbourne to Goulburn and awaits continuation.

The radical proposal has HSR coming under the Snowy Mountains at Talbingo (top photo), to Canberra and up Lake George (bottom photo). Beyond Goulburn is yet to be plotted.

Political activity around High Speed Rail appears to be like watching a pot of very thick soup warming up – an occasional bubble of heat bursts the surface, but mostly it is a boring scene of inactivity. It was thus since 1984 and the first of them, the Very Fast Train (VFT).

Honestly if they had just built the godforsaken thing back then, the owner would have:

a) lost a lot of money and probably opened it during the 1991 recession when it would have looked quite empty but,

b) during the pilot’s strike of the late 1980s, wished they had it and would definitely confirmed it was a good idea,

c) run it for nearly 30 years to this point, upgraded the rolling stock several times over and it would be making a bucket load of cash and finally,

d) less desperate need to build a new airport for Sydney.

Watched pots, of course never boil, so maybe it would be better to leave Australia and come back in 20 years to see if High Speed Rail has emerged.

This blog has maintained through the years that the major impediments to High Speed Rail are NOT the ones usually thrown around like lack of population (at the ends of the route, or en route), nor any cultural aversion of the Australian people towards rail.

The three main factors this blog identifies, in no particular order are:

• The journey lengths between the capital city nodes, which are really on the edge of ideal journey times to compete with air city-to-city;
• The lack of ‘last mile’ (or ‘first mile’) connectivity between any HSR and the places people really want to go – their own homes; business addresses and other likely drawcards which suffer from poor local public transport; and
• Cost of construction, which is impacted severely by poor topography near the cities.

This last point will be the topic for this post.

Not only has Australia struggled with High Speed Rail, but so has the USA. While arguably the USA does already have some very low end High Speed Rail between Boston and Washington DC, the classic development of High Speed Rail has not proceeded in the USA despite some of the obvious factors that should be driving it. These include:

• Very high populations and population densities in several corridors including the North East; Great Lakes; Florida; California and in Texas;
• Congestion on parallel roads and congested, unsatisfactory aviation alternatives;
• In some of these corridors; favourable topography.

Technically the USA started into the medium speed rail domain consistently with the Metroliners in the Penn corridor in the 1960s.

What are the factors repelling high speed rail then?

• Political and ideological, which this blog would characterise as stretching along a continuum from the classical liberal idea of avoiding government subsidies, right through to a nationalist/fascist/contrarian idea that if other countries find HSR works well it MUST be opposed in the USA
• Political corruption related to the previous point
• Lack of firm and consistent government policy towards HSR, even where the economic case is strongest.

Once upon a time, US railroading was respected, especially by the big business-leaning right wing. Now they hate it.

This blog has followed the progress of the Texas proposal, which seems to be right out of the textbook of being a sensible business idea, along a route of ideal length, plenty of potential to squeeze traffic away from road and air, favourable topography and good population.

Texas rail concept shot of Japanese 700 class on elevated viaduct across Texas plain

This proposal seems to be simmering nicely, but very, very, slowly.

One thing that seems to be impacting on the Texas proposal is that Texas does not have a credible rail option at present, let alone one that simply needed ‘speeding up’ to make it competitive.

There is no direct train from Dallas to Houston, albeit there is a very infrequent one, the Texas Eagle of Amtrak, running direct from Dallas to Austin, which is near, but not near enough.

This contrasts with High Speed Rail in classic markets like Japan or Europe. Usually there has been an existing fast train, trying its damndest to cruise the route at some speed, but hampered by the current line. Japan has very fast, for narrow gauge, trains running the Tokyo-Osaka Tokaido route before the Shinkansen. In the UK, we think of how much effort has gone into speeding up the London to Edinburgh journey, now four hours at its very fastest.

Before JNR built the Tokaido Shinkansen in the 1960s (which NHK are there to remind us in the poster for the Wives of the Shinkansen) they ran express trains along the old alignment as fast as possible on the skinny gauge.

The argument in these countries, for the public and politicians alike, is to improve the rail service using High Speed Rail, rather than effectively create a new one.

That said, the Texas people seem to be arguing their hardest against the naysayers, and are just waiting for all the stars to align, financial and political, before embarking on an extremely conservative design along their route.

The Texas proposal is conservative, based on proven Shinkansen models of the Central JR.

With favourable topography, the Texas proposal is not the most interesting one. That would be California.

For a project initially approved way back in 2008, progress has been extremely slow. Some of this is hard to fathom for outsiders. First, the total population of California, and even of the affected route, is very high, at least 20 million. All the relevant freeways and airports experience congestion.

While federal intervention might be welcome (depending on the colour of that government) the project is wholly within the ambit of the state government, making inter-jurisdictional conflict less likely. California, despite all to frequent economic crises, is not a poor state, far from it.

And besides the high populations at both ends, there are plenty of cities en route to add to the patronage, as will be discussed below.

California Geology

In one respect, Californian geology and topography could not be more different.

California, unlike the Australian east coast, sits on the Pacific Ring of Fire where active plate tectonics are creating relatively young (in geologic terms) and severe land forms such high uplift mountains and volcanos.

One of the defining features of California is the Central Valley, a very long (hundreds of kilometres) nearly flat valley sandwiched between ranges of mountains that each represent a time of geological activity in the past. These ranges, unlike most Australian mountains, are very sharp and tall, not worn down by the hundreds of millions of years of erosion, but much more recent.

The high ranges have a second feature, which is to be high enough to attract rain and snow that can water what would otherwise be fairly arid land.

High mountains line the US west coast and inland western states, catching the rain and snow.

The Central Valley also represents a past when sea levels were higher, and the valley had both the Sacramento and San Joaquin basins as inland bays, continuing the pronounced feature that San Francisco has, where a long bay reaching right down to San Jose and Silicon Valley defines the area, and has San Francisco itself at the head of a very long peninsula.

With the sea gone, the Central Valley has been densely farmed and with farming has come not inconsiderable settlement, with sizeable cities with inhabitants numbering in the 100,000s (including the suburban areas) in Bakersfield, Fresno, Stockton, Modesto and Merced.

San Joaquin train in the Central Valley, this picture showing the flat topography between the mountain ranges.

But in another respect, California’s geography is all too similar to South Eastern Australia, specifically Sydney and Melbourne.

San Francisco and Los Angeles, to differing extents, have been placed in the most favourable places for their 19th century growth and development, in a way that helped in the 20th century, but is now hindering.

Both sit on narrow coastal topography, limited on all sides by sometimes abrupt mountain ranges.

The California HSR project

Early in the project Californians and their political masters had to choose two broad routes from San Francisco to Los Angeles. The first went via the Central Valley, a route that would ensure much of the journey would be relatively flat and easy to construct. Using the existing rail easements for much of the journey, such as the Union Pacific rail corridor, would make it even more so. The major expense would be reaching the plains at both ends, especially the LA end.

LA Union Station

The current rail experience

California used to have two passenger rail corridors between SF and LA. The Central Valley route, using the Tehachapi Pass between Los Angeles and Bakersfield; and the Coastal line which still runs.

Service in California on these routes is a mixture of ‘national’ Amtrak; Amtrak California (operating under contract to the California State Government) Los Angeles Metrolink long distance commuter rail and Caltrain in the San Francisco Bay area.

Amtrak long distance train

Since the Bakersfield to LA service by Amtrak no longer runs, the Amtrak California San Joaquin service runs several times a day from Bakersfield to both Oakland (for the San Francisco Bay area) and Sacramento.

The Tehachapi Pass route is currently closed to passenger rail due to congestion on the single-track freight route, and on current alignment, it would also be a slow journey. The line remains in use by Metrolink from Lancaster into LA.

LA also has metro and light metro services

The main blockage at the LA end

The second, via the coastal cities of Santa Barbara and San Luis Obispo, would be considerably more expensive through coastal ranges and not attract a lot of en route patronage besides those two cities.

The population is less dense along the coastal route.

At the San Francisco end, the peninsula is served by the Caltrain service to San Jose and Gilroy, which while it makes physical connection with the San Joaquin service is not by design trying to connect with it (it is also not necessary from San Francisco downtown, as a better option is to use the BART underground to Oakland and connect there).

BART is an effective access to downtown SF, if only it was very easy to connect onto it from the Caltrain system.

The coastal route is served by National Amtrak as the Coast Starlight from LA to Seattle via Oakland, serving Santa Barbara and San Luis Obispo en route, while these cities have an additional connection to LA a few times a day via the Amtrak California Pacific Surfliner (which runs from San Luis Obispo through LA to San Diego) and local Metrolink trains to East Ventura.

None of these trains are setting speed records, and probably don’t fit the ‘precursor’ idea of ‘fast-as-you-can’ rail that you see from London to Edinburgh. All are subsidised by the state of California.

Staging High Speed Rail

Sensing, one suspects, an opportunity to kick much of the can down the road, it seems to have dawned on the political and bureaucratic class in California to construct the Central Valley section first, at relatively lower cost, to build a working HSR there, and only afterwards worry about the much more expensive connections to the SF and LA areas.

While the full route has been drawn on maps, generally the only construction taking place is in the Central Valley, from Merced to Bakersfield.

The difficult political problems of alignment and expense are therefore negated upfront.

From Los Angeles, there are essentially two problems:

• The alignment through the suburban area of Los Angeles, where the coastal plain is reasonably flat but already extremely well developed. There would be limited scope to use the existing freight/Metrolink corridors to the foot of the ranges and journey times would be too slow.
• There are two ranges between LA and Bakersfield. Metrolink penetrates the first of these, via the San Fernando tunnel, to get to Lancaster, but the Techahapi one already mentioned will be extremely expensive. It currently takes Metrolink more than 2 hours from Lancaster to LA via the Newhall Pass– the HSR would have to do it in no more than 30 minutes to achieve the proposed SF transit times.

LA Metrolink stock

Tackling both these issues can only be achieved expensively including some long distance tunnelling.

At the San Francisco end, urban rail congestion and an indifferent corridor down the peninsula towards San Jose will limit the options for HSR, as well as crossing of the Diablo range to Merced. This mountain range is not as dramatic an obstacle as at the LA end – but still adds to the expense as well as the controversy.

If the Central Valley alone was an HSR?

According to Wikipedia, the total population of the Central Valley is 6.5 million. However, not all of these are along the proposed route, mainly due to the rail turning off the San Joaquin alignment at Merced and heading to SF rather than continuing along the Valley.

If the route was not from Bakersfield to Merced then San Francisco, but instead continued following the favourable topography to Stockton and Sacramento, a population around this level, or maybe higher, would hardly be a compelling case for HSR, but it would be a start.

Construction for the HSR in Fresno. Note how flat and spread out the city is

These cities follow the unfortunate American pattern of being spread-out, lacking a dense ‘downtown’ and otherwise poor environments for local public transport. Only Sacramento has local light rail system.

If HSR was to thrive in the valley, it would need to bring with it all the trappings of end-to-end transit. While SF and LA are hardly world exemplars for this, they are leaps and bounds ahead of the cities of the Central Valley.

In terms of national politics and culture, irrespective of their populations, the Central Valley cities do not loom anywhere near as large as LA or SF do. It would therefore be difficult for such an HSR to imprint itself on US citizens and voters the way a line from LA to SF will.

Governor Newsome, in his recent speech, attacked those who suggested the idea of a train only from Merced to Bakersfield would be a ‘train to nowhere’ as offensive to the residents of the Valley, who he noted have the worst air pollution in California.

Politics

It would be foolish to think the opposition to HSR was due to money alone, though a sizeable part of it does relate to that. So specifically the challenge of the mountain passes at both ends has added to that.

The game in California is the proponents versus the ideologues of the right, fighting the very existence of the thing but with 10,000 small objections against route, against land acquisition, against financing, against whatever they can find. This was inevitable even with an easily sellable, easily built proposal like Texas.

But the proponents knew this, and should not have been surprised at the response. Managing their own affairs – working out their own strategy, was the key to their success.

The total, true cost of such a line, inflated by American overcharging, was obviously beyond the California authorities to put to the people in a referendum. However, and unlike Australia, the people, having voted by referendum for a portion of the cost, have given the authorisation for that portion to be spent, and the government in essence cheated by building the Central Valley section first.

This is unlike other staging we have seen.

For example, the London to Paris Eurostar route was built as TGV across France, a medium speed channel tunnel, followed by very slow (130km/h) operation across legacy lines through Kent to London. While this provided a sub-optimal route at commencement, it was good enough to get the line going, while the HSR through Kent and into London (by tunnel) was built. That line then consolidated the case and took market share from aviation on the route.

In France, the line from Paris to Bordeaux started first as a TGV from Paris to Le Mans and Tours, tackling the most congested section first. The line beyond to Bordeaux had always been fast and the overall journey time was enough. A full line was later built that reduced times further.

Japan, where the legacy gauge was different, did not really have such an option, though in northern Tohoku, changing from Shinkansen to ordinary train between the boundary at Morioka and Aomori provided still a much faster journey than using the ordinary train the whole journey and was therefore more popular than flying.

In Japan the only reduced-cost option for staging Shinkansen works was to dual gauge some sections, such as the Akita line shown here.

This blog does not think attempting to outwit the Californian public has worked. The new line is really not much more than an enhanced San Joaquin service. Though therein lies what might have been a better option.

A new line from LA to Bakersfield, expensive though it would have been, could have allowed through service via the Valley to be restored. And this service could have been upgraded, using a faster Talgo than the Starlight, to enable, by way of example, say a 2-3 hour service to the Valley cities from LA. Such a train, diesel powered, could still have continued into SF or Sacramento.

Caltrain has ordered new trains for San Joaquin, Pacific Surfliner and other trains, but these will not be medium speed, as should have been contemplated.

If it had maintained 200km/h along the flat, duplicated and grade separated Valley line, while continuing to hold freight, and initially a 200km/h transit of the tunnel into LA, it would have proven a medium speed solution to the Californian public. It would not be fast enough to replace LA-SF aviation, but it would replace some of the intrastate Valley aviation.

The tunnelled route from LA to Bakersfield would of course be engineered for 350km/h running and with provision for electrification. It would also improve the service to the Antelope Valley at Palmdale and Lancaster for local commuters.

Techahapi Loop, a major blockage on the southern route into the Central Valley from LA. Dicklyon@wikipedia

And therein it would not be trivial -but would have real commuters from the Mojave Desert but also the lower Central Valley into LA each day helping defray the cost.

At the San Francisco end there are some complicated politics still to resolve. The first element is to provide, as Caltrain intends, a much better service than at present. The downtown station for San Francisco is very poor and connectivity into BART is also poor. A much faster service via the peninsula to San Jose is required in any event.

Caltrain, seen here at Millbrae, is to be electrified.

Such a service, if it also reached Gilroy and Merced faster, could be a game-changer for the upper Valley. These items could and should also have been delivered without the HSR moniker but would help HSR when/if eventually delivered.

There is also still a lack of overall plan for Californian rail. The BART has been extended, at considerable expense but also with poor frequencies and transit times, towards San Jose. Santa Clara has a light rail system, also covering much of the same turf as BART, Caltrain, Amtrak California and the HSR within San Jose. Yet it is hard to see how it all joins up, in the way you would expect in a city of that size in Europe or Japan.

An outsider might have thought the BART extension less important than getting a good medium speed solution from San Jose into downtown SF as well as to Oakland and on to Sacramento.

Elsewhere in California

There is proposal, still not under construction but probably more advanced than the Texas one, to have high speed rail to Las Vegas from Southern California. Note – not necessarily, at least at the outset, from Los Angeles.

Topography, again is a killer. The same topography, with the coastal ranges between LA and Victorville, creating the need for tunnelling. Beyond Victorville and towards LV, the topography is much easier.

The market is definitely there: Las Vegas metropolitan area has about 2 million people; it is a major US tourist destination and the Interstate 15 freeway to LA is always congested (and not a pleasant drive). Flights are popular, but a fast train could get there quicker, depending on the way the service is designed and run.

This blog took SW Airlines from LA Burbank to Las Vegas.

So the proponents want to start the service from Victorville, on the other side of the range. In so doing, they have avoided much of the drama of the government proposal, albeit their operation relies on the same topographic ‘cheat’. Perhaps the proponents believe, that if the Victorville to LV section works, this will clear the opposition to building it into LA.

Victorville is only 135km by road from LA. The train (the long distance Southwest Chief) takes 2.5 hours, the bus even longer. Even by road in peak time it will be a long drive.

The questions that remain to be answered:

• Whether people will drive part way, and catch HSR the balance. The aim is to get people out of their cars completely, even if it meant some sort of ‘last mile’ solution for the USA like limo or van pickup to get people to the station.
• The government proposal is for a tunnel through the range to Palmdale, while the LV line would require one to Victorville. A compromise – a tunnel to a point between Palmdale and Victorville, might be able to be shared by both. Of course, as the government and the private (Virgin Group) people are separate, and no coordination is formally proposed.
• Could a HSR (albeit diesel powered) get from LA to Victorville along the current route in 2 hours, and add this to the current proposal to achieve a LA-LV transit time around 3.5 hours. Not great, more like the first London to Paris Eurostars…but better than nothing.

Photo LAtimes

Lessons for Australia

The USA, despite its high population, often pleads inability to do HSR, for reasons also often cited for Australia – namely low population densities, cultural preferences for and cost of driving, and topography.

This blog is certain these issues are overstated for the USA, and also for Australia.

Of course the USA, with its large population on a large continent, has a low overall population density, but this has always been a spurious argument, as the rail is not intended to serve the whole country, but specific corridors.

These corridors often do have quite creditable populations and densities.

Much of this is in the definition.

Many European or Asian urban areas refuse to be described as ‘suburban’ as they have their own legal status and long histories of independence. This tends to inflate the notion that the travel from some of these places is not ‘suburban’ or ‘commuter’. City pairs such as Paris-Tours, Nagoya-Osaka or London-Brighton might seem to fit the classic definition of intercity routes, and in many respects they are. But at the end of the day, transit time is transit time, and population is population.

Paris-Tours TGV

The US cultural attachment to driving is well known (as if Europeans or Asians don’t drive – a ridiculous proposition) but how much of this is because of circumstances after WWII.

As in Europe and Asia, the railroads after WWII needed massive upgrades due to demand for travel, lack of investment for 20 years, and changes in technology. But unlike in Europe or Asia, the government generally refused to make this investment in the privately owned lines, and many laboured into the 1960s and 70s with 1920s equipment.

What investment there was sensibly gave freight traffic priority, so much so that the operation, with high tonnages but low speeds, became unsuitable for all but what this blog refers to as liner trains – going slow enough to enjoy, not fast enough to use.

Copyright as shown. The Pennsylvania Railroad entered the 1970s in very poor condition, despite the then new Metroliner electric shown at left.

The USA also failed to make the investment on transport within cities – other than freeways – so intercity rail travellers would not find the local transport they needed at each end.

We have seen in this blog post the impact of topography.

In Australia, the past posts noted the South East of Australia, is blighted by topography, in fact some of the worst that rail can encounter. Rather than the U-shaped (flat-bottomed) glacial valleys of Switzerland, we have V-shaped river valleys that have eroded in jagged shapes that make using the valleys for rail alignments poor.

Swiss trains can run along the bottom of glacial valleys.

Between Melbourne and Sydney, the V-shaped valleys in the Great Dividing Range typically run south-east to north-west, meaning the straightest rail route would have to confront each new range or spur ‘head on’, rather than running along the valley bottoms.

This blog’s strategy was to have the line follow the Melbourne to Albury line on the current alignment, which avoids just about all these perpendicular mountain ranges. But from Albury to Canberra, tackle the issue head on with a series of extremely long tunnels.

Between Albury and the tunnel entrances – the line can use the flattish valley of the Upper Murray, defined by the shore of Lake Hume. From Canberra to Goulburn, there is conveniently a flood plane of Lake George that provides a flattish route all the way to the Breadlebane plain.

In other words – some aspects of topography cannot be beaten except by tunnelling, so factor that into the project.

Population as such is not the issue. People who live alongside the line, except where near a station, are not relevant to the forecast demand. In fact for new lines, the fewer people en route the better.

The problem of ‘density’ is really the same ‘last mile’ problem as in the USA. If the Sydney-Melbourne route had a station and diverge near Campbelltown, and with most trains going to Sydney CBD but some to Parramatta (say at 160km/h along the current alignment) this would take away the usual argument that people in Sydney’s west would not use HSR. In Melbourne you could do the same thing at Somerton, with some traffic direct to the CBD, but others to Sunshine and Geelong not through Melbourne City.

The Californian Government ‘cheat’ – building an affordable line that is not that useful, hoping to demonstrate its value – is something we should avoid. If we build something less than HSR in Australia, it still needs to be useful.

For example, Sydney to Canberra rather than Sydney to Melbourne. Sydney to Newcastle would be an immediate success, even if its route would be the most difficult on the whole line to Brisbane.

Tim Stewart wiki photo

And unlike the USA, the HSR can be made to connect with reasonable local transport, for example, a station at Ourimbah would connect with local Central Coast trains.