01 March 2014

High Speed Freight Rail?

The biggest factor in determining the economic viability of high speed passenger rail systems is the population density of the corridors that they serve.  The cost of building high speed rail over non-mountainous and non-swampy terrain per mile of tracks pretty much constant (adverse terrains are about 50% more expensive per mile of track).  The amount of expected trips on high speed passenger rail varies greatly between areas with relatively low population density and areas with relatively high population density, all other things being equal.

For high speed passenger rail systems to be viable, the infrastructure cost per trip over the useful life of the tracks has to be sufficiently low.  If you double the number of trips, you cut the infrastructure cost per trip in half.

Suppose that you take a high speed passenger rail car, leave its exterior dimensions and connections to the track essentially unchanged, gut the interior, use that car for cargo rather than to carry passengers, and move that cargo via high speed rail on the same basis as a passenger car would move.

Adding the freight traffic increases the amount of paid passenger trip equivalents in system revenue without changing the track infrastructure costs at all, to the extent that the freight charges are more than the variable operating costs associated with adding the extra car to the day's traffic and the infrastructure costs of the cargo car per freight trip over the useful life of the cargo car, and each freight trip's share of any infrastructure necessary to facilitate loading and unloading cargo as opposed to passengers.

Certainly, building new rail lines, which is necessary for high speed rail in most instances because turns must be softer, grades must be more flat, train gauges may be different, and electric supply systems for electrically powered high speed rail systems must be supplied.  But, it isn't at all clear that a mile of new high speed rail line is all that much more expensive than a new mile of interstate highway or a new mile of a low speed rail freight rail line.  The land acquisition and materials costs and construction effort aren't all that much different.

It also isn't at all obvious that the operating costs per ton-mile of freight moved using low speed rail is all that much cheaper than the operating costs per ton-mile of freight moved using high speed rail.  Indeed, there are economies of scale even in variable high speed rail operating costs.  A lot of the cost of operating a train comes from the salaries paid to the personnel the run it, and the fuel and maintenance costs aren't all that huge by comparison.  But, the personnel costs of running a high speed rail train with two cars and one with four or five or ten cars, is pretty much the same.

Suppose that it is possible to move freight via a high speed rail system at a cost only marginally more than moving freight via a low speed rail system, or for that matter at a cost comparable to (or even a bit less than) the much more expensive option (relative to low speed freight rail) of moving freight via a semi-tractor trailer on an interstate highway with speeds that a little bit higher that sending freight on a truck on an interstate highway.  If this is true, then high speed freight rail would have the potential to win over a very large market share of the intercity truck freight market, vastly increasing high speed rail traffic overall and vastly reducing the infrastructure cost of high speed rail relative to a passenger rail only model.

Let's be a bit more concrete.  Suppose that a truck on an interstate highway costs 37 cents per ton-mile (this source puts the figure closer to 17 cents per ton-mile for trucks, but confirms a 3 cent per ton-mile figure for low speed freight rail) and moves freight at about 55 mph on average from point to point including time spent stopped (with peak speeds of 75 mph on rural interstate highways).  Now, suppose that moving the same freight on high speed rail costs 33 cents per ton-mile (eleven times the cost of slow speed freight rail) and moves freight at about 80 mph on average from point to point including time spent stopped (with peak speeds of 160 mph), and time spent transferring the freight from rail to trucks for for first and last 15-20 miles of the trip.  Why wouldn't high speed freight rail capture almost all of the truck freight traffic on competing routes and vastly increase the volume of high speed rail traffic per mile of high speed rail tracks.

I think that these speeds and prices are attainable with existing high speed rail technology as part of an overall mixed passenger-high speed freight system.

Now, to do this, one needs to be able to very quickly move a high speed rail system compatible shipping container from train tracks to a waiting truck for the freight to go the last few miles from the destination high speed rail station to its final destination up to about 15 to 20 miles away.  Otherwise, the delay in getting freight on and off the train will slow down the passenger service so much that the benefits of high speed rail will be lost.  But, working out a solution such as having a big crane that lifts cargo cars off the tracks and puts them on trucks in a method similar in principle to existing multi-modal shipping container systems is very viable technologically.

The prospect of shifting a huge share of all truck freight to rail at prices closer to truck freight prices than to current rail freight prices would improve the economics of high speed rail and would also dramatically reduce interstate highway road maintenance costs, traffic delays from serious freight truck accidents, and other harms from freight truck accidents.

This is not just an accounting gimmick.  Fundamentally, it should work for freight because the labor costs of having one driver per truck are much higher than the labor costs of adding a few freight cars to a passenger rail train, or of running a purely freight trip with perhaps a dozen freight cars, on rail tracks that aren't vastly more expensive to build and maintain per mile than the alternative of interstate highways.  Rail is also just inherently less prone to collisions or other accidents than freight trucks.  Fuel costs for rail per ton-mile are generally lower than for trucks, but even if they are slightly higher, the total operating costs for high speed rail per ton-mile should be lower.

High speed freight rail is essentially equivalent to having self-driving cars without the huge technology risk associated with that kind of technological leap.

A 50-50 revenue mix of high speed passenger traffic and high speed freight traffic would cut the population density necessary for high speed rail to be viable roughly in half.  This could very well be the difference between viability at all, and non-viability, of high speed rail in places like the I-25 corridor in Colorado.

It might very well be that it is just impossible to get the cost of high speed freight rail down to the 3 cents per ton mile of low speed freight rail.  But, I wouldn't be surprised at all if it proved to be economically viable to provide high speed freight rail services at costs that are closer to 3 cents a ton-mile than they are to 37 cents per ton-mile for trucking.  And, if it is, high speed freight rail might not only virtually wipe out trucking in reasonably dense traffic corridors, but also might pick up a lot of the higher value freight that is currently transported by low speed freight rail.  Also, the speed of high speed rail relative to air traffic for medium distant trips (where non-flying time cuts into the benefits of high flying speeds) might make it highly competitive with air cargo at 122 cents per ton-mile.

Mixed passenger-freight high speed rail is not unprecedented, but typically, for example, in France, has been used only for postal freight.  But, the possibility of much wider high speed freight traffic seems at first glance to be something that can be made economically viable with existing technology.

No comments: