Why Can’t Amtrak Just Go Faster?
Amtrak, particularly outside the northeast corridor, has a reputation for being slow and unreliable. I’ve previously written about the main cause of unreliability — freight trains are involved in over 40% of Amtrak delays — so let’s talk about Amtrak’s speed. It’s no secret that Amtrak is often significantly slower than driving: Rochester to Toledo is 6.75 hours on the Lake Shore Limited vs 5.75 on the I-90 and Jacksonville to Columbia is 5 hours on the Silver Star vs 4.25 hours via car. Kansas City to Chicago is takes 7.5 hours on both the Southwest Chief and by car, but only 3.5 hours via plane once travel to and waiting at the airport is accounted for. Even though train travel is substantially better for the environment than both driving and flying — and more comfortable — it can be hard to justify the increased journey time, particularly for short weekend trips.
Getting Amtrak to go faster might sound absurdly difficult: California’s High-Speed Rail program is notoriously expensive (albeit overblown), and replacing Amtrak’s rolling stock would be wasteful and costly for a company subsidized by our tax dollars. Upgrading Amtrak’s long-haul routes to the 200+ mph standards of a high-speed network is just not practical currently.
That notion is a bit of a red herring, though! The question was whether we can make Amtrak go faster, not whether we can make it high-speed rail. As a mathematician, I can assure you there are numbers between 80 and 200. Can trains go those speeds? Here, the answer is yes, and it’s surprisingly simple. Let’s take a look at what’s going on with our trains.
I live in Albuquerque, New Mexico, which is a stop on the Southwest Chief. The Southwest Chief is unique among Amtrak’s long-haul routes in that the trains can go 90 mph for about 300 miles of the 2,265-mile route. Why the Southwest Chief can do this and not other trains isn’t because of topography, frequency of stops, or even the capability of the train, but rather simply the speed limit of the track. Unlike road speed limits, track speed limits fall into only a few discrete buckets. Like road speed limits, finding data about what parts have what speed limit is annoyingly difficult.
We can approximate knowledge of the speed limits by simply tracking how fast the Southwest Chief travels on its journey. Thanks to TransitDocs, I can pull historical train information, so let’s look at the 14 Southwest Chiefs that departed between August 7 and 13, 2022. Since this isn’t an article about delays — again, I already wrote that one — we’ll throw out the 5 trips that were delayed by more than 2 hours.
As you might hope, the Southwest Chief spends nearly two-thirds of the route (1400 miles) traveling in the Class 4 and 5 speed tiers that are the standard for Amtrak long-haul routes. What is concerning, however, is the amount of track where the train is going slower, in particular the 600 miles the train spends in the Class 3 band. Worse, this track is Class 3 not because of sharp turns or slopes; we know this because this route is exactly the route run by the Super Chief over 50 years ago, which traversed it 20% faster than its modern descendant.
Instead, a large portion of the track is Class 3 because BNSF — the freight company that owns most of the track — pays the bulk of the maintenance, and freight companies today don’t consider higher class track classes to be worth the increased maintenance cost. You might wonder, how much does it cost to maintain these tracks?
That would be a difficult question to answer, but again the Southwest Chief makes a great example. In 2016, the track contract between BNSF and Amtrak needed to be renewed, and discussions over dividing maintenance and repair costs over the previous years resulted in a 2014 study on those costs. It determined that keeping the central third of the route at Class 4 — from Lamy to Newton — would cost about $53,000 per mile in maintenance (2022 dollars). For comparison, states spend an average of $67,000 per mile to maintain four lanes of state highway (and states notoriously don’t spend enough on highway maintenance).
The next question is what effect does this Class 3 track have on trip duration? One way to answer this is to create a similar pie chart:
Here we see the effectiveness of traveling fast: while 66% of the track is Class 4 or 5, only 42% of the trip time is spent on that track. Instead, a substantial amount of time is spent either stopped or in the Class 1 and 2 bands — without in-depth study, this is likely unavoidable transit through the cities the stops are located in. Still, 27% of the trip time is spent in Class 3. What would happen if, during each stretch we were going between 32 and 75 mph, we instead went 75 mph?
It turns out that this simple change reduces the overall trip speed by about 10%, and that 7.5-hour trip from Kansas City to Chicago drops to 6.6 hours. But, while we’re making simple changes, why stop there?
Amtrak’s long-haul trains are far over-qualified for the tracks they run on. The cars are generally either the 30-year-old Superliners, which have a top speed of 100 mph; Viewliners, with a top speed of 110–125 mph; or Amfleets, which can go 125 mph. For locomotives, Amtrak is currently replacing its GE Genesis-es (top speed 110 mph) with Siemens Chargers that can go 125mph.
Allowing Amtrak’s trains to go at their designed speeds would be transformative to passenger rail in this country. Continuing the analysis from before, if the Southwest Chief’s average speed when traveling faster than 32 mph was 100 mph (a Class 6 speed with a 10 mph buffer), the trip would be 25% shorter. That’s a pretty big improvement given that we need only minor re-allocations of Amtrak’s rolling stock and no removal of stops!
That 25% improvement means the Southwest Chief travels from Kansas City to Chicago in 5.5 hours, the Silver Star between Jacksonville and Columbia is 4.25 hours, and the Lake Shore Limited between Rochester and Toledo takes 5.3 hours — all comparable to or faster than driving, and with the added ability to walk around, eat, and relax on the trip.
The infrastructure needed for this “higher-speed rail” is much less expensive than the eye-popping numbers for high-speed rail. Requirements for level crossings — where railroads cross roads — are the same for all tracks up through Class 6. The long-haul routes by their nature are mostly straight and in rural and unpopulated areas, reducing costs compared to the densely-populated Northeast Corridor. Finally, the United States has a long history of working with and maintaining Class 6 track: most of the Northeast Corridor and part of Michigan accommodate speeds at least that high.
In fact, the LOSSAN Rail Corridor in Southern California is doing exactly this: at a cost of less than $17 million per mile, they are performing various track upgrades that both increase capacity of their rail and increase its speed limit to 110 mph. For comparison, in 2021 Arizona announced a project to widen a rural section of its I-17 at the cost of over $19 million per mile.
Of course, all of this is pointless unless we can rein in the scourge of freight train interference. It doesn’t matter how fast the track lets you go if you’re stuck behind a 40 mph freight train, although this would at least help Amtrak make up lost time. There is legislation to fix that: if you live in the US, call your congresspeople and local government officials and ask them to support it!
As we ramp up efforts to fight climate change, we will increasingly need to reduce emissions from transportation. Passenger trains will play a large role in this effort, being by far the most efficient way to transport people long distances (yes, even diesel trains can be more efficient than an electric car). With governments at all levels looking to both reduce our carbon emissions and invest in transportation infrastructure, we shouldn’t pass up the opportunity to make the long-haul routes as fast as, if not faster than cars.
This relatively simple process would be a massive step toward building robust passenger rail service and establishing the habits and demand for future high-speed infrastructure. It’s not the only step needed — the number of stops and frequency of service are important too — but it has a broad, positive impact across the entire train system. Let’s do it.