Desegregate Amtrak’s rolling stock acquisition

Back in 1954, Brown vs Board of Education established that “separate but equal” was not a permissible standard for providing services. The same concept applies to a proper fleet modernization strategy. Right now, Amtrak’s plans involve a completely separate set of acquisitions for the current Northeast Regional service and the Acela replacement and NextGen HSR service they plan on. This leads to needless expenses when they could, in fact, receive greater benefits at lesser cost by merging the acquisitions.

Currently, Amtrak plans on replacing about 484 Amfleet I cars over a 5-7 year period. On the high end, the new bilevel car buy for California and the Midwest is proposing $551 million for 130 cars, or about $4.24 million per car (page 78). On the low end, the regular price for one of Metrolink’s Guardian cars is $2.35 million with contract options bringing it down to $1.8 million and a total contract value of about $2 million per car. So simply replacing the Amfleets will cost about 968-2,105 million dollars. It's actually rather worse than that as Amtrak plans on purchasing 825 single level cars over the next 11 years (including the 130 Viewliner II order currently under contract), but many of these are for long distance travel. If all except for the Viewliners were counted for NEC and connected service, this would bring the total to $1.39-2.95 billion.

In addition to this, we have Amtrak’s NextGen HSR plans for the NEC which entail acquiring 12 new high speed train sets in 2020 and 32 in 2025. Since the separated right of way that Amtrak envisions will not have been complete yet, we may very well assume that these are going to be subject to all the fun and joy of FRA regulations and overweight under current Amtrak plans. That aside, Amtrak plans on spending $5.2 billion on 58 train sets out till 2040, making for an average expense of 90 million per train set for an expense of $1.08 billion for 2020 deliveries and $2.88 billion for the 2025 sets.

So let’s scrap this nonsense. We have a budget of $4.93-6.06 billion dollars to play with if we simply stop segregating the NEC fleet and new equipment acquisition. What could we get if we injected some sanity into the process?

We could, to start with, buy some New Pendolinos. These come in tilting and non-tilting versions and are capable of speeds up to the current track and regulatory maximum of 150mph (160mph if the FRA ever comes to its senses and synchronizes the signal and track speed limitations). The British, back in 1997, purchased 53 train sets, custom made for the British loading gauge, at a cost of £1.2 billion (extending the fleet to 9 cars may have cost additional however; a later contract would have been about £1 million per car). Adjusted for inflation, that’s about $2.7-3.6 billion USD, depending on your choice of measurement (simple CPI is $2.8 billion) or $52.8 million per set with CPI inflation. With the current fleet acquisition plans, that would represent 93-114 train sets that could be purchased instead. Obviously that’s a fairly ludicrous number greatly in excess of what’s needed: Amtrak’s fleet strategy plan only anticipates a daily availability of 16 Acela sets and 45 electric locomotives (which includes hauling long distance trains and any trains with multiple locomotives). If we assume a necessity for 61 train sets to replace current Acela, Northeast Regional, and Keystone Corridor service, we can recapitalize the Northeast Corridor’s fleet with custom build Pendolinos for approximately $3.22-4 billion (with 100-80% availability rates; Alstom does claim it has a 100% rate of availability maintaining Spain’s AVE fleet [page 18]).

Of course, more standardized equipment is cheaper. Czech Railways purchased seven 7-car Pendolino sets for approximately two hundred million dollars (~$28.5 million each), though these were not without their share of problems. More recently, Poland has purchased non-tilting ETR 610 New Pendolinos for 47.61 million each, a price which includes 17 years of maintenance by Alstom. Alstom also sold twenty-five 10-car AGVs, a non-tilting train capable of 186mph, for an inflation adjusted $994 million, about 40 million per train.

As a result of desegregating the Northeast trains and pursuing a rational fleet strategy, Amtrak could save billions of dollars while simultaneously greatly increasing the service speeds of what are currently the Northeast Regional trains and, depending on the performance characteristics of what it buys, significantly improve upon its current Acela running times (this is almost a given with any significantly lighter rolling stock: a Talgo official calculated that a Talgo train set, limited to only 125mph, would be only four minutes slower than the Acela on a Boston-Washington run [page 7]). In turn, these savings can be used for track and catenary upgrades which minimize slow zones and otherwise speed up the train service allowing Amtrak to meet many of their NEC goals for significantly lower expense.