Dick Spotswood is a supporter of SMART and an optimist regarding its success, but his insistence that it could function with the same form as Caltrain shows a lack of understanding of how either system must work. Back in July, Spotswood argued:

When in Oceanside, [former general manager Lillian] Hames' crew should have walked across the depot to ride Coaster, the excellent passenger rail line linking the San Diego County coast. There, they'd find an off-the-shelf commuter railroad using high-capacity cars that are America's standard.

They would work perfectly in the North Bay hauled by environment-friendly Tier Four locomotives… It's all proven technology. Think Caltrain on San Francisco's Peninsula.

This, he says, would result in $120 million in savings and provide twice the capacity over the Sharryo DMUs SMART ended up buying. The savings would come from:

Using non-customized trains
Cutting specialized track work
Cutting specialized signaling systems
Cutting high loading platforms

This is simply bonkers. The Federal Railroad Administration (FRA) has rules of crashworthiness that come into force when freight trains run with passenger trains, rules that Caltrain doesn’t meet. The Sharryo trains cost $6.3 million apiece. A comparable FRA-compliant and Caltrain-style train costs $11.7 million*, almost double the original cost. Not only that, but the locomotive makes the train too long to fit within a normal city block, meaning streets would be blocked while the train is at a station. ~~Caltrain is elevated~~ Caltrain's stations are grade-separated and so does not have that issue.

FRA rules regarding freight/passenger interaction also dictate the specialized track work and signaling systems, which total only $36 million. ADA and FRA regulations conspire to require level boarding at stations, but the stations cost less than $3 million apiece. Even cutting them all entirely would only save $27 million. I cannot understand how one finds cost savings of $120 million by purchasing more expensive trains and cutting legally required capital expenditures.

The point is that train type doesn’t dictate much with regards to SMART’s capital costs. Specialization does come with a price, and there may be one to pay in maintenance later, but SMART’s cost overruns are not the result of purchasing DMUs and so cannot be fixed by replicating the Caltrain model in the North Bay. Indeed, Caltrain’s model is unsuited to the SMART corridor because those corridors are different. Caltrain cannot run with freight, its trains are too long to run at street level, and it is more expensive than the custom-built DMUs SMART already has. Making them fit SMART's constraints would only cost more.

If someone wants to build a boondoggle, running Caltrain on SMART’s tracks would be a good place to start.

UPDATE: Multiple commentators have pointed out that Caltrain is already FRA-compliant, and that the waiver is for future, rather than current, service. This was an oversight on my part, but the point still stands: traditional trainsets are too long and too expensive for SMART.

-----------------------------

* Assuming 1 Amtrak locomotive at $6.7 million and 2 Bombardier Bi-Level cars, to accommodate the 312 seats Spotswood argues are necessary, at $2.5 million apiece.