3 ms·
you're not exactly making a good faith argument here [0] but sure i'll play ball i'll ignore the infrastructure costs, it's free to drive on I-5 after all. an
by andbberger 3y ago
you're not exactly making a good faith argument here [0] but sure i'll play ball
i'll ignore the infrastructure costs, it's free to drive on I-5 after all. an N700S costs about $50M and holds 1300 people. amortized over 30 years, you're looking at say $3M/year including a very generous maintenance budget. i'll throw in another $0.5M/year for misc personnel costs (driver, station staff).
even if you ran only a single weekday peak hour round trip per trainset with tickets priced at $20 one way, revenue would be about $13M/year. so you could charge $10 for tickets, follow american "best practices" of running only peak service and idling your expensive rolling stock for the rest of the day and still make ends meet.
in reality, the infra would be built following best practices to bring the runtime between portland and seattle to just under an hour[1]. this facilitates an hourly takt. running short turns between portland and seattle each trainset would make 5-8 round trips a day, which even with low occupancy factors is still tremendously profitable.
[0] see the other comment where you say that "transit dorks" never talk about how much things cost. "transit dorks" are _obsessed_ with how much things cost. alon levy, who is by all accounts one of the more influential voices in internet transit world these days has a multi year project in collaboration with several other transit activists to find out exactly why transit projects are expensive as they are in the US. not to mention dozens of blog posts analyzing individual projects. Clem Tillier has a site going back 15 years where he publishes very detailed analysis on the bay area portion of the california high speed rail project, its cost overruns and policy driven technical failings. that "transit dorks" don't care about cost is a conclusion you could draw if you only follow reductive normies like alan fisher, but if you uncritically accepted that such sophomoric reddit tier dialogue was the state of the art in transit twitter - well, one would have to really question how much critical thought you've put into your stance on car dependency and EVs.
[1] a route following I-5 is 260km so this is a reasonable assumption.
- forrestthewoods 3y ago> i'll ignore the infrastructure costs, it's free to drive on I-5 after all. Ugh. Like I said. =\ We aren’t building a new city with no initial infrastructure. This isn’t Sim City. And we aren’t rebuilding after getting firebombed and flattened in WW2. What we do have is an actually pretty good interstate highway system that was built in the 50s by eradicating minority neighborhoods and rampant use of eminent domain. The only relevant question “given where we are today, what makes sense”. How much does it cost to go from where we are to a rich and robust train system? So, we have a bunch of highways that require maintenance/growth and a bunch of train track that needs to be built. If you want to move budget from highway maintenance to train track construction and maintenance that’s totally cool. But you can’t just handwave it away. I’ve read all the articles about why America sucks at building rail and what the per mile cost is. So. Napkin math. How much to build, operate, and maintain a high-speed rail system in the PNW. Full system costs. Face value. With zero complaining or snark about how much we spend on cars/highways today.
- andbberger 3y agoabout $8B for portland to seattle. in comparison, a single bridge that carries I-5 over the columbia river is being replaced for $5-7.5B. you do the math.