21stcenturyurbansolutions.wordpress.com/2009/07/27/bay-area-transit-efficiency-how-bart-caltrain-vta-light-rail-and-muni-metro-stack-up -> 21stcenturyurbansolutions.wordpress.com/2009/07/27/bay-area-transit-efficiency-how-bart-caltrain-vta-light-rail-and-muni-metro-stack-up/
Overall, Bart and VTA have been the overwhelming winners with numerous new extensions, Muni Metro has received some improvements and one extension, and little has changed for Caltrain. Yet, Caltrain will receive the most dramatic makeover in the next decade with grade separations and electrification from HSR. How do these systems stack up against one another and their peers nationwide? Ridership per mile allows for easy comparisons between the ridership of like transit systems regardless of the size of a system-rather than just showing total ridership, it shows the ridership density on a given line or system. Rapid transit (often subway) systems have the highest ridership per mile, since surrounding population density leads to a highly concentrated ridership base and justifies the higher capital costs. Compact streetcar systems have the next highest ridership per mile, followed by more spread out light rail systems, and finally commuter rail systems. Bart's low ridership per mile also means that Bart extensions are generally not nearly as cost-effective as many of its peer systems. Commuter Rail Systems System Ridership Route Miles Ridership per Mile 1 Metro-North Railroad (NYC) 265,000 384 690 2 Metra (Chicago) 316,000 495 638 3 Long Island Railroad (NY) 367,500 700 511 4 Caltrain 39,100 77 508 5 SEPTA Regional Rail (Philadelphia) 118,600 291 408 6 MBTA Commuter Rail (Boston) 149,900 368 407 7 Trinity Railway (Dallas-Fort Worth) 10,000 34 294 8 New Jersey Rail 276,000 951 290 So, while Caltrain's ridership is noticeably lower than other commuter rail systems, it is still the (just barely) 4th most efficient system in the nation. However, if you take away the 6 trains to Gilroy (which add 25 miles but only about 500 riders), Caltrain's efficiency becomes 752 passengers per mile, making it the most efficient system in the nation. Caltrain's extremely high efficiency speaks to the fact that Caltrain operates more like a single light rail line than a sprawling commuter rail system (even though Caltrain has never had the capital investment that a light rail line would have). Muni Metro's efficiency is extremely high, and could be higher if not for the routes which originally had light rail but were changed to buses back in the 1950s-the Geary and Mission corridors have 100,000+ daily riders (including adjacent lines), but are stuck with buses, while all Muni Metro lines (except the N-Judah) have less than 30,000 daily riders. These lines were saved in the early postwar era because of their tunnel infrastructure (to travel to more suburban areas) while lines along Geary, Mission, and Columbus were scrapped. But, the biggest variable in this analysis is cost-Bart costs more than more than Muni Metro, which costs more than VTA Light Rail, which costs more than Caltrain. Now, here's where the calculations get a little shaky, and highly theoretical. Let's establish an arbitrary base cost of $100 million per mile, and adjust ridership for each of the four systems accordingly. hardly none for VTA and Muni), and different years of completion (all). While Bart to San Jose's ridership numbers are questionable, either way both the system impact and the impact of the individual line wouldn't affect Bart's numbers very much. Systems with greater frequency hands down get more riders. Is this post just a bunch of wild estimates and speculations? Yes, but I think this is about as good of a comparison as is possible under a multitude of constraints. Well, in spite of the murky details, it appears that Muni Metro blows other Bay Area transit agencies away when it comes to cost-effectiveness, and Caltrain will probably be on par with Muni Metro's efficiency within the next decade. VTA Light Rail and Bart are the least efficient systems in the Bay Area for their price (only 1/4th as efficient as Muni Metr0) and therefore are using inappropriate transit modes for the ridership needs. Nationwide, Muni Metro is extraordinary, and Caltrain performs extremely well for a commuter rail system. Investments in the next ten years will put Caltrain on par with many light rail systems at a fraction of the cost. Bart performs poorly nationwide compared with its peers, and spends a lot more money to get its riders than other rapid transit systems as well as other rail systems in the Bay Area.
Reply I think the Teco line estimates aren't quite right. If you look at all the other quarters for ridership on the APTA site, it's actually about 1,000. It would be nice if it carried 17k, but that simply isn't the case and the APTA numbers should be fixed. I also might argue that more urban extensions of lines, whether it's BART or Muni would be more efficient than the lines that are currently planned. Van Ness, Geary, Broadway, Telegraph and the like would make those numbers change. Currently they are all slated for BRT, mainly because LRT lines such as T Third are over constructed and basically rebuilt the entire street, along with a couple of expensive drawbridge retrofits. It really skews the numbers of how much LRT should really cost. But this analysis also really shows Caltrain's strength and shows how electrification would be a strong investment. Eventually it would be great to include it in the second transbay tube so you get a seamless connection between Martinez and San Francisco and from San Jose up to Oakland.
One caveat: the definition of route miles can get a little jiggy where you have significant segments of overlapping lines. Such is the case with both BART and Muni Metro, although one is getting the apples treatment here and the other oranges. That 104-mile figure for BART is not the total mileage of all routes; in other words, it doesn't double-count where there are two or more lines. The actual network length for Muni Metro is in the ballpark of 29 miles, which gives you a per-mile figure higher than all other US LRT systems except Boston and possibly Houston-as you would expect.
Reply Thanks Steve, those details really help make sense of things. I just redid the calculations, and it's much clearer now that Muni Metro is unbelievably more efficient than Bart and VTA, and it really makes you wonder what would happen if Geary, Columbus, Van Ness, etc. It makes it even harder to understand how we keep putting money into Bart projects when Muni Metro is where the ridership density is.
Reply Or if capacity weren't so limited by the Market subway and two-car trains. One can easily imagine the N carrying another 10 or 20 thousand riders per day. Geary buses, by the way, do around 8,000 riders per mile, which as your post makes clear is more than many heavy rail lines.
Reply The capacity issues take on even more meaning with Tom Radulovich's recent article. I wonder if/when we will finally make a large enough investment to bring Muni Metro into the 21st Century and actually make it a truly comprehensive metro system.
Yes, that's a lot of money, but I bet a full build out could capture as many as ten times the riders as Bart to San Jose for the same price. I might have to look into that more for a future post...
Reply Also, as if I couldn't get any more theoretical, my Oakland streetcar idea would get about 2100 riders per mile, which seems pretty reasonable if not a little too low. The adjusted efficiency though would be 8400 which is twice that of Muni Metro, which either means an Oakland streetcar network would be the most cost-effective system ever, or my cost estimates were a little too low. Either way, from this analysis it looks like a no-brainer.
Reply Wouldn't it be more useful to look at the rides per vehicle revenue-mile? I think that would be a better measure of the service actually provided than the miles of track.
Reply @ Steve, John - grade-separated tracks costs a ton of money, so it's not unreasonable to compare pax per mile of track. If sections are used by multiple lines, that doesn't increase the up-front infrastructure cost, just operating cost. However, shared segments create bottlenecks in the throughput limit, so non-shared track can never be utilized to its full potential. Sharing still makes sense when passengers from a highly distributed catc...
|
http://csua.com/feed/