How much money could All Door Boarding save?

This is sort of a post-live blog from Transportation Camp NE. One of the first sessions was regarding all door boarding on the MBTA. There are a lot of ins and outs—notably, that you have to account for all scenarios where people could access the system, for example potentially without paying a fare (Silver Line airport) or having proof thereof (boarded with a friend who paid and parted ways)—but it was a good discussion, and something that is moving forwards, but needs to move faster. I pointed out that the discussion needs to not be pushed by the small minority who complains (loudly) about fare evasion, or really by fare evasion at all, but by vehicle speed and efficiency, since 95% or more of passengers already pay their fare: we need to improve service for the vast majority.


Often, when we talk about all door boarding, we talk about the real and potential time savings. Muni, in San Francisco, started experimenting with all door boarding, and it turned out it worked really well, and they went system-wide, and it has saved passengers time. According to their final report, it saves 1.5 seconds per passenger boarding or alighting, and speeds overall vehicle speed by 2%. 1.5 seconds does not seem like a large number, but it begins to get a lot bigger when aggregated over a large number of passengers.
SF Muni and the MBTA have a similar number of surface passengers: about 500,000. (The T has about 400,000 bus passengers and another 100,000 or so surface boardings of light rail; looking at only surface lines, the T and Muni are actually quite similar in terms of size.) So, if we can save 1.5 seconds per person—we’ll look only at boardings, since many trips either end at a terminal station where all doors are used or are surface Green Line boardings that end in a tunnel—we wind up with 750,000 seconds saved per day. This is, rounded down a bit, 200 hours saved. The cost of operating an MBTA bus is about $163 per hour, and for a light rail vehicle $250 per hour. Let’s assume that half of that is direct operating cost: operator wages and such. Assuming the lower bound, it would save $16,000 per day. Even if there were no savings on non-weekdays, in 250 weekdays it would equate to operational savings of $4 million
Savings add up, quick.
Let’s look at it a different way. A full, two-car Green Line train in the morning carries approximately 300 passengers. On the B or the D lines, the surface portion of the route takes 32 to 34 minutes to run at peak rush hour (according to the T’s scheduled time). Saving 1.5 seconds on each of these boardings, would equate to 450 seconds, or 7.5 minutes: more than a 20% savings for the above ground route. With the addition of signal priority on the B line, you could be looking at speeding the route by 30% or more—a game changer for one of the slowest—and most heavily-used—surface lines.
The transit planners will say “well, these savings will probably just be added in to headway recovery time.” First of all, if you actually do realize a 7 minute saving, you’re talking about an entire rush hour headway, so I doubt it will all disappear, unless you are going to be lining up multiple vehicles at rush hour. But second of all, if these wind up making headway recovery times much more even, that’s great. That means you’ll have the same capacity without having to dispatch a train as soon as it arrives, but rather on even headways. This is likely to reduce the number of vehicles that wind up bunching, overloading and slowing down.
But let’s run with the $4 million figure. There are, on buses and the Green Line, probably about 1500 doors that would need car readers. If a pole-mounted reader costs $4000, the system could be paid for in a year and a half. Or if the system were assumed to last five years, you’d have $20 million to put towards the cost of the readers ($6 million) and additional enforcement ($12 million, or $2.4 million per year).
Oh, and customers? They’d get a faster ride. It’s a win-win, for everyone. Except the few curmudgeons who are less concerned with how the vehicles run, and more about the anecdote about the person they once saw jump a fare gate.