This is the category archive for "Use Cases".
Main

March 04, 2010

Days of Reckoning, Part Three

Today, thanks to skepticism from those who think my position solidified over six years on this subject is because of predetermined bias rather than actual study, I'll switch from my original plan of doing use cases by "estimated level of commute interest" and instead hit what I would guess are the two best possible cases for the Red Line.

Since shuttle-buses are obviously a problem, and since even in the commute to UT (you know, the obvious primary destination for people riding transit in our area, that unimportant little spot) from the furthest out station in Leander, the speed of the train can't make up for the time lost to the shuttle-bus, let's try to assemble one of the few commutes that might not require a shuttle-bus, although that's relatively hard to do.

Frost Tower is just on the edge of the 1/4 mile circle that most transit planners view as the maximum distance people will walk to work from a transit stop. It's also the ONLY major office building within what's commonly considered acceptable walking distance from the 'downtown station'. (Me, I might actually have to take the shuttle even on that trip some days due to my feet, so I'll plan that out too). Let's run there from both Leander (far out park-and-ride) and Crestview (supposed TOD which will supposedly provide the only real walk-up traffic for Austin).

This case also benefits the Red Line disproportionately because both the express bus route from Leander to downtown and the #101 limited first run past UT, and then past the Capitol, then through the rest of downtown; so we're at the very end of the slowest part of that route here. IE, we've picked the destination that makes the bus look its absolute worst.

Continue reading "Days of Reckoning, Part Three" »

March 02, 2010

Days of Reckoning, Part Two

Today's entry: Somebody who fell for the "TOD" hype and moved into Crestview Station so they could walk to the Red Line and take it to work at UT. Morning commute this time around; assume they want to get in comfortably before 9:00AM. Note that the Red Line shuttle drops off on San Jacinto; the two bus options here drop off on Guadalupe; the typical UT office is, if anything, closer to Guadalupe than San Jacinto.

Spoiler: Even the local bus beats the Red Line, because of the shuttle-bus trip. Yes, even though that local bus travels through half of the congestion on the Drag.

Continue reading "Days of Reckoning, Part Two" »

March 01, 2010

Days of Reckoning, Part One

Using the new schedules on Capital Metro's spiffy new MetroRail site; this afternoon in the 5 minutes I could spend, we now know that, according to schedules, if you're leaving UT for Leander and want to take the first available trip after 5:00, the express bus that currently takes you 68 minutes is on tap to be replaced by a shuttle-bus plus Red Line option that will take you either 71 or 76 minutes, depending on if you feel like taking your chances on maybe not fitting on the second shuttle bus for the 5:40 trip heading up to Leander.

TripPickup at UTArrive MLK stationLeave MLK stationArrive Leander stationTotal travel time
#987 express bus5:04 PMN/AN/A6:12 PM68 minutes
Red Line with #465 shuttlebus (first one)5:16 PM5:28 PM5:40 PM6:32 PM76 minutes
Red Line with #465 shuttlebus (second one)5:21 PM5:33 PM5:40 PM6:32 PM71 minutes

I wonder if there was anyone who predicted way back when that the Red Line would be slower, thanks to its reliance on shuttle-buses, than existing express bus service? Nah. Couldn't be. Nobody could have predicted this debacle way back in, say, 2004.

July 15, 2004:

The current commuter rail plan, for reference, requires both of these constituencies to transfer to shuttle buses to reach their final destination. This, as I've pointed out before, means that anybody who has a car and can afford parking will never ride this route.The shuttle transfer kills the performance of the transit trip to the point where only people who don't own cars or have difficult parking situations would consider it, as is the case with today's express bus lines.

More references:

March 05, 2009

Capital Metro express bus changes screw Austin in favor of Leander

Take a look at the following charts (done quickly; please forgive my lack of time on the business trip) showing some of the express bus routes proposed for elimination when commuter rail service begins:

The really fast express bus from Leander only runs obscenely early (6:00 - 6:30 AM). After that, you need to take the #987 (the one that runs down Mopac, 38th, Guadalupe), which, at least for the 'late' (7:30ish) trip, shows to be slower than commuter rail. So far so good. But what about the Lakeline Park-and-Ride, you know, the one that's "in Austin"?

Continue reading "Capital Metro express bus changes screw Austin in favor of Leander" »

February 25, 2009

Do people know they're going to have to ride shuttlebuses?

Commenter "breathesgelatin" pointed out 2 posts ago:


Mike, I have a great story for you. I went to the Crestview Station open house on Saturday. In front of me in line was a guy who asked the woman explaining the fare system the following question:

"I take the express bus in from Leander currently. It drops me off a block from my office. What bus will I need to take to get to my office now?"

The woman was completely unable to effectively explain the shuttle system, the fact that the shuttle system was different from normal bus routes, or the normal bus routes. She had clearly either not been trained, been poorly trained, or trained to cover up the idea that you need a transfer. It was really striking.

I don't think the guy was a plant; I think he was a genuine guy who wanted to use the train and was surprised it didn't actually take him to his office.

To natrius: I used to live sort of near the MLK station, on Manor. There are things you want to go to on Manor... but it is too long of a walk, to be honest. And... people are actually buying houses at Chestnut Commons?

I would rent at Crestview Station but it's probably going to be too expensive for me. Not that I would actually take the train anywhere though. I'd take the 1.

I am seriously wondering how long it will take everyone to realize that Mike has been right all along. So many are being duped by this "light rail" bullshit PR.

It's been my experience on the Capital MetroBlog that most commenters labor willfully or mistakenly under the misapprehension that they're going to walk to work from the train station. What have you all noticed with your peers, if any of them even talk about it?

Crestview Station and Commuter Rail

So Capital Metro's showing off stations. One of the ones they're most proud of is at the supposed TOD at Lamar/Airport called "Crestview Station". Let's imagine we're a new resident there and thinking about leaving the car at home to get to our job at the University of Texas.

Take a look at the following chart. Looks pretty good, don't it?

Local bus route was the #1 which seemed to get to 24th/Guadalupe as close as possible to 8:30. "Express bus" is the #101; same location and roughly same time. Pickup times at Crestview estimated to be 2 minutes from NLTC. Commuter rail travel time straight from Capital Metro's schedule to the "UT station" (MLK).

But wait. There's more.

Continue reading "Crestview Station and Commuter Rail" »

December 03, 2008

The downtown station, drawn optimistically

Erica from Capital Metro, in comments to this post, brings up the fact that the third image (originally from the city's old OnTrack newsletter, updated with green and yellow dots by yours truly), had an error in how the circles were drawn around prospective rail stations on the extension to Seaholm many people unsuccessfully lobbied for in 2004. The point of this image was to show the locations of the office buildings -- not the circles (although that is not inherently obvious if the image is viewed in isolation), and the error wasn't mine (somebody at the city drew a 1/4 mile diameter rather than radius) - but I've known about it for quite some time; using the image just to show the office locations since I have not yet created a new map with a better representation of offices. Typically when I discuss this issue on other forums, I prefer to use a google maps link like this one which shows a walk of 0.4 miles to 6th and Congress.

However, some folks at CM just produced the image below, which is about the best light you can put this 'downtown' station in, and which I will post even though it has its own problem: an attempt to fudge the issue by presenting both the legitimate 1/4 mile circle and a far less legitimate 1/2 mile catchment zone. Another discrepancy between the maps, not anybody's fault, is that in 2004, the station location was projected a half block or so farther east.

Please see comments after the image.

Important things to note here:

  • Most major office buildings are outside the 1/4 mile zone. Most are also inside the 1/2 mile range. However, using the same principle as above, note that, for instance, the second-newest big office building downtown is more than a half-mile from the train station. Essentially all major office buildings downtown, including this one, would have been within 1/4 mile of the 2000 light rail route, whether on Congress or Colorado or even Guadalupe/Lavaca.

  • The 1/2 mile radius is used as a fallback 'rule' to declare that you can attract a few more choice commuters to excellent high-frequency rail service than the 1/4 mile rule would suggest. The problem here, of course, is that the service we are providing is neither high-quality (doesn't go to UT or the Capitol or anywhere else worth going if your origin is 'downtown') nor high-frequency (runs only every 30 minutes and only during rush hours). In addition, the expanded catchment area is most suited to the residential end of the trip - i.e. you might walk farther from your home to pick up the train if it's really good - but surely not to take the train if the walk FROM the train station TO your office was extra-long - this is borne out by New York's transit agency's project to spend billions to bring the LIRR a bit closer to employment centers (see also: non-trivial unwillingness of choice commuters to tolerate transfers even from 'good rail' to 'good rail', even in Manhattan).

  • We don't have a large population of people who would be willing to walk 1/2 mile to work from the train station (and risk mistiming a 1/2 mile walk back to the train station in the afternoon only to maybe miss the once-every-half-hour train) who, and this is critically important here: aren't already riding the bus. The same people who would give the train such an incredible time investment are already going to be riding the buses from all over the city that head straight to their offices downtown. I speak from experience here: a long walk to pick up transit from the office isn't sustainable in the long-run even for transit-positive people like me. If I had to pay $10/day to park, I might think differently, but then I'd already be taking the bus, wouldn't I?

  • And, most importantly, if Capital Metro really believed that the average choice commuter would consider this train station to be within a quick, comfortable, walk of their office, they wouldn't be providing these three downtown shuttles, one of which runs right up Congress Avenue.

December 02, 2008

Hop on the Shuttle

I'm probably much more amused by myself than warranted. Judge for yourself:

Been itching to climb aboard a Capital Metro train? Understandable, given that we’ve been talking about light rail/commuter rail around Austin since the mid-1980s.

Well, that first chance will come next week when Capital Metro and the Downtown Austin Alliance host a “hop ‘n shop” at Brush Square. Up to now Capital Metro has allowed only the media and few selected others to take an up-close gander at the red-and-silver-and white train cars.

[...]

and my response:

There should really be a requirement that people spend 15 minutes sitting on board a stationary shuttle bus before disembarking and boarding the stationary train, shouldn’t there?

September 17, 2008

Getting to the ACM on transit

Here's some examples to back up the previous post about the ACM. I'm picking major intersections near some neighborhoods in the news the last couple of years. I am granting the Mueller Town Center a stop at Aldrich and Airport - which isn't really IN the Town Center, but as close as you can get today (and, I believe, the closest you'll get in the future except perhaps on the streetcar). Using times of 11:00 AM arrivals on a weekday.

Note: I had to use Capital Metro's Trip Planner instead of Google - since the Google planner defaults to "closest time" rather than "minimize transfers" or "minimize walking". This actually didn't make any difference for trips to Mueller, but it did lead CM to propose trips with transfers to downtown which arrived 5 minutes closer to the desired time than a much shorter non-stop would do (most parents would choose the non-stop that arrived at 11:05 over the longer trip-with-transfers that arrived at 11:00 on the dot). The irritating thing about CM's planner is that each and every time, you have to say "yes, I meant Congress, not South Congress" - it still won't let you say "North Congress" to avoid this. Guh.

I also had to limit walk to 1/2 mile to avoid some ridiculous options like dropping off at Hancock Center. No parent with child, not even ME, is going to walk that far to go to the ACM. Sorry.

Those who will blithely reply that buses will be rerouted to run past or through Mueller should please reconsider. The major bus routes in this city have run on essentially their current paths for decades now - none of the major N/S routes are going to move miles out of their way to run down Airport Boulevard. The most likely transit improvements as Mueller builds out are the streetcar (just improves access to downtown and the UT area, which already have a direct bus to/from Mueller) and improved frequency on the routes that currently serve Mueller - like the #350; meaning you'll still see options like the ones Capital Metro gives you below, just more often.

Remember, the point of this exercise is to think about whether this is a good long-term move for the ACM. If you are confident that gas will be cheap 10 years from now, then this is clearly a good move, except for those who don't own cars, but if you think gas might be 8 or 10 bucks a gallon by then, maybe it's worthwhile to think about how realistic it will be to get there by other means, wouldn't you say?

Additional suggestions welcome.

1. From Burnet/Koenig:

To downtown:Direct, 32 minutes on the #3; 39 minutes on the #5
To Mueller:49-73 minutes, with 2 of the 3 transferring at Northcross; the other up on St. Johns

Continue reading "Getting to the ACM on transit" »

June 13, 2008

Transportation Microeconomics Bites Me In The Butt

So you may have heard me talk about the new suburban office. For a while, we were trying to keep making a go of it with just one car - my wife driving me in most days and picking me up sometimes; other times me taking that hour and 45 minute trip home with a long walk, 2 buses, and a transfer involved. I tried to work from home as much as possible - but the demands to be in the office were too great; and we couldn't sustain the drop-offs and the long bus trips.

Well, we relented. Just in time; I got my wife to agree on a color and we now own a second Prius - this one obtained right as the waiting list shot up from zero to many months (ours was ordered; but there was no wait beyond that so it took about 2 weeks - arriving right as the house exploded so ironically I ended up working exlusively from home for a few weeks longer anyways). Do not argue with the M1EK on the futurism/economics predictions is the lesson you should be taking away from this.

So that's the intro. Here's the microeconomics lesson.

Assuming $4 gas, the trip to work in the car costs $1.56 according to my handy depreciation-free commute calculator. The morning drive takes 20 minutes. The afternoon drive more like 30.

The transit trip costs $1 (although soon to go up to at least $1.50). That means I save $0.56, at least before the fare increase, right? Not much, but every bit helps, right?

Well, the transit trip takes an hour and a half in the morning; an hour and 45 minutes in the afternoon; and I can't afford that much extra time anyways, but even if I could, it would be placing an effective value of 23.1 cents per hour on my time, which seems a bit, uh, low.

So it's gonna take a lot more than $4/gallon gas, sad to say. You might be seeing some marginal increases in ridership around here, but only in areas where transit service is very good and where people should have been considering taking the bus all along. And there's no prospect for improvement - the reason bus service is so bad out here is because Rollingwood and Westlake don't want to pay Capital Metro taxes, although they sure as heck enjoy taking my urban gas tax dollars to build them some nice roads to drive on. In the long-term Cap Metro plan, there may be a bus route on 360 which would at least lessen the 30 minute walk/wait involved, but that could be a decade or more - by then we'll probably be getting chauffered through the blasted alkali flats in monkey-driven jet boats. Not gonna help me.

Also, those who think telecommuting and staggered work schedules are more important than pushing for higher-quality transit and urban density can bite it, hard. If even people in my business often get pressure to come into the physical office, there's no way the typical workaday joe is going to be able to pull it off in large enough numbers to make any difference.

March 19, 2008

Commuter Rail Use Case #2: Leander

Continuing yesterday's post, here are a couple of use-cases from Leander; the endpoint of the line. Since the train trip would be the longest here, one might expect the train to do well - let's see.

Each table below is again based on a commute leaving the origin point at roughly 7:30 AM (for bus scheduling). I'm still taking Capital Metro at their word that the average shuttle bus trip length will be 10 minutes even though I suspect it will be worse. It certainly won't be reliable - but the train schedules will. In each table, a row just indicates a step (a travel or wait step).

Train times taken from page 4 of the PDF. Note that I now include a drive to the park-and-ride. The last example, folks, was supposed to be the "let's pretend we believe that Crestview Station will really be a TOD that people will really walk to the train station from". Updated walk time for UT for car case to 10-15 minutes based on input from Kedron et al. Note I'm assuming faculty/staff here, not students.

Leander to UT

StepDriveExpress Bus (#983)Rail
132-60 minutesDrive to park/ride (5-15 minutes)2Drive to park/ride (5-15 minutes)2
2Walk 10-15 minutes to office3Wait for bus (10 minutes)2Wait for train (10 minutes)2
3 Bus: 45-80 minutes5Train: 48 minutes
4 Walk 0-5 minutes to officeTransfer to shuttle bus (5-10 minutes)4
5  Bus: 10 minutes5
6  Walk 0-10 minutes to office1
TOTALS
Total Time42-75 minutes60-100 minutes78-103 minutes

Notes from superscripts above:

  1. Offices are more likely closer to the Guadalupe end than the San Jacinto end of campus, but that still presents a range of walking times.
  2. For the train you'll really want to be out there 10 minutes early (penalty for missing is a 30-minute wait), and 10 minutes for the bus (unlike the Crestivew case, these buses don't run very often), and the bus is less reliable to boot, but I'm including "late time" in the bus range for the actual trip.
  3. The walk from parking around UT to office is going to vary widely, but almost nobody gets to park right next to their office, whereas some people get dropped off by the bus essentially that close.
  4. A load of passengers headed to UT will actually require more than one bus to service. In other words, if we assume that the train has 300 passengers, and a third are going to UT, those 100 passengers are going to require several shuttle buses - and loading even one bus from zero to full is going to take a few minutes. Of course, if relatively few people ride the train, the bus loading would be quicker.
  5. The shuttle bus is going to drop off on mostly San Jacinto, so no need for a range here. The express bus varies widely (from personal experience) - so big range here. These express buses actually will run ahead of schedule if traffic permits - the 40 minutes is my estimate of a "quick" run based on driving time of 32 minutes uncongested. On my old reverse commute on a similar route (but only to Pavilion P&R), in no-traffic conditions, the bus took about 20 minutes compared to 15 for my car. Note that in uncongested conditions, the bus will actually get you there faster than the train leg alone - that's because the bus goes straight to UT; while the train goes quite a bit farther east, and the bus actually has a higher average speed in uncongested conditions than the train will (since the express bus goes on 183 and Mopac for miles and miles with no stops).

Conclusions for trip to UT:

  1. Like yesterday, if the destination was really anywhere near the "UT station" out east on MLK, the rail trip would be a slam-dunk winner, even with its low frequency. Even with the 10 minute wait on the front-end, it's competitive with the car and would destroy the bus. (A guaranteed 58 minutes versus a car trip which ranges from a bit better to a lot worse). Remember this when we talk again about light rail. Too bad we're not trying to build offices around that station - only residential TAD.
  2. A multi-door vehicle will be essential for loading/unloading. But even with two doors, it's going to take a few minutes to fill the seats. And the claim that the bus will always be there waiting for the train is not likely to be true based on experience with Tri-Rail in South Florida.
  3. A transfer to a streetcar would improve this only slightly. If running on reserved-guideway for most of its route, it would be more likely to be there on time, and the trip to UT would be a bit more reliable (although I'm being charitable right now and just accepting "10 minutes" for shuttle-bus anyways), but on the other hand, a streetcar that carries 1.5 to 2 busloads of people is going to take longer to load too. There's a reason transit people talk about the "transfer penalty", folks.
  4. Remember, the shuttle bus is dropping people off on San Jacinto, not Guadalupe. Go to UT sometime and see how many offices are along SJ sometime. Big mistake - but the administrators who run UT are apparently more interested in providing another spur to eventual rejuvenation of that side of campus than they are at actually serving their staff's needs.
  5. If I were in their shoes, I'd be taking the #983 already, but would actually try the train when it opens Unless you had to pay a ton for parking, though, practically zero drivers would likely not give up the drive for this train trip. If you valued being able to read/work instead of drive to this extent, in other words, you'd already be taking the express bus.
  6. Effect of future congestion increases? Much bigger than in the Crestview case. A much larger portion of the rail/shuttle trip is on the train itself - and the drive to the park-and-ride probably doesn't change; so the train ends up inching closer to the car as congestion increases - but only until we put an HOT lane on US183 and Mopac, assuming they don't do the stupid current design which wouldn't actually work. Again, though, it becomes clear that it will take unrealistically large time savings on the one leg to begin to make up for the fact that you don't get taken anywhere useful on it.

Downtown will have similar enough results that I'm not going to cut/paste for now, unless somebody really wants to see it.

Next: Mueller!

March 18, 2008

How much time are you going to save on commuter rail: part one

Capital Metro has put up a new presentation on rail-bus connectivity which also includes schedule times for the train service. Now we can see how much of an advantage this service will provide its potential passengers. Step one is "Crestview Station", a supposed but not really TOD which is located within walking distance of a train station.

Each table below is based on a commute leaving the origin point at roughly 7:30 AM (for bus scheduling). I'm taking Capital Metro at their word that the average shuttle bus trip length will be 10 minutes even though I suspect it will be worse. It certainly won't be reliable - but the train schedules will. In each table, a row just indicates a step (a travel or wait step). Updated walk time for car case based on input from Kedron et al. Note I'm assuming faculty/staff, not students.

Train times taken from page 4 of the PDF.

Crestview Station to UT

StepDriveLocal Bus (#1)Express Bus (#101)Rail
115-25 minutesWait for bus (10 minutes)2Wait for bus (10 minutes)2Wait for train (10 minutes)2
2Walk 10-15 minutes to office3Bus: 19 minutes5Bus: 12 minutes5Train: 10 minutes
3 Walk 0-5 minutes to officeWalk 0-5 minutes to officeTransfer to shuttle bus (5-10 minutes)4
4   Bus: 10 minutes5
5   Walk 0-10 minutes to office1
TOTALS
Total Time25-40 minutes29-34 minutes22-27 minutes35-50 minutes

Notes from superscripts above:

  1. Offices are more likely closer to the Guadalupe end than the San Jacinto end of campus, but that still presents a range of walking times.
  2. For the train you'll really want to be out there 10 minutes early (penalty for missing is a 30-minute wait), and 5 minutes for the bus (less penalty for missing), but the bus is less reliable, so I give both 10 minutes of "waiting time" for the bus running late.
  3. The walk from parking around UT to office is going to vary widely, but almost nobody gets to park right next to their office, whereas some people get dropped off by the bus essentially that close.
  4. A load of passengers headed to UT will actually require more than one bus to service. In other words, if we assume that the train has 300 passengers, and a third are going to UT, those 100 passengers are going to require several shuttle buses - and loading even one bus from zero to full is going to take a few minutes. Of course, if relatively few people ride the train, the bus loading would be quicker.
  5. Taking CM's word on the bus schedules here. There is going to be some unreliability built into here, but since I took their word on the shuttle bus time, I did it here too to be fair (similar traffic interference in both cases). Not as bad as the downtown case below - since I'm assuming a dropoff at 24th/Guadalupe for the local/express bus cases, there's only about a half-mile of truly congested conditions to worry about. The shuttle bus is going to drop off on mostly San Jacinto, so no need for a range here.

Conclusions for trip to UT:

  1. If the destination was really anywhere near the "UT station" out east on MLK, the rail trip would be a slam-dunk winner, even with its low frequency. Even with the 10 minute wait on the front-end, it's competitive with the car and would destroy the bus. Remember this when we talk again about light rail. Too bad we're not trying to build offices around that station - only residential TAD.
  2. A multi-door vehicle will be essential for loading/unloading. But even with two doors, it's going to take a few minutes to fill the seats. And the claim that the bus will always be there waiting for the train is not likely to be true based on experience with Tri-Rail in South Florida.
  3. A transfer to a streetcar would improve this only slightly. If running on reserved-guideway for most of its route, it would be more likely to be there on time, and the trip to UT would be a bit more reliable (although I'm being charitable right now and just accepting "10 minutes" for shuttle-bus anyways), but on the other hand, a streetcar that carries 1.5 to 2 busloads of people is going to take longer to load too. There's a reason transit people talk about the "transfer penalty", folks.
  4. Remember, the shuttle bus is dropping people off on San Jacinto, not Guadalupe. Go to UT sometime and see how many offices are along SJ sometime. Big mistake - but the administrators who run UT are apparently more interested in providing another spur to eventual rejuvenation of that side of campus than they are at actually serving their staff's needs.
  5. If I were in their shoes, I'd be taking the #101 already, and would continue to do so after the train opens.

Crestview Station to 6th/Congress

StepDriveLocal Bus (#1)Express Bus (#101)Rail/BusRail/Walk
120-30 minutesWait for bus (10 minutes)2Wait for bus (10 minutes)2Wait for train (10 minutes)2Wait for train (10 minutes)2
2Walk 0-10 minutes to office3Bus: 25-45 minutes5Bus: 20-35 minutes5Train: 18 minutesTrain: 18 minutes
3 Walk 0-5 minutes to officeWalk 0-5 minutes to officeTransfer to shuttle bus (5-10 minutes)4Walk 10-20 minutes to office6
4   Bus: 5-20 minutes1 
5   Walk 0-5 minutes to office 
TOTALS
Total Time20-40 minutes40-45 minutes33-38 minutes38-63 minutes38-48 minutes

Notes from superscripts above:

  1. Shuttle bus is likely to be much less reliable on the two routes being proposed for "downtown" than for the UT area based on traffic conditions. I've abandoned CM's 10 minute estimate in favor of a range here - 5 minutes for places close to the Convention Center on a good day; 20 minutes for the far reaches on a bad day.
  2. For the train you'll really want to be out there 10 minutes early (penalty for missing is a 30-minute wait), and 5 minutes for the bus (less penalty for missing), but the bus is less reliable, so I give both 10 minutes of "waiting time" for the bus running late.
  3. People driving downtown often have parking in their exact building (0 minute walk); but many have to park a block or more away - up to a 10-minute walk.
  4. Still going to be a bus loading wait here - varying depending on actual number of people using this service.
  5. NOT taking CM's word on the bus schedules here. Lots of unreliability when you have to go all the way past UT and then through half of downtown. I've taken their schedule times of 30 and 23 minutes respectively as about 1/4 through the range, because if buses get too far ahead of schedule, they'll actually slow down and/or stop in certain places to avoid missing pickups.
  6. The walk time here is to 6th/Congress, per my own estimate. Note that hardly anybody works anywhere near the Convention Center.

Conclusions for downtown trip:

  1. Again, the shuttle is the killer. Streetcar wouldn't help a whole lot on the loading front; but would be dramatically better on the travel-reliability front, if we get reserved guideway (would make a bigger difference downtown than on the route to UT).
  2. Note that if you were lucky enough to work at the Convention Center, your trip time would range from 28-38 minutes. In that imaginary scenario, I ride the train. Too bad we don't have much developeable land around the Convention Center for future office use. Again, this is the fatal flaw in deciding to run the train service where the tracks happen to be rather than where people actually need to go - and in this case, we can't fix it with office TOD because most of the land around the CC station is already developed - the Convention Center itself, recent hotels, etc..
  3. I'm staying on the #101, again.

One more question some are likely to ask: will worsening traffic make commuter rail more competitive on this trip? Answer: not likely. If bus travel times increased by 10 minutes in the downtown case, for instance, the shuttle bus trip is likely to increase too (5 more minutes, say) -- meaning that the two modes' total travel time really just continues to overlap, and on the low end of the rail/shuttle range to boot. Again, fatal flaw time: if you're trying to sell people on a transit trip with reliable time characteristics, you can't run a shuttle bus for the last half of the trip!

Next: Leander.

January 15, 2008

TFT: Suburban wasteland

As alluded to at the end of this crackplog, my company just opened a physical office in a truly awful part of the suburban wasteland. Today was the test case for "how bad is the trip home on the bus", after getting rides to/from work with my wife and a travelling coworker all of last week (not so bad in the morning; but awful in the afternoon, especially for my wife, who had to invest 30-40 minutes getting to the office to pick me up to then spend 30-40 minutes going home). Ironically, this would be a great bike commute, if I could still ride my bike any non-trivial amount.

I'm still not sure how often I'm going to need to come in, but there's a sliding scale here - at some point it'd require us to get a second car, which I don't want to do for many reasons, not least among them financial (we couldn't have taken our trip to Hawaii if we'd had a second car payment, after all). There's a certain number of days per month on which we could tolerate a both-ways drive (very little); a larger number where we could tolerate a drop-off in the morning and a bus ride home (determining that right now); a larger number which might be achievable on something like a scooter, if I can get past some emotional barriers; and anything else requires that second car. At which point I also have to consider other options, because if I have to lay out the money and time for two cars, might as well look for somewhere that can make up the gap (or maybe downtown, or at least in a less awful suburban part of Austin where you can actually take the bus).

I am writing this on the bus - filling in links later. It's a crackplivebusblog!

Google transit called this trip a 10-minute walk, a 26-minute bus ride, a transfer, and another 20ish minute ride from there, the last leg being one on which I can take about six different routes home, so no worries there. I was highly dubious of google's estimation of the walk, having ridden this route many times on my bike, back when I still could, so I gave myself 25 minutes to walk and 5 minutes to wait (buses can and sometimes do arrive early).

Update on the next day: Now google is accurately saying 19 minutes for the walk. Huh.

Walking trip: Got to the elevator at 4:03 (after having to run back in and use office phone to call home, since cell phone battery had died). Started on the long, not so scenic, walk through suburban Westlake. Guh. No sidewalks, of course, on Allen (behind the Westlake High tennis courts and other fields). Pretty decent sidewalks after that on Pinnacle, which I took the rest of the way down. Walked past some middle schoolers who will doubtlessly be telling their friends they saw a Real Adult Walking - must have been a bum or a predator. Got to the bus stop at 4:20. Whoops - although google was way too optimistic, I was a bit on the pessimistic side. Would budget 20 minutes for the walk next time, if it happens, plus the 5 minute wait.

First bus leg:

  1. 8 people were on the #30 bus as it pulled up (exactly on time at 4:33). I made 9.

  2. 5 more people got on at Walsh Tarlton and Bee Caves. Total on bus counting me now 14.

  3. 1 more guy got on in the weird office park at the end of Bee Caves. 15 people on the bus now. Bus goes through a road at this complex and then turns up Spyglass to make a short loop in the wrong direction, at least for me.

  4. 1 more got on somewhere along Spyglass at one of the apartment complexes. 16 people now!

  5. #17 got on at Spyglass / Barton Skyway.

  6. At Spyglass, near north intersection with Mopac, one got on and one got off. Still 17.

  7. Turned back onto southbound Mopac at 4:44. Guess that loop was worth it after all. Stopped for a couple minutes at the Bee Caves light, and then another 3 got on! We're essentially at standing room now - one standing, although there are a couple of seats left. 20 passengers.

  8. At 4:48, we turn into a bus bay to pick up a guy with a bike. That makes 21 passengers.

  9. We cruise through Zilker Park without stopping and arrive at Robert E Lee at 4:51. Not a good day to be hitting the park anyways - but someday remind me to write a crackplog about how the city needs to jack up the parking prices there in the summer quite a bit higher. Still 21 passengers. A Barton Hills bus (#29) turns off Lee with about ten people on board that I can see (maybe more).

  10. Amazingly, they're still working on that Villas of Lost Canyon project. We arrive at the backup for the Lamar light at 4:53 and almost hit a bicyclist stopped in the right lane for no apparent reason. We're back in civilization, as I see real adult people with apparent jobs walking about like actual pedestrians. Hooray! Stuck for a bit behind our friends on the #29 as they load a bike. Boo. Driver may not make my promised 4:59 drop-off if he keeps this up.

  11. 4:54: Somebody finally pulls the chain to be let off in front of the Armstrong Music School. Down to a mere 20. The bus is practically empty! The suburbanites are right!

  12. 4:55: Lady gets off at the corner of S 1st. Down to 19 people! I think I see a tumbleweed.

  13. 4:58: D'oh. Somebody signals they need off just past Riverside. Going to be hard to make my best transfer at this rate. Time to hibernate the laptop now, though; the rest of first leg is from memory. About 10 people got off at that stop! Holy cow. Down to 7 passengers now. All of those passengers walked over to S Congress to hop on one of the many buses that pick up on the other corner, by the way.

Transcribed later on from here on out.

The wait: Had my bus been just a minute earlier, I could have immediately jumped on the 4:59 #7 bus which was a few minutes late. Rats. As it turns out, my #5 bus was quite a bit more late.

Second bus leg (transcribed today from yellow legal pad - since the ride was way too jerky and crowded to crack open the laptop):

  1. 5:10: Bus arrives; I board. About 15 people on the bus.
  2. 5:11: 14 people still on at 7th/Congress.
  3. 5:13: 3 more get on at 9th/Congress.
  4. 5:14: One got off at 10th/Congress
  5. 5:16: 3 got on as we turned in front of the Capitol at the bus stop that our asshat governor is forcing to move. There were about 30 people there at that time. Up to here, 'rapid bus' on this corridor would have saved about 30 seconds of the 4 minutes it took to traverse Congress which is actually a bit better than I would have guessed. Not that the #5 would get that treatment anyways, but it was something to look at while we were stuck in traffic with the #1/#101, which would be the rapid service. Streetcar would have been no better than the bus I was on in this part of the route - but at least no worse.
  6. Note for comparison's sake that light rail on this route ala 2000 would have probably taken about 2 minutes. About two stops; no being stuck behind cars or other buses. Moving on...
  7. 5:17: Lavaca at 12th and 13th, one got on at each. Ride is getting even jerkier and crappier. Good thing I didn't take out the laptop.
  8. 5:18: One more gets on at 16th.
  9. 5:18-5:24: We're stuck in a very long backup from the light at MLK/Lavaca. This is where LRT would really have helped. As it turns out, streetcar would have been even worse because we saved a minute or two at the end by prematurely jumping into the center lane (bypassing a stop on the right where nobody was waiting). The streetcar, stuck on the tracks in the road, can't make that decision. This helped a bit because the primary backup from this light was traffic heading to I-35 - the tailback in the right lane was about a block longer than the one in the center lane and moving much more slowly too.
  10. 5:24: Driver guns it to try to make up some time, as by this point we're really really late. Note: this is why people who say you shouldn't have rail until you can run the buses on time are idiots - the driver did everything in his power, but all the cars and a few other buses made it impossible for him to meet his schedule.
  11. 5:26: We slowly approach light at 21st/Guadalupe, having been stuck through several light cycles. Now we see why "Rapid Bus" won't work at all - and the same thing would apply to "Rapid Streetcar". The entire corridor is congested - we can rarely make the first green light we see all the way past UT, and quite often don't even make the second one. At this point, a whole ton of people get on, and the bus is now standing room only, with 3 people standing and every seat full.
  12. 5:29: Stuck short of 24th. Once again, rapid bus shows its uselessness - as we could have held that light green till the cows came home, but the traffic from 26th through 29th would have still stopped us dead. At this point we're probably more than 10 minutes behind schedule.
  13. 5:32: Finally made it to near the Dean Keeton / Guadalupe intersection; finally about to leave the "rapid bus" route (and also the light rail route). Note that light rail as planned in 2000 would have breezed through this stuff - making a couple of stops, but never getting stuck in traffic. The driver really goes fast on Dean Keeton - feels like 45, although it's very hard to tell.
  14. 5:34: We pull over near the ped bridge over Dean Keeton and pick up a few more people. About 5 people standing now.
  15. 5:36: Finally on the way home. No more delays/obstructions.
  16. 5:38: Three people, including yours truly, disembark. Some of the remaining standees find seats. Bus has improved to only 9 minutes late, thanks to some speeding and 'flexibility'.

Things learned:

  • Don't trust the pedestrian part of google transit's directions. I kind of suspected this before, but they clearly assume you can take a bees'-line. It would be a much better idea if they were to assume you had to take the same route as your car - they'd be erring in the conservative direction if at all - which is definitely the better way to err when walking to a bus stop!

  • They might be able to run the #30 a bit more often, if this is any indication. At least a bit more frequent during rush hours, as the people on the bus were (mostly) clearly headed home from work.

  • As another commenter alluded to on his blog, this is the kind of thing Ben Wear should be doing from time to time.

  • Rapid Bus is shelved, of course but today's experience yet again confirms how useless it would be. Likewise, streetcar on this corridor in a shared lane would be an absolute disaster - even worse than the bus. Broken record time: Light rail as conceived in 2000 would have greatly helped this corridor - giving people a transit alternative which would be superior to the private automobile and FAR superior to slow, unreliable, jerky buses or streetcars.

January 04, 2008

Why transit service doesn't work on frontage roads

This has come up frequently in the past in regards to the idiocy of claiming that major retail belongs out on the frontage road (where I have claimed in the past that it's impossible to practically provide good transit service). Here's a much better version than my previous one, and as a bonus, MS Paint was still tangentially involved!

(For non-Texas readers who may have wandered in from Jeff's excellent transit portal, almost all limited-access highways in this state are built from pre-existing major arterial roadways - where property access is maintained via the construction of new "frontage roads" which unlike perimeter roads often used for that purpose in other states, also serve as on-and-off-ramps. The incredibly wide road footprint that results makes it far more expensive to build new or maintain existing crossings over or under the highway).

Both images from google transit; click through for full details. This is basically the "how do I get from the drop-off for the express bus at the park-and-ride on the west side of the road to the entrance to all the office parks on the east side of the road". Note that the address for the park-and-ride you sometimes get (12400 Research) doesn't match the actual location, which is on Pavilion Boulevard back towards Jollyville.

First, the transit directions, which look pretty good at first:

Then, the driving directions, which look like this:

Huh. Wait a minute. If I can just jump across the road, why do the driving directions have me go down a mile and back? Let's look at the satellite image:


Oh. Now I see. Note that the bus stop images you see on the other side of the road are for a poorly performing cross-town route which suffers from the same basic problem - if you need to leave an office on that side of the street and go southbound on 183 back home, you get to walk to the next crossing - which on a normal street wouldn't be that big of a deal, but crossings of frontage roads are few and far between. Farther to the northwest, crossings are even less frequent - you face a walk of close to 3 miles in spots to make this trip across the freeway. Taking that cross-town route would be even worse than taking the express plus the incredibly long walk, because it would require a long slow trip down the frontage road and then a transfer to a second bus, and because the service on the frontage road is inevitably low-demand, it doesn't run very often either.

Keep in mind that this is just to cross the freeway. If you work at the Riata office park, you then face another walk of a half-mile or so inside the complex. I used to do this commute on my bike, with bus boost in the morning at times and am very familiar with the area - ironically, proximity to the Pavilion transit center was supposedly touted as a positive for this development when it was originally proposed. I was always pretty sure Pavilion used to connect with what is now called Riata Trace Parkway when 183 was just a six-lane divided arterial but have never been able to find a clear enough old satellite image to confirm, but our Tennessee correspondent has already confirmed in comments that it did cross.

For reference, my last job before this one was also on US 183, but between Balcones Woods and Braker Lane, which was much more accessible by transit - and yes, I did sometimes take the bus even on days where I wasn't biking. I tried the bus commute once to Riata and never did it again - that walk, in addition to being far too long even for a nice comfortable express bus, is just dreadful, even compared to conditions down by Braker.

And, yes, there's a personal reason this is coming up now too. All I can say now is dammit, dammit.

January 01, 2008

Why Streetcars Suck: Simple Example

Very quick hit today; not even any links, although I may fill them in later if I get a minute.

My family took the #5 down to Town Lake yesterday for the First Night festivities (the parade was outstanding - best one I've ever seen). One simple thing we experienced shows why streetcars in shared lanes are completely useless.

We're travelling southbound in the right lane of Congress (where shared-lane low-budget streetcar would inevitably operate as well). Oops, a car has stopped and is unloading a bunch of stuff with their flashers on. The bus driver quickly changes to the center lane to get around them and then moves back right.

A block later, somebody starts to pull out of one of the angle-parking spaces and stops. I was never able to figure out why - they may have been spooked by traffic. Again, the bus driver changes lanes and moves around the obstruction.

Anybody see the problem with shared-lane streetcar yet?

You get enough little blockages like that and the performance and reliability of the streetcar gets so bad that even the mystical streetcar fairy dust that supposedly makes commuters forget how to read their watch won't help.

True light rail, with reserved guideway ("running in its own lane") is a slam-dunk win for Austin. But shared-lane streetcar is a complete waste of time that actually performs worse for passengers than does the city bus that most of them won't even take today.

Unfortunately, I have my suspicions that the Wynn/McCracken rail plan will end up having to rely on mixed-traffic streetcar service for a good chunk of its proposed route (and that's only one of the two impending problems; the other being that the route absolutely must go up Congress and then Guadalupe, rather than over the east side of UT and then out to Mueller as in the useless Capital Metro proposal). So, once again, we're scrod by our pal Mike Krusee - because of his push in 2000 to destroy Capital Metro, and then his push in 2004 to force commuter rail instead of light rail, urban Austin will probably end up with no rail at all, or, at best, rail which is actually less useful than city buses.

September 27, 2007

TWITC: The Domain and The Bus

Starting a new category - "This Week In The Chronicle" where I post a short response to a couple of articles matching my subject matter here. Subtitle for this category should be "In which M1EK performs the critical analysis that we used to rely on the Chronicle to do, instead of just fleshing out Capital Metro / city press releases".

Both about The Domain today, which is actually a pretty nice little project in the middle of suburban crap.

First, the main article which includes this:

Each TOD, inevitably, has separate demands, different problems, and a different mix between the core components. "No TOD has everything," said Lucy Galbraith, TOD manager for Capital Metro. "Some will primarily be employment centers, some retail or residential. Nobody ever gets everything in there – except maybe Downtown Manhattan."

So what do they have in common? "It's the three D's: density, diversity, design," explained Galbraith. Density isn't about buildings per acre but bodies. It means enough people to make the area feel like a community. There's a psychological factor, that a busy street is a comfortable street. "If you're the only person walking, it can be a little lonely," Galbraith said. "If there's 50 people walking, you feel fine." Similarly, diversity is supposed to reflect not just the usage but the culture of a TOD. "It's incomes, housing types, ethnicity, everything you can find," she added, "because the full range creates the kind of all-day use that makes it a healthy, lively place."

But the third and most critical component is design. Transit plans depend on road design, and a transit plan that hopes to balance public, private, and pedestrian traffic needs to get it right early on, because fixing a road is a lot harder than building it in the first place. According to Galbraith, for a really successful TOD, that means putting people-on-foot first. "There's many technical details, but basically you think about how you make life easy for the pedestrians, and then you fit in everything else."

And my response:

As I've said before, you never, ever, ever get TOD with anything but high-quality rail transit. Note: the rail transit has to be within walking distance of the TOD for this to work - a 'circulator' shuttle bus will absolutely NOT work. Also note, the same lady quoted here has previously attempted to claim that the Far West and Riverside student ghettoes are TOD.

Wishful thinking pushed by the Feds aside, the general opinion in the field is that obvious and frequent bus service is arguably an impediment to high-quality TOD, because it drives away the tenants most in demand (choice commuters). The only thing that appears to work is rail transit within walking distance, period.

Sub-article, on "Getting There":

One concept being considered is a circulator shuttle-bus service that will pick up train passengers and distribute them through the area. It will mean less of an overall dependence on the ubiquitous Cap Metro big bus, but it's not exactly virgin territory for the city's public-transport system. "Our range is a little longer than people perceive, because not everyone sees our express buses or our smaller special-transit service shuttles," said Lucy Galbraith, TOD manager for Cap Metro.

Response:

Even in true downtown areas, circulators are a huge disincentive to choice commuters. In an area like this, which is a pale shadow of downtown, they're going to be a killer. Imagine the use case here, from either central Austin or Leander:








#Segment typeDestinationNotes
From Leander
1DriveTo park-and-rideNot realistic to pick up circulator buses on residential end in Leander
2WaitFor commuter rail trainRuns every 30 minutes during rush hour only for first N years, maybe as often as 15 minutes many years later
3TrainTo Kramer stationStation is way east of Domain - behind IBM/Tivoli
4BusFrom Kramer station to DomainProbably no wait here (circulators timed to train arrival) but bus stuck in traffic
5WalkFrom bus stop to destination(short walk)
From Central Austin
1WalkTo shuttle bus stopNo parking at the few stations closer in than Kramer, so only way there is bus
2WaitFor shuttlebusModerate to long wait. (Timing only guaranteed on train end).
3BusTo station (one of three)Slow, jerky, stuck-in-traffic ride
4WaitFor commuter rail trainRuns every 30 minutes during rush hour only for first N years, maybe as often as 15 minutes many years later. Only one reverse commute per day initially.
5TrainTo Kramer stationStation is way east of Domain - behind IBM/Tivoli
6BusFrom Kramer station to DomainProbably no wait here (circulators timed to train arrival) but bus stuck in traffic
7WalkFrom bus stop to destination(short walk)

Now, compare to driving. Does either one of those trips look remotely attractive enough to get you out of your car? The whole point of transit-oriented development is that the trips to and from the development must be served as well or better by transit as they are by the automobile. Unless you're smoking a particularly potent brand of crack, commuter rail service plus shuttlebus to The Domain will never in a million years, even with gridlock, be better than just driving there.

What could have been done differently? The 2000 light-rail proposal would have knocked off items 2 through 4 from the Central Austin use case above; and light rail could eventually have been routed directly into The Domain (someday removing the other shuttlebus trips from both cases). The DMUs being used on this commuter rail, on the other hand, will never be able to be run in the street, even up there, because they can't make anything but the widest of turns. Once again we see that the decision to implement commuter rail instead of light rail not only buys Austin absolutely nothing now, it prevents us from doing anything better in the future.

September 19, 2007

Driving: Fixed versus variable

Wanted to point readers to a discussion between Austin Contrarian and myself about fixed versus variable costs of driving, and how best to account the fixed costs. One thing many commute calculators get just absolutely and stupidly wrong is the idea that depreciation is a factor of miles driven (it's actually far more a factor of age - miles a distant second). For instance, this is a comparison I ran for one of the early comment's on AC's post: a 1998 Honda Civic LX, automatic, all other values default. All values as "private party" and "excellent condition".

Miles drivenValue
20,0006,540
90,0005,640
180,0004,825

(KBB said 86,000 would be "typical", but actually seems a bit low to me).

A 1998 Civic would be either 9 or 10 years old today, depending. The added depreciation due to driving normally versus the little-old-lady case is no more than $100 per year ($8.25 per month). The depreciation due to age swamps this figure by a factor of 10 or more. This stands to reason - would you really pay a ton of money for a 9-year-old Civic just because it wasn't driven very much? Of course not.

What does this mean? Ignore the commute calculators which include depreciation, insurance, and registration, unless you're one of the vanishingly rare few people who can completely get rid of a vehicle. Instead, use one of the calculators which only includes truly variable costs, like mine (originally written for bike commutes, but can be used to compare the cost for transit commutes just as easily - just zero out the cost of bike tubes and tires and put bus fare in "extra costs"). For instance, at gas prices of $3.00, and with $80 tires (about what our last set cost, each), you end up with these values for some of my old commutes (assuming I got to use our Prius instead of what I actually drove back then, which only got 38 mpg):

TripCarBus
Home to 183/Braker (Netbotz)$1.31$1.00 / $2.00 (regular / express)
Home to downtown (free parking)$0.46$1.00
Home to downtown ($8 parking)$8.46$1.00

Now, if AC's parking cost was unbundled - charged per-day, his commute would actually come out cheaper on the bus by a fair margin, as indicated above. He indicated a monthly cost of $100, and I'm just guessing that $8 might be the price on the spot market, but that means that if he drives even about half the time, it'd be smarter to pay for the parking pass and then drive every day.

How can we fix this? If more of the costs of driving were borne directly by drivers, at the time they drove (or at least paid for gas), it wouldn't be so artificially cheap. For instance, when I drive downtown, I'm using roadways which were paid for out of property and sales taxes - not the gas tax. If we were to pay for all major roadways out of the gas tax, well, first, Round Rock would start to have to finally pay something approaching the cost of their infrastructure without free-riding on Austin, and second, at the extra buck a gallon I figure it would take, the math would shift a bit. It'd shift more if we could get auto insurance priced by the mile (although you keep hearing about it, it's never been an option for me or anybody I know personally). And, of course, if we paid for the costs of our Iraq adventure by gas taxes instead of through income taxes, the story would be even more different. But in the meantime, it rarely makes sense on purely economic grounds to ride the bus, even at our currently way-too-low fares so we're going to have to keep working on the other reasons. Like reliability. Light rail, dependable in time and at least competitive with the car, on which you could comfortably work or read, would be an easy winner. City buses - well, I salute AC and Tim for being able to work and/or read on the jerky city buses, but I was never able to, and I doubt most people would consider it acceptable even if saving a couple of bucks. Unfortunately, of course, our brand-new commuter rail line is going to inflict two of those jerky bus rides on every single rail passenger every single day. Oops.

September 07, 2007

Difference between streetcar and bus

Since many people still think that if you build streetcar, they will come; here's a set of use case-like tables which I hope will explain what the actual difference is between streetcars and buses. The first case is for "why can't we just fix commuter rail by building a streetcar line to which they can transfer?". The second case is for "won't streetcar get more residents of central Austin to take transit to work?".

Some shorthand below explained up here:

"Stuck in traffic": Does the vehicle have its own lane, or is it sharing a lane with cars? This affects speed and reliability.

"Detourable": If there's a traffic accident in the shared lane, can the vehicle in question change lanes to get around it? This is a drastic impact on reliability.

"Fast/slow": Is the vehicle capable of accelerating/decelerating quickly? Speed, obviously.

ModeStuck in traffic?Detourable?Fast/slow?
Circulators as applied to commuter rail service
ShuttlebusYesYesSlow
StreetcarYesNoSlow
Mode by itself (for residents of actual central Austin)
ShuttlebusYesYesSlow
StreetcarYesNoSlow

Notice anything? Whether you're using the vehicle as a circulator or as your primary form of transit, it performs exactly the same. I know this seems obvious, but I still get people thinking that there's some magic fairy dust that will make streetcars turn into good transit service for the people who actually wanted it, in both 2000 and 2004. No, credulous fellow residents of Central Austin, streetcar doesn't bringing anything more to the table than bus does - arguably LESS, for daily commuters. Note the "Detourable" column. Yes, I've had times on the bus when I've benefitted from this capability. They won't detour just to get around heavy traffic, but they darn sure will to get around an accident.

So what are some of the other benefits of streetcar not mentioned here? It provides a perception of permanence that bus service does not. This is worth something if you're trying to stimulate development somewhere - but downtown Austin doesn't need the help. It also provides a minor benefit for tourists - making it more obvious that transit exists, and making it more attractive (people from out of town are unlikely to want to ride the bus given the stigma of bus service in many other cities).

The only advantage streetcar has is for tourists - which is why, IF we build this thing, it should only be funded out of hotel/rental car taxes. Even if it ran through the dense residential parts of Austin, it would provide precisely nothing of benefit to those residents, who, by the way, pay almost all of Capital Metro's bills.

April 04, 2007

You can't have TOD without good T

Don't gimme no crappy transit, fool!

So the Statesman and the good folks at Austinist are falling prey to the hype about the TOD around the new commuter rail line. Let's see how attractive the "T" component of the "TOD" will be for Crestview Station, the one the Statesman most recently covered. Remember that without high-quality transit, you don't achieve the true benefits of TOD.

First, let's consider Paula Professor. She lives at Crestview and works at UT. The first map below (click for expanded version) shows her ride on the commuter rail train. So far so good! She's able to walk to the train station, and even though the trains only run every half-hour, that's not that big a deal on this end of the trip; she just plans ahead. The train ride is quick; and is not held up by traffic.

But wait! Why is the train stopping out here off of MLK, way out in East Austin? Paula wanted to go to UT; her office is between Guadalupe and San Jacinto near 24th street. Well, the signs at the station inform her that this is the UT stop, so she gets off. Ah, here we go: a shuttle bus marked "UT". Well, she's rather committed now, so might as well get on and see. Here we go:

The shuttle bus took 15 minutes to travel about two miles. Stuck in traffic behind the cars of all the people that drove to work. "What a pain in the ass," thinks Paula, "if I was going to be stuck in traffic on the bus anwyays, why didn't I just take the #1, or better still, the #101 express, which go straight where I want to go? Or better yet, just drive. Maybe in 2006 2007 2008 2010, I'll just take the Rapid Bus there".

On the way home from work, Paula missed her shuttle bus by five minutes, and ended up having to wait 25 minutes for the next one, which again took her back through heavy traffic, very slowly, to the commuter rail station. "What happens," Paula wondered, "if my shuttle bus misses the train departure because it's stuck in traffic? This thing only runs every half-hour during rush hour and not very late into the evening"

Paula ain't gonna ride this thing again, folks.

Now on to a worker at the Capitol, who I'll call Steve Staffer. Steve does the same thing as Paula; he walks to the train station. So far, so good! He rides the train, just like she did. Great! But at this station off MLK way out in east Austin, he sees that Capitol workers are supposed to depart, just like UT workers. Hmmm. Well, on to the shuttle bus:

"Wow," said Steve, "I didn't believe Paula when she told me how lame this ride on this slow, jerky, stuck-behind-cars shuttle bus was. Now I do."

What's Steve's better option?

Wow. Looks just like the 2000 light rail proposal, doesn't it?

Finally, Larry Lawyer, even after hearing the complaints of Paula and Steve, decided to ride the train anyways and catch up on his paperwork. "Wow," he thought, "this is a lot more comfortable than the bus - and easier to work, but why the heck have I gone so far out to the east only to loop back here to this corner of downtown where there's nothing but bums and the blank wall of the Convention Center?"

"I got off the train," Larry explained later, "and there was a shuttle bus there that said 'downtown', but I already was supposed to be downtown, since that's what this station is called! So I just started walking. I walked. And walked. And walked. By the time I got to my office on Congress Avenue, I had walked half a mile. More than I ever wanted to walk from the train station. I thought this thing was supposed to be right in the middle of downtown? On the way home, I took the shuttle bus instead. Not much better - a ten minute tour of downtown on a herky-jerky bus just like that Dillo that I tried once a few years ago and never went back to. I think tomorrow I'll just take the Lexus straight in. Isn't there a better way to do this?"

The common thread in all three of these "direct" pictures, in case you missed it, is that they all precisely match the expected route from the 2000 light rail proposal, which is now impossible to build thanks to commuter rail. We may get higher-density development at these spots simply because City Council upzones them to closer to what the market would like to provide in Central Austin, but it's pretty darn clear that most "choice commuters" (people who can afford to drive to work, and, obviously, afford to live in these developments) will just be driving to work as usual unless we deliver transit service which doesn't require a stupid shuttle-bus or even streetcar transfer. Go back to the the link from VTPI about the difference between TOD and "transit-adjacent development", and pay particular attention to this item:

Transit service is fast, frequent, reliable, and comfortable, with a headway of 15 minutes or less.

Even if we run commuter rail trains more often, a trip which relies on a shuttle bus travelling through mixed traffic for the last two miles or so will never be reliable or comfortable. This is why our friends at Tri-Rail have egg on their faces year after year after year as the promised TOD around stations never materializes. Here in Austin, we're likely to get at least medium-density development at Crestview Station, but the residents still aren't going to be enjoying the true benefits of TOD, and neither is the city.

March 27, 2007

The Great Interconnections Lie

The North Burnet/Gateway presentation, which, frankly, looks very very appealing in the alternate universe where we had the guts to stand up to Mike Krusee and develop light rail, continues the rationalization of poor transit service by calling it "interconnected", which is a euphemism for "you're going to have a lot of transfers". Specifically, a resident of this area trying to go downtown would need to first board a circulator (probably a bus) to get to the commuter rail station between Metric and Burnet, then wait for the train, then ride the train to MLK or the Convention Center, then switch to another circulator (probably again a bus) to get to UT, the Capitol, or the parts of downtown where people actually work.

And yet nobody sees this as a problem.

Today, all you have to do is spend some time outside the transfer centers at Northcross or Highland Mall, and it becomes abundantly clear that the only people who use bus service that requires a transfer are the utterly transit-dependent (not the choice commuters we're supposed to be serving). So we're going to build a rail spine for our transit network that requires at least one transfer to bus for anybody to use.

And yet nobody sees this as a problem.

Light rail, as promised here in 2000 and delivered everywhere else in the meantime, on the other hand, is designed to serve as a one-seat ride for the majority of riders (two seats for suburban users of park-and-rides). Let's compare and contrast again:

Suburban users:

Light Rail a la 2000Commuter Rail With Interconnections(tm)
1Drive to park-and-rideDrive to park-and-ride
2Wait for train (every 10 mins rush hour; every 20-30 otherwise)Wait for train (every 30 mins rush hour; no service otherwise)
3Ride train to stationRide train to station
4Walk to officeWait for circulator (probably bus)
5 Ride bus (stuck in traffic)
6 Walk to office

And now, for this "second downtown", we're being sold on the idea that "interconnected transit" with "circulators" is the way to go, meaning that the commuter in the right column will actually be adding another bus ride at the beginning of their trip.

Folks, even in Manhattan, routes that require transfers see a substantial drop in ridership, yet somehow we think that our comparatively low-density city is going to do better? Even when our transfer is to a jerky, slow, stuck-in-traffic bus? And now these idiots working on the Burnet plan think a bus ride on the OTHER END is actually a POSITIVE?

(No, streetcars won't help; they're still stuck in traffic behind everybody else's car).

Somebody other than me's got to start talking about this stuff so it's not such a surprise in 2008 when nobody rides the thing. Please, for the love of god, somebody speak up. Ben Wear? Wells Dunbar? John Kelso? Somebody hep me!

And, no, this is not a problem we can fix with better circulators. Remember, the Manhattan transfer commuters go from one reserved-guideway rail vehicle to another reserved-guideway rail vehicle, and yet it still cuts their ridership by a substantial percentage. And that's in a town where you have to lay something like 50 bucks a day just to park that car.

Start here to learn about all the places New Yorkers are still trying to eliminate transfers.

September 07, 2006

Chronicle remains credulous

In today's story about the new effort to align CAMPO dollars to Envision Central Texas goals, not once, in the entire story, was this fact mentioned:

The three biggest "nodes", now and in the future, by orders of magnitude, are UT, the Capitol, and downtown; none of which are served by commuter rail, and not well by streetcar. If you live at Mueller and work at the Capitol, you can take the streetcar to work, but it'll be as slow as the bus is today, and that's the only use case that makes sense. All existing residential density in the city continues to be provided with nothing but slow, stuck-in-traffic, buses (mislabelled as "Rapid" though they may be).

Summary: Until the elephant in the tent is addressed (those three nodes), all of this is just useless ego-stroking wastes of time.

August 23, 2006

Think Like A Passenger

Neighborhood groups are crowing over the results of the Capital Metro streetcar workshop which is, frankly, just a load of barely-informed fluff that anybody who's bothered to ever ride a transit line of any type knew about three minutes after getting on the bus or train. Capital Metro holds these things mainly in order to appear as if they're accepting input from the community - I'll write about that someday if it bugs me a bit more than it already does.

As usual, what's missing from this entire thing is, getting back to the old microeconomical view, why would somebody decide to ride this thing instead of driving their car?

Take as a given that we're talking about 'choice commuters' - i.e. those who could, and today do, drive to work. So look through the series of comments from this workshop and see if you can find even one which addresses, even obliquely, the reasons why people don't take the bus today (the entire streetcar corridor is served quite well by buses which run almost as frequently as this streetcar would).

See anybody talking about signal pre-emption (a la Rapid Bus)? Nope.

See anybody talking about reserved guideway (a la light rail)? Nope.

There's about one place where the "why is this better than a bus" question is even asked/answered, and it boils down to what I always say: a modest improvement in attraction due to perception of permanence and a slightly more comfortable ride. It's not any faster than the bus; nor is it going to be any more reliable. People who try it are very quickly going to figure this out - so you're left with luring tourists, which is, I suppose, a worthy goal, but then why are we spending all the money to drag this thing out Mueller-ways? Again - people living in Mueller and working downtown are going to figure out after a couple of trips that the streetcar may look nicer than the bus did, but it's still very slow and still very much stuck in traffic, so might as well go back to driving.

Think about it this way: We've got a passenger. His name's Joe Mueller. He lives in the new development out at the old airport. He drives to work today at the Capitol. Many days, traffic is bad, and he has to either suffer through traffic, or shift a few blocks over and try to make up some time on a different road. Why doesn't he take the bus today? Well, he sees the buses every day on the same road he (usually) drives. They stop a lot; accelerate poorly; and can't shift to another street when there's an accident or congestion on Manor, for instance. What could you do to get this guy on transit? Well, cost isn't going to work - he has free or cheap parking, and the variable cost of driving is trivial. But taking a big chunk out of the current gap in speed and/or reliability might do it - and in other cities, actually has worked. So, is the streetcar going to be faster than the existing bus? More reliable?

Somewhat depressing is the Chronicle coverage of the session - in which the author conflates light rail with streetcar, and continues the Chronicle's perfect record of refusing to analyze the difference between "good rail" and "bad rail". At least they gave my colleague Patrick Goetz some play - but that makes it sound like the only choices are streetcar or monorail, which plays right into the hands of Krusee. Light rail as in 2000 would have run on the ground, for a fraction of the cost of monorail, and provided most of the speed and reliability benefits of truly grade-separated transit. Somehow, I've got to find somebody else in the world who can get a bit deeper than "rail bad" or "rail good" to "this rail bad BECAUSE".

The most depressing thing of all, though, is that TWO CITY COUNCIL MEMBERS are apparently dumb enough to fall for this hype and think it's going to make any difference. Sigh. I had hoped that McCracken, at least, was going to be pushing for something like light rail for the center-city, but now I see all he's doing is pulling the same crappy sled as the rest of them.

July 19, 2006

Where does the commuter rail line end downtown?

On this forum, some folks are naively optimistic about how close the commuter rail line comes to major employment centers downtown (one even argued, although was corrected, that people would walk the 2+ miles from the MLK station to UT every day!). I dug up the picture below, and added in a legend and drew in the route of the 2004 commuter rail line as well as the 2000 light rail line. I'm not enough of a photoshop wizard to remove the other three "possible station locations" - this image was originally from a city of Austin newsletter about possibly extending the commuter rail line west to Seaholm.

Note that the typical 1/4 mile catchment area around the station at Red River and 4th Street doesn't go anywhere near any big office buildings - the only big buildings it captures are some hotels - whose employees aren't the "choice commuters" a new rail start should be going after anyways. A quarter-mile radius is typically used as an estimate of the maximum amount of distance that the typical daily commuter would be willing to walk from the train station to their office - any more than this, and they won't take the transit trip (or, as Capital Metro would hope, contrary to all of the evidence from Tri-Rail in South Florida, they'll be excited to be "whisked to their destination on shuttle buses").

Also note that the Capitol and UT are much, much, much farther from any stations for the commuter rail line - this image only shows the southern half of downtown. Not even the most optimistic people are thinking anybody would walk to work at UT or the Capitol from this thing.

I've also put green dots on the biggest buildings in this area from emporis.com's list of Austin high-rises (top 20 only), and yellow dots on other future big buildings / employment centers in the area (mostly residential high-rises under construction). Note the complete lack of any current or proposed big buildings anywhere near this commuter rail stop.

July 18, 2006

Local puff media fails again

Apart from KUT, nobody bothered to get a remotely critical reading on Capital Metro's latest PR blitz other than Jim Skaggs' Neanderthal Act. And even KUT let Cap Metro off the hook, as it turns out. (Note that the Cap Metro flack responding to my comment that the shuttle bus or eventual streetcar would not be reliable or fast since it's sharing a lane with cars said that the commuter rail train would take the same amount of time every day - which is true - good dodge, CM flack; I salute you).

There's really only one question you need ask Capital Metro:

How are passengers on the train going to get from the train station to their office in the morning, and how are they going to get back to the train station in the evening?

The rail line doesn't even go close enough to downtown offices for people to walk; and there's zero chance anybody's going to walk the mile or two to UT or the Capitol. So, again, why is nobody asking Capital Metro how they're going to get to work in more detail?

February 09, 2006

The thing people aren't getting about the library

With the call to build it somewhere pretty or where they can build it bigger is:

The people who most need and use the library currently are quite likely to get there on the bus. Yes, the bus you think nobody uses; although if you stand outside the current library and look at those buses go by, you'll quickly be disabused of that particular brand of suburban idiocy.

The current library works well because it's on one of the two most heavily bus-travelled corridors downtown (Guadalupe). A location on Cesar Chavez too far from Congress, on the other hand, won't be an easy trip for many of the current patrons.

Look at the map (zoom in on the lower-right inset). Notice how many buses go right next to the thing. Most of the rest of the buses are three blocks away on Congress. So, a huge chunk of routes don't require any walk at all, and most of the rest require a 3-block walk at most.

Now, consider the proposed new site at what's now the water treatment plant. Going by current routes, two come fairly close, but the big conglomeration coming down Guadalupe/Lavaca will be about two blocks away; and the Congress routes about five blocks away.

This doesn't sound like much to walk, and it wouldn't be for most of us. However, as somebody who hasn't been able to walk well for quite a while now and used to serve on a commission where we were often taking up issues important to those who are mobility-impaired, I have more appreciation than most for what a pain in the ass this is going to be. Oh, and don't forget, unlike most of the people involved with this decision, I've been to this library many times - and I can tell you that at any given time, a huge number, possibly even the majority of the patrons arrived on the bus, and a large fraction of those are either elderly or in wheelchairs or both. For THOSE people, two more blocks is a lot to ask.

Don't move somewhere which makes the library less accessible to those who need it most just for the sake of being pretty. Please say no to moving the central library off the main bus lines.

Update: Several commenters have commented along these lines (paraphrased, with my response):

"Isn't commuter rail going to a transit hub at Seaholm anyways?" - please do yourself a favor and read this category archive and start with this post, OK? Short summary: It ain't going to Seaholm for decades, if then. And Seaholm is still a couple-blocks'-walk from this site.

The buses will just be moved to go by the library - this isn't going to happen either, folks. Long-haul bus routes don't make two-block jogs just for the hell of it (people already complain about how supposedly indirect these things are). Each one of those bus routes might deliver a dozen passengers a day to the existing library - enough to make it a valuable part of the demand for the current route, but not enough to justify hauling a long, heavy, bus around a bunch of tight corners.

January 31, 2006

"Build it and they'll come" is no way to run a city

So the end-result of the Parlor problem appears to be that the neighborhood isn't going to budge on the parking variance, which means that another local business is in danger of going under unless the notoriously neighborhood-friendly Board of Adjustment suddenly becomes more responsible.

The end of the thread on the hydeparkaustin mailing list occurred when a member of the "Circle C in downtown Austin" party commented that a plan (in the works now for a long time and seemingly not close to fruition) to arrange for parking at the State Hospital (across Guadalupe) to be used for employees of businesses on Guadalupe would be the only way out of this mess.

I replied that it was unlikely that any customer or employee of those businesses would find it attractive to park at the state hospital, walk out to Guadalupe, wait a long time for the light at 41st and Guadalupe to change, walk very quickly across the street, and then and only then arrive at their destination (as compared to parking on a side street or Avenue A).

The person replied (and was supported by the moderator, who then ended the discussion with the attached unpublished rebuttal in hand) that "the boss can make the employee park whereever they say". This may be true in an abstract sense, I replied, but it's unlikely that any such boss would want to spend the energy enforcing a rule which prevented employees from parking in PUBLIC spaces such as on Avenue A, even if they did want to keep employees out of their own private lot.

This goes back to thinking of a type which is unfortunately prevalent here in Austin and among many other progressive cities - that being that people will do things that are good, as long as we provide opportunities to do them. IE, build it and they will come. What you build, given this thinking, doesn't have to be attractive compared to the pre-existing or forthcoming alternatives; its mere existence will suffice.

For instance, in this circumstance, they think that simply providing available parking in an inconvenient and unpleasant location will get people to park there who would otherwise park on neighborhood streets. Likewise, Capital Metro thinks simply providing any rail will get people to use it, even if the individual incentives are pretty awful, given the shuttle bus transfers.

I have a whole blog category analyzing 'use cases' which I think is a far more useful way to look at the problem. In this case, for instance, put yourself in the shoes of that potential parking consumer a few paragraphs back and remember that your boss probably (a) isn't going to be able to stop you from parking on Avenue A, and (b) probably couldn't catch you even if he tried.

But like with the naive pro-transit suckers that bought the MikeKrusee ScrewAustin Express, it's unlikely that it's possible to get through to these people. And so, the consequence is that another local business which probably would have improved Guadalupe as a place we actually want to be is thwarted. Good work, geniuses.

This is not to say that we should never build transit or highways. What it does mean is that somebody ought to spend at least a few minutes figuring out whether the thing you're going to expect people to use is actually attractive enough for them to choose to use it. By that metric, light rail in 2000 was a slam dunk, despite the lies spread by Skaggs and Daugherty. But in this parking case and with this commuter rail line, nobody seems to have bothered to put themselves in the shoes of the prospective user.

my sadly now never-to-be-published response (remember, this is to somebody who said "But the Heart Hospital doesn't let their employees park in their lot!" follows.


Those cases have some clear and obvious differences to the one
we're talking about here -- one being that the employees are being prohibited from parking in a private lot (which is still difficult to enforce, but at least defensible). You're asking that these business' employees not only refrain from parking in the business' lot (private) but ALSO from the public spaces on Guadalupe and the street space on Avenue A. And nobody's 'requiring' those state employees to park in Siberia - if they could find an open metered space somewhere else, for instance, they're free to take it. Likewise, the Heart Hospital can't force its employees to mark at the MHMR pool.

So it's easy to prohibit people from parking in a given private lot. Unless you're going to turn Avenue A into RPPP as part of this, though, they'd still park there instead of across Guadalupe. And any boss who tried to force them otherwise would probably be experiencing the fun world of employee turnover.

November 28, 2005

Use Cases Part Three: Reverse Commutes

In case you thought I'd never pick one which works well with commuter rail, we've got one (although light rail would have worked a little bit better).

Analyzing a couple of reverse commutes:

Case 1 is a young downtown resident (of one of the condo buildings now under construction, for instance) who works at IBM (which as the draft environmental impact assessment states, will be right next to one of the stations). Parking up at IBM is free, of course.

Most of the residential development downtown is on the west side of Congress (except for the Milago and the 555, which are within walking distance of the train station). This puts the majority of housing units within a 5 minute walk of the 2000 light rail line with a short shuttle bus ride for the commuter rail station; with the Milago and 555 being the opposite.

For a minor variation, my own commute when I was working at IBM was from my condominium in Clarksville, from which I could have ridden a bus to either rail station from a couple of bus options - add 10 more minutes for extra bus travel for those trips.

Numbers indicate "seats". IE, if the number gets up to 3, you had to ride in 3 vehicles to get there. T indicates transfers. W indicates wait. P indicates pedestrian trip.

Passenger TripCommuter RailLight Rail (2000)BusCar
Downtown condo to IBM For the majority: (P). Walk to shuttle bus stop.
(W). Wait for shuttle bus.
(1). Ride shuttle bus to rail station at Convention Center
(W). Short wait (we hope) for train
(2). Ride commuter rail (not stuck in traffic) to station near IBM
(P). Walk to office at IBM or Tivoli
Estimated time: 40-50 minutes (5 minute walk on each end; 5-15 minute range wait and ride on shuttle bus)
(P). Walk a few blocks to Guadalupe.
(W). Short wait for train
(1). Ride light rail train (not stuck in traffic) to station near IBM/Tivoli
(P). Walk to office.
Estimated time: 40 minutes (5 minute walk on each end).

(P). Walk to downtown bus stop for #174 express bus.
(W). Wait for bus.
(1). Bus ride to stop near IBM (far from Tivoli).
(P). Walk to office
Estimated time: 50-70 minutes (5 minute walk to bus stop; 5-10 minute wait for bus; 35-45 minute bus trip; 5-10 minute walk to office)

(1). Drive (stuck in traffic, but reverse commute is free-flowing in morning; quite bad in evening) to office
(W). Find parking in own parking garage
(P). Walk to office
Estimated time: 15-45 minutes

Unless you live in Milago or 555, this commutes would be better on light rail than on commuter rail, but the car still kicks both to the curb during the morning commute and probably always will. The afternoon is where this commute really gets competitive - this is the route I used to have to drive when I worked up north and lived in Clarksville, and it's not pretty. You can sometimes save a bit of time by using alternate routes, but it's never quick; the problem is that the express bus on Burnet isn't going to be quick or reliable either since it's stuck in stoplight and slow-speed traffic conditions. Rapid bus isn't an option for this commute (at least, not initially - the long-term buildout indicates a route up Burnet). Both commuter rail and light rail allow passengers to at least obtain a more reliable commute, and in some cases even a faster one.

Having lived this commute, I'd pick light rail and MAYBE commuter rail over the car - a comfortable transit ride which took on average 5 minutes longer but was reliable and allowed me to work or read would have been a big winner. The scary thing about the commuter rail trip would be (of course) the bus transfer (if your shuttle is running late due to traffic, you're on the next train ride 30 minutes later). Light rail would have run about every ten minutes during the peak hours; so the penalty for missing a train would not be as scary.

Either rail line could pick up a small number of passengers who match this travel pattern (small because most workers at the IBM-area complexes live in Round Rock and other north/northwest suburbs; only a handful live central). The other thing this travel pattern has going for it is that the car trip is only going to get worse; while both the light rail and commuter rail trip are unlikely to get much slower since neither one relies heavily on a bus component.


Case 2 is the same downtown resident but he now works at one of the tech businesses on the 183 corridor (let's not even talk about the apalling amount of office space on Loop 360).

I've worked in several offices along this corridor while living in central Austin, so I know the area very well. An interesting fact about the light and commuter rail plans is that despite claiming to be alternatives to the 183 corridor, neither one goes anywhere near a parallel line to US 183 until they approach Cedar Park from the east. This means that the predicted rerouting or elimination of the 183-corridor express buses is really going to hurt transit in this area.

Numbers indicate "seats". IE, if the number gets up to 3, you had to ride in 3 vehicles to get there. T indicates transfers. W indicates wait. P indicates pedestrian trip.

I'm picking the first office I had at S3 in 1998 - because it happens to be located directly across Jollyville from the Pavillion Park and Ride (I would take the express bus up many mornings and ride my bike home).

Passenger TripCommuter RailLight Rail (2000)BusCar
Downtown condo to 183-corridor For the majority: (P). Walk to shuttle bus stop.
(W). Wait for shuttle bus.
(1). Ride shuttle bus to train station at Convention Center
(W). Short wait (we hope!) for train
(2). Ride commuter rail (not stuck in traffic) to station near IBM or station at Howard Lane
(W). Wait for transfer bus (no high-frequency circulator in either of these areas).
(3). Ride transfer bus to 183-corridor stop (stuck in traffic and slow)
Estimated time: 45 to 85 minutes (5 minute walk on each end; 30-35 minute train trip; 10-45 minute range wait and ride on bus)
(P). Walk a few blocks to Guadalupe.
(W). Wait for train
(1). Ride light rail (not stuck in traffic) to station near IBM or station at Howard Lane
(W). Wait for transfer bus (no high-frequency circulator in either of these areas).
(2). Ride transfer bus to 183-corridor stop (stuck in traffic and slow)
Estimated time: 45 to 85 minutes (5 minute walk on each end; 30-35 minute train trip; 10-45 minute range wait and ride on bus)
(P). Walk to downtown bus stop for 983 express bus.
(W). Wait for bus.
(1). Bus ride to stop near IBM (far from Tivoli).
(P). Walk to office
Estimated time: 50-70 minutes (5 minute walk to bus stop; 5-10 minute wait for bus; 35-45 minute bus trip; 5-10 minute walk to office)

(1). Drive (stuck in traffic, but reverse commute is free-flowing in morning; quite bad in evening) to office
(W). Find parking in own parking lot/garage
(P). Walk to office
Estimated time: 15-45 minutes

Unfortunately, neither light rail nor commuter rail is going to work for this trip, even if you brought your bike along and wanted to ride from the station to your office. (There are no good bike routes from either the prospective Howard Lane-area station or the IBM-area to the Jollyville corridor). Express buses today aren't horrible (you'll spend a good deal more time in the morning and be nearly competitive in the afternoon), but might be going away as part of this rail plan. Clearly neither rail line would gain a non-trivial number of passengers falling into this travel pattern.

November 20, 2005

Use Cases Part Two: Central Austin to Central Destinations

This use case analyzes a typical central Austin resident.

Let's consider a lawyer who lives in one of those expensive houses in Hyde Park and wants to get to his law office downtown. Mister Law-Talkin'-Guy probably has free parking available in his office building, but many downtown workers don't (they would have to pay to park). Today, Mister LTG doesn't take the bus, because it's a lot slower than his car, and he can park for free in his building.

Numbers indicate "seats". IE, if the number gets up to 3, you had to ride in 3 vehicles to get there. T indicates transfers. W indicates wait. P indicates pedestrian trip.

Passenger TripCommuter RailLight Rail (2000)BusCar
Hyde Park to Downtown Office Building (6th/Congress) (P). Walk to bus stop.
(W). Wait for bus
(1). Take normal city bus (new route) to commuter rail station out in east Austin or north on Lamar.
(W). Wait for train.
(2). Ride commuter rail to Convention Center station (not stuck in traffic).
(W). Hopefully shuttle bus is waiting for you (short wait).
(3). Ride shuttle bus "circulator" (stuck in traffic) to 4th/Congress
(P). Walk 2 blocks to office
Estimated time: 35-50 minutes

(P). Walk a few blocks to Guadalupe.
(W). Wait for train
(1). Ride light rail train (not stuck in traffic) to 6th/Congress
(P). Short (sub-block) walk to office
Estimated time: 15 minutes

(P). Walk to Speedway (for #5), Duval (for #7), or Guadalupe (for #1, #101, or Rapid).
(W). Wait for bus
(1). Ride bus (stuck in traffic - yes, even the Rapid Bus is stuck in traffic) to 6th/Congress
(P). Short (sub-block) walk to office
Estimated time: 25-40 minutes

(1). Drive (stuck in traffic) to downtown
(W). Find parking in own parking garage
(P). Walk to office
Estimated time: 10-20 minutes

To me, the only transit option which seems remotely palatable to Mr. LTG is the light-rail trip, because it could save time over his drive through rush-hour traffic. None of the other options are likely to be remotely competitive in time or reliability - in fact, the light rail trip might be a BIT slower than his car too. But if you're a downtown worker who has to pay to park, or parks a few blocks away from your office, the light-rail option would be a clear winner. The light rail trip might even win Mr. LTG over since he'd have a smooth comfortable ride where he could read the Wall Street Journal, which of course he can't do when he's driving, and probably not on the bus, unless he's unusually carsickness-resistant.

Note how unreliable the trips are which involve navigating traffic. On a good day, the car would beat even the light rail trip; but on a bad day, light rail would be faster. Light rail's speed doesn't change, in other words, because it has its own lane. The bus and the shuttle-bus both suffer from this worse than even the private car does, since you can always change your route when you're driving.

This particular passenger type maps well to UT students who live at the Triangle, or to UT staffers who live anywhere central, etc. Essentially, the entire central Austin residential market could have been very well-served by light rail, but will not be served at ALL by commuter rail.

Most people in Central Austin are transit-positive. That is, even if they own a car, they're willing to seriously consider using public transportation. A good number of these folks take city buses today; but the idea that Rapid Bus is going to get a non-trivial number of the remainder to leave their cars at home is ridiculous.


What about streetcars? The Future Connections Study, as I previously noted, has settled on a route which winds from downtown up to UT, then east to Mueller, so it won't be of much use for actual residents of Central Austin. Even if it DID go "straight up the gut" as intelligent folks asked for, it wouldn't be able to beat the city bus (or Rapid Bus) - unlike light rail vehicles, streetcars share lanes with cars.

Use cases Part One: From Leander / Northwest

Start of a new series - for those who are still optimistic about this commuter rail line. A "use case" in my business (software) describes how a customer might perform a certain task using your product - in this case, we'll describe how a few prospective transit customers would get to work using 4 transportation products.

Today's example is a Leander resident who works at the University of Texas or the State Capitol. Both locations don't provide much in the way of free convenient parking, so workers at both locations currently provide a good deal of business for the 183-corridor express buses. Leander residents are much more suburban and conservative than Central Austin residents, so the performance and reliability gap between transit and the car would need to be smaller, in my opinion, to attract new riders to choose transit than it would be for the analogous central Austinite. I expect most of those who are motivated by expensive or inconvenient parking are already taking those express buses, in other words. (and the express buses are actually pretty nice; most of the time I can read in them without getting carsick).

Numbers indicate "seats". IE, if the number gets up to 3, you had to ride in 3 vehicles to get there. T indicates transfers. W indicates wait. P indicates pedestrian trip.

"Current" is indicated next to the bus trip because there are some indications that Capital Metro might eliminate some of the 183-corridor express buses in order to induce more commuter rail ridership.

Note that the "shuttle bus" portion of this trip will, even if made on a streetcar, still have the same traffic characteristics (i.e. a streetcar running in mixed traffic will still be as slow and unreliable as a shuttle bus).

See notes after the table for more.

Passenger TripCommuter RailLight Rail (2000)Bus (current)Car
Leander to the University of Texas (1). Drive to Leander park-and-ride.
(W). Wait for train.
(2). Ride commuter rail to MLK station (not stuck in traffic).
(W). Hopefully shuttle bus is waiting for you (short wait).
(3). Ride shuttle bus (stuck in traffic) to UT
(P). Walk to office
Estimated time: 1 hour, 25 minutes to 1 hour, 45 minutes

(1). Drive to Leander park-and-ride.
(W). Wait for train.
(2). Ride light rail all the way to UT (not stuck in traffic).
(P). Short walk to office
Estimated time: 1 hour

(1). Drive to Leander park-and-ride.
(W). Wait for bus.
(2). Ride express bus (stuck in traffic) to UT
(P). Short walk to office
Estimated time: 1 hour, 15 minutes to 1 hour, 45 minutes

(1). Drive (stuck in traffic) to UT area
(W). Find parking
(P). Potentially long walk to office
Estimated time: 40 minutes to 1 hour, 5 minutes

Leander to the state Capitol (1). Drive to Leander park-and-ride.
(W). Wait for train.
(2). Ride commuter rail to MLK station (not stuck in traffic).
(W). Hopefully shuttle bus is waiting for you (short wait).
(3). Ride shuttle bus (stuck in traffic) to UT
(P). Walk to office
Estimated time: 1 hour, 35 minutes to 1 hour, 55 minutes

(1). Drive to Leander park-and-ride.
(W). Wait for train.
(2). Ride light rail all the way to UT (not stuck in traffic).
(P). Short walk to office
Estimated time: 1 hour, 5 minutes

(1). Drive to Leander park-and-ride.
(W). Wait for bus.
(2). Ride express bus (stuck in traffic) to UT
(P). Short walk to office
Estimated time: 1 hour, 20 minutes to 1 hour, 50 minutes

(1). Drive (stuck in traffic) to UT area
(W). Find parking
(P). Potentially long walk to office
Estimated time: 45 minutes to 1 hour, 10 minutes


In general, I assumed you would get to the express bus stop and wait 5-10 minutes for the express bus, and I was charitably assuming it would be on time. The remainder of that trip is from the 7:25 route in from Leander, and assuming a 5 minute or less walk from the stop. The drive is me estimating what I suppose it would take that time of day (I'd like to hear from a Leander resident that makes this trip in their car for a more accurate estimate). The commuter rail time has such a wide swing because of the shuttle bus component - buses fare worse than cars in heavy traffic due to their acceleration characteristics and the fact that they can't change their route to get around heavy traffic. In general, I assume that the more time you spend on a bus, the less reliable your trip (could be faster or slower than the average). (The express buses don't try to slow down to avoid hitting stops early on the way in in the mornings, unlike city buses, so you actually could get dropped off earlier than schedule indicates).

Note that one of the key attractions to the 2000 light rail route is its reliability. A route which doesn't require that you take shuttle buses can dependably get you to work at the same time every day. The train isn't stuck in traffic, and you don't have to make any transfers.