I was reading a rather famous attack on the notion of transit, and wanted to set the record straight on a couple of things about transit.
I’m writing this on an MD-88 over southern Virginia between IAD and ATL. The clouds are rushing by and I feel perfectly safe. Most air mishaps happen within 1,000 feet of the ground.
This flight is full, like so many in the renaissance of passenger air in the US. Airlines have gotten very good at selling the number of seats we need, and no more. This airplane is as efficient as it could be, per passenger mile. It would certainly require more energy to move me from Washington to Atlanta in traffic. It may even be more energy efficient that Amtrak, as trains are not usually this full.
Up in the air, over North Carolina, we have no land use impact. The great thing about t flying is that it’s done in plentiful 3D space. With a 5-mile horizontal and 1-mile vertical separation requirement, crowding of airways between airports is not a problem in most of the world. There are two places on this trip that require a great deal of land, however: The two airports, IAD and ATL. Two of the largest airports in the US, by site area. The runways alone involve acres* of the thickest, stiffest pavement you can get. It is richly important that the runway be smooth when landing this 80-ton jet. The highway circulation and parking around each airport requires additional square miles of land at each airport**.
Even though I’ll spend most of this trip out of anybody’s way, this trip does have a land use needs. Big ones. This flight could not have occurred without all that space being used on the ground at my origin and my destination. Our more everyday trips, like driving, taking transit, biking or walking also require space and resources at the beginning, end, and along the way.
This is one reason that transportation is about land use.
Every day, I get in a car in the parking lot near my home and drive along lanes that are twice as wide as my car to other parking lots near my cafes, work, dry cleaners, stores, or school. If I want to go to the area’s best parks and hiking trials, I drive to their parking lots. My car only takes up 0.7% of an acre at each parking lot, but the places that I go don’t just serve me, they are available to the maximum possible load of possible cars that could ever want to park in each lot. That’s why we have 800 million parking spaces for 300 million cars. Property owners build parking lots for dozens, hundreds or even thousands of people with the assumption that everyone they want to see in those places has to arrive by traffic.
And they’d be right.
There are a couple of good restaurants and a drug store within walking distance of home, and three cafes within biking distance. I can walk out my front door and walk to their front door, within 5 minutes. To bike, I unlock my bike garage in our back yard, haul the bike out, and drive it to the café. A bannister usually suffices for a bike rack when I get there.
The problem is that the walk is pretty barren, and the bike drive is downright dangerous, So more often than not, I just drive. The value of real estate and infrastructure given to me to allow this cheap seat in traffic may be greater than the cost of gas I put in the car.
The landscape around me is built for traffic, and is minimally usable for walking or biking. Walking or biking are an afterthought and an inconvenience to the free flow of traffic.
Speaking of inconvenience, consider transit. This is the only mode where I can’t just get going, but have to wait for a bus or train to pick me up at a designated stop and take me to another stop. The land needed for this is negotiable, including a bus pullout off the side of the road, a train station with bridges to pavilions in different neighborhoods, an elevator to an underground platform, or a pole on the side of the road with the transit agency’s logo on it. There is no inherent need for space at a transit stop, but sometimes consuming a lot of space makes the transit more visible, as with elevated stations, or usable, as with bus pullouts.
Just as there is no inherent size for a transit stop, there is no inherent need for a transit rider to get in traffic after getting off transit. A transit rider can walk or bike to where they are going, provided what they need is within walking or biking distance of their transit stop. I haven’t got the data to hand, but I hypothesize that most transit riders are traveling to a destination within walking distance of their destination stop. The impact of each transit stop could be magnified with an easy model for biking to transit stops. The 15 minute walk distance is a quarter mile, the 15 minute bike distance is almost 3 miles.
What’s within a quarter mile or three miles of most transit stations is a place built for traffic: signalized intersections, crosswalks that allow occasional access, roads wide enough to carry peak traffic at a tolerable pace and traffic for the rest of the day at a dangerous pace, and parking lots required of every property to keep cars off the roadways when not in motion. Traffic doesn’t work without routine impacts to land use that have been required for the last 80 years. Parking spaces everything apart to unwalkable distances, high speed traffic makes biking the realm of adrenaline monkeys and daredevils, and the scale of a landscape built for traffic is unusable for transit routes.
With so few people able to walk to each transit stop, there is little reason to run buses very often to much of America. An empty bus is more of a gas guzzling, carbon belching monstrosity than even the International CXT
Yes, they weren’t kidding.
So what would it mean for transit to affect land use?
More jobs, housing, shopping and parks (not parking) within walking and biking distance of each stop or station. Many transit stops are completely surrounded by drivable suburbia. A bus stop on the edge of a parking lot next to the highway is a frequent and futile site for all but the people who work in that strip mall or live in those garden apartments.
What would it mean for biking or walking to affect land use?
More jobs, housing, shopping, and parks within biking or walking distance of where they live or work, and along streets that are safe and interesting to bike or walk along. Interesting because they are full of places to bike or walk to, not spaced apart by mandatory parking lots between the street and the buildings. People in historic city centers and some New Urbanist developments get this at a price premium. It is more expensive because it is
A) better and
We don’t build a lot of walkable/bikable places in the US because they violate the default. They have no place where they must be, the way a parking lot must be next to every development that wants a driveway.
A good place for walkable and bikable places would be next to transit stops.
* One reason this article was delayed was that I was trying to get a runways, taxiways and parking lots polygon layer into GIS. OpenStreetMap is remarkably coy on this front. If I had such areas, I would divide that area by the annual passenger traffic and divide that by the number of working hours in the year to get an approximate space consumption per passenger. Assuming a passenger uses an airport for an hour. I’ll add this to the list of lampposts I’d like to look under for data
** Many air passengers have to drive their vehicles to the airport and leave them there while on their trips. My own car was just one of the thousands parked at IAD this weekend.
This is a follow-on from an article where I meticulously compared various modes of transit in cost and energy use. I promised I’d get that done for a time series, and I’m ready to write about that.
First, Cost. The Cost per transit passenger trip is pretty simple to calculate, as we measure transit use in passenger trips, not miles. A passenger trip is far more useful than a vehicle mile, as a trip represents someone getting where they want, rather than just moving miles for the sake of moving miles.
I chose the maximum for this graph as the average cost of traffic trips in the US (~$2.06). We travel about half a light year in traffic every year, and the cost of owning and operating vehicles on our roads far exceeds the cost to government actually build and maintain the roads.
A lot of transit systems are more expensive per trip than traffic, which is why they are not shown here. There are no commuter rail modes (blue) shown on this graph, and Light Rail (brown) and Heavy Rail (green) are clustered here around the $1.50 price. The general upward trend in prices may be inflation. These values are not adjusted for inflation, but are a collection of annual reports strung together into a shambling wreck of data.
What you don’t see are the highest cost lines. The rouges gallery includes Connecticut Commuter rail, Dallas Area Rapid Transit Light Rail, and San Francisco BART. These agencies have been spending over $25 per passenger trip, which makes financing difficult at best.
With the same color code, as above, and repeating the same look at BTUs per-trip as a measure of energy. The colors are same.
The highest consumers of energy per trip are New Jersey Transit (Commuter Rail), New York Subway (Heavy Rail), and MBTA (Boston) (Light Rail). Each of these systems averages over 5x the per passenger energy consumption as a car in traffic. I’m pretty surprised by these number, considering that New York should be able to use economies of scale to improve their energy efficiency. They are also one of the most affordable systems, per person. NJ Transit is one of the largest Commuter Rail Systems, and MBTA operates the oldest, largest light rail system on the East Coast. It could be that NJ Transit’s great schedule service (short headways through much of the clock, over many miles of track) means that many of their trains run nowhere near capacity. A commuter rail train weighs dozens of tons, and even on the best rails it takes considerable energy to move the thing. It is not energy efficient to move transit vehicle when they are empty, but the schedule must be obeyed.
This is one reason schedules should be carefully balanced between potential and real customer service, and land uses should be concentrated around transit stations to ensure round-the-clock transit demands in both directions. If a large transit vehicle is going to be serving a place anyway, it is a waste to have it run empty.
One last thing: It is worth exploring what systems have high energy efficiency, if the high energy users are so surprising. Oddly, NJ Transit ranks well on energy efficiency for its light rail service. If I recall, that is just The Hudson-Bergen Line and the River Line, but there could be more. Northern Indiana, the one-time leader in interurban service, now has the most efficient commuter service. Finally, Staten Island Transit, also within New York City, has the most energy efficient heavy rail service. Curiouser and curiouser.
The energy use of electrically powered modes is not the same thing as the energy put into making the electricity, however. Though my earlier findings differed, the efficiency of burning coal, gas, oil, or protons to boil water to turn a turbine which turns a dynamo is listed in the transit literature as 40%. That is, we consume 60% of our energy on electricity just operating the machine to make the electricity, with only 40% of the energy output as useful electricity.
More on that later. Still wrangling my spreadsheet.
TS2.1TimeSeriesOpExpSvcMode TOS 2.xlsx
I have long admired the trip and distance data collected sporadically by the National Household Travels Survey. Not because of its fine scale (it is national only) or because of its annual updates (irregularly between 5 and 10 years), or even its consistent methodology (It, like the census, has shown a nasty drift in survey methods and assumptions between surveys).
What I admire about the NHTS is that it surveys all the modes (traffic, transit, walking, and biking, along with boat, and air trips.. That is not to say it surveys them with perfect accuracy. It surveys people after the fact, who may or may not exaggerate or misremember their travel use in even the recent past. That is also not to say that it achieves the grail of transportation surveys: the “tour”. Many of us walk to our cars before parking them near our destination where we might walk some more to get to our destination. Transit users always have to get to their stop or station, be it walking, biking, traffic, or a different mode of transit. Only the luckiest Brooklyn Heights commuter walks right from their front door to the ferry and thence to their work on Wall Street. They might need a CitiBike to complete their trip.
MY final frustration with the NHTS has been its evolving acceptance of Walking and Biking as modes. Though they do a good job of delivering 1995, 2001, 2009, they don’t deliver the 1989 or 1983 results in anywhere near such a convenient form. Which is just as well, because their attitude towards walking and biking have been evolving from invisible to minorities.
But then it occurred to me to look at the reports to see if they had any information on biking and walking before 1995. It took some hunting, but I was able to get some numbers of trips by walking and biking among all the less useful summary tables. Assuming that the average walk distance was 0.5 miles and the average bike distance was 1.3 miles, I was able to estimate the total miles traveled for these unsung modes. I based those estimates on the the trip distances from the three NHTS dates I did have trip numbers and distances for: 1995, 2001, and 2009.
So here’s the longer term trends in miles per capita for not only traffic and transit, but also biking and walking, expressed in miles traveled per capita.
Notice the downturn in traffic miles traveled down to 1990 levels, the uptick in transit use since 1996, and the climb of biking and even walking from their lows in 1991. 1991, by the way, was the first year that federal transportation policy shifted to acknowledge walking and biking as anything but a local concern.
I was putting together a presentation and needed an image of a rail transit parking garage. So I went to what I know.
The North Parking garage at the Vienna/Fairfax Station at the end of WMATA’s Orange line is pleasingly rectangular, and most usably, closest to the walk bridge across I-66. Bing Maps has long provided a nice feature: oblique maps form four cardinal directions. These are useful for understanding the massing and facade of buildings better than mere overhead or street view shots.
So, here it is.
This 9 AM summer shot is hilarious.
The 2,000 space garage is obviously full. Beyond full, as we can see from the couple of crd packed in the striped areas at the corners of the parking garage. As this is a garbage off a traffic exit for mode switching, the number of transit passengers deliver daily by this garage is not much above 2,000, or the capacity of two Metro trains. I don’t know how many families can agree on a commute into the city, a time to do it and a return time back, but I suspect it’s not that many.
The two cars circling the lanes are not going to find a space, as we can see. There are probably two more like them on the levels below, spiraling hopelessly to the top. Note the SUV at the lower left entering the garage, in the hopes there ill be a space for its berth. There will not, as we can see.
However, they will find is that a certain block of reserved spaces opens at 10 AM, and can hold a few dozen desperate cars long after their work shifts began, in a pinch. Its a clever overflow, but the fact is that most of these cars are just being stored for the day. Once the garage is full, it is full. The only way transit passenger can reach the station after that is to arrive by bus, be dropped off, bike, or walk to the station. Pity the station is surrounded by full parking spaces and single family homes.
As we got set for the weekend, I gazed out over a sobering sight to remind me that Transit Oriented Development (TOD) is not always simple to achieve.
We were waiting to board the bus for New York for a weekend with our niece and her girlfriend. My wife looked out over the scrim of new condos at the edge of a bare field, and expressed her disgust at the “ticky tacky” new developments. I told her that this was the MetroWest development, which was an exciting redevelopment of a formerly single family complex right next to the the metro station.
This metro station at the end of one of the busiest lines in DC’s system had long been designed as though its sold purpose was to lure drivers off the interstate and have them take the last 20 miles of their ride on the train. Once 5,000 cars filled up the decks, then the function of the metro station was essentially done for the day, and it would run trains at wide intervals to serve the few stragglers who arrived by bus or *gasp* foot. The number on foot would be very low, as the station was situated in the middle of a highway next an offramp. The housing developments that surrounded the station had nothing to do with it when it was built around 1984.
Over the years, subdivisions came in that were higher density in recognition of the rail line’s capacity and connection to jobs downtown. But there were nothing but housing developments here and there. They were “Texas Doughnuts”, with parking surrounded by 3-4 stories of housing to allow every apartment or condo 2 parking spaces. They were built dense for transit, but they still had to allow lit e in the suburbs. Resident used transit only for the commute, but for everything else they needed a car in traffic. The station was built in the middle of an interstate next to an offramp.
We lived in one of these Transit Adjacent Developments (TAD) for 2 years during my first job out of school. I used to walk through the metro station to get to work. But there was nothing but housing.roads and parking.
This made MetroWest exciting, as it was designed a s a complete suite of hosing , jobs and retail to satisfy our needs for walkable stimulation and service. I remember thrilling to this during an interview with AECOM (nee EDAW) and looking forward to a Fairfax County that finally got it.
But now they’ve built all the residences they can stand in MetroWest, and the Commercial and Retail promise is a grass stormwater promise awaiting development. The residents walk across the field to get to the station.
The problem is that the commercial real estate market is not what it needs to be. One sure way for a developer to go bankrupt is to build a multi-million dollar office tower that sits as vacant as downtown Detroit for an indefinite period. They need tenants for this to happen. Without the commercial, there is no structure and no market of the retail, mostly restaurants for the office lunch crowd and resident dinner crowd.
Office demand has not gone down a great deal in the DC area, but it has gotten much more compact. It was an embarrassing adage ten years ago that offices devote as much space to parking each worker’s car as they do to the worker’s office. Now that ratio is down to twice as much space for each workers car. An office simply devotes much less space to each worker than it use d to. This is great for businesses that lease, but a catastrophe for those that build the buildings and that they would actually want to lease int. And of course a problem for those of us who took commercial office development as a granted and important piece of every thriving TOD.
12 years ago, when I was just learning about the filed, a keynote speaker proclaimed that commercial office was the lynchpin of TOD. Without it the TOD will fail. Well now what?
So MetroWest will have to wait for a critical mass of successful tenants to come occupy the space. And they will have to of this in competition with all the other office buildings in the DC area.It a great place for an office retail compile,x but there has to be the real estate market to support it. Full Stop.
MetroWest is a great wiring that if you draw it, they may not come. At least I can be confident that these real estate forces are not particular to TOD, but affect traffic dependent and walkable developments the same.
* BTW, I wrote this in a bus going down I-95 and the New Jersey Turnpike. It added a certain piquancy to my ordinary typographic style.
Lingering on the patterns of transit reported in the National Transit Database a bit further, I thought I’d look at the modes differently than last week, in their patterns and their top performers.
These are the distributions of performance by decile. Dividing their average performance by their maximum value, and binning their distribution into 10 bin s provides a rough histogram of their distribution. Too often I use averages to show data because that’s all the data I can find, but these graphs show that rare world data rarely looks like the way we assume data should group about an average. I smoothed the lines that were the most “normal” in their distributions, and left the other ones jagged. I also only considered those with enough data to provide a regular distribution, throwing out those with less than ten examples in the US.
The first decile on the left shows how what percent of agencies operating each mode were less than 10% as efficient as the maximum. The middle of the graph shows what percent are half as efficient as the maximum.
As promised, I also want to praise the top performers in some of the transit modes:
- Best performer overall Trips/$ : City of Portland Aerial Tramway, even better than biking
- Best Funicular Trips/BTU : Chattanooga Area Transit
- Best Heavy Rail in trips/ BTU and $ : MTA New York City Transit
- Best Light Rail in both Trips/BTU and $ : MTA Harris County, TX Houston
- Best Commuter Rail Trips/BTU : SEPTA Philadelphia Transit
- Best Commuter Rail in Trips/$ : Peninsula corridor joint powers board, San Francisco/San Jose, CA
- Bus best Trips/$ : UGA Transit system, Athens, GA
- Bus best in Trips/BTU : Ames, IA Transit agency (notably both college towns)
- Best Bus Rapid Transit in Trips/$ Regional Transportation Commission, Southern Nevada, Las Vegas
- Best Bus Rapid Transit in Trips/BTU Los Angeles MTA
For a much later post, I’d like to compare the performance of these modes from a decade ago, to see if they have consistently been this good. I’ll want to look at changes in these pattern over time, and other stats, like route extent, vehicles in service, vehicle miles, and firebox recovery, but a bit later. For now, look at the performance patterns from last week, Trips/BTU and Trips/Operational $.
By Friday, I’ll talk a bit more about population and jobs patterns, I’ve got a busy weekend coming up and that will be good practice for it.
So, how does transit stack up against other modes, really? I’ve approached this question in earlier posts, but I decided I’d look at it a little differently this time.
It is relatively easy to get agency data on trip numbers, energy, operating costs, miles traveled, route miles installed, and so forth. Each transit agency reports this data to the FTA as a condition of receiving federal aid, no matter what mode they operate, A lot of transit agencies receive federal aid, and a lot of them dutifully report this data. 536, in fact, from Surprise Dial-A-Ride to Chicago Transit Authority.
I figured it would be more interesting to look at the data as it was, rather than trying to calculate some dull average. That way, the bus service in Akron would be represented along with that in New York, and we could see what the range of performances was.
I picked the first two of those, energy, and operating costs, tabulated for each transit agency. Some cities might only operate a metrobus and van pool, while others could operate everything from ferries, to subways, to aerial tramways. This data let’s us compare the performance of 16 different “modes” of transit, vastly different in way and propulsion*.
They are all similar in their primary purpose, however: trips. Getting people from where they are to where they want to be is a function of any transportation mode, and is a far better measure of transportation performance than distance traveled or time spent traveling.
So how do the various modes of transit stack up against the average energy and operations costs of traffic, biking and walking?
Biking and walking are not shown here or on the next graph because they perform much better than transit or traffic. on energy (0.18 and 0.25 Trips/BTU for biking and walking, or 150x the scale shown above) and price (8.78 and 2.34 Trips/operating $, or 2x the scale shown below. Then again, they don’t move as fast, and require us to be “outside”. How quaint.
So, we see that Trips/ BTU for most transit modes is higher than the average for traffic, the green line across the bottom of this graph. The mode that is closest to traffic in energy inefficiency is also the most like traffic: VanPool
In the case of operating cost, transit does not stack up nearly as well as traffic. Only Inclined Plane (IP), Bus Rapid Transit (RB) and Aerial Tramway (TR) offer more rides per dollar. There is one aerial tramway reported in the US**, though I can think of at least two, and many more if skiing is your goal.
Of course, transit agencies typically have to operate the ways as well as the vehicles, while traffic has an installed base of over 8 million lane miles to drive across. Those are the numbers, however, and I’d like to explore them further in the coming months.
I’ll get into the star performers in each of the modes in a later post, perhaps as early as next week.
By the way, the modes are:
CB COMMUTER BUS
CC CABLE CAR
CR COMMUTER RAIL
DR DEMAND RESPONSE
HR HEAVY RAIL
IP INCLINED PLANE
LR LIGHT RAIL
MG AERIAL GUIDEWAY
RB BUS RAPID TRANSIT
TR AERIAL TRAMWAY
YR HYBRID RAIL
** And it’s not the one I’ve ridden, it so happens.
My favorite bus these days is the Fairfax connector 926.
On the way home, it has an express route that zips past all the rigamarole in Herndon, getting me right to the new Wiehle Avenue Metro station. On the way out, this same expedited route means that it sits at the station for minutes before taking off westbound. I haven;t gotten off the train yet when it hasn’t been there.
I worked out a clever little schedule of the 926 and the 952, the two buses that deign to pass within a suburban block of my work. They have roughly the same headway (a a suburban 30 minutes), but the 952 just doesn’t measure up to the 926 in getting me expeditiously to and from the station. The 952 is never at the station when I get there in the morning.
There was one advantage to the 952. I was able to get a lot of work done on it one evening,as it wound its way through five miles of rush hour traffic in the exurbs or Herndon and Reston. Time on transit is different than time in traffic. I am drafting this article from a seat on that bus; something I could not do if I was stuck in traffic.
The thing that inspired me to write this piece in particular, was the notion of dwell time. How long is your bus going to wait at your stop. This is an advantage afforded transit users at the ends of routes. Along the middle of the route, the train or bus comes and goes in a matter of seconds. If there is no one at the stop the bus will just drive on by. Only at the very busiest stop will the bus wait through a cycle of the nearby traffic signals for passengers to make it across the traffic arterial without being killed by blunt force trauma. Once the bus is gone, it is gone, which is a big deal for a bus that only comes every 30 minutes. At the end of the route, you can take your time.
What if a bus knew what passengers it was going to take on each run at each time, and wanted faithfully at each stop?
It would be great customer service for those trying to get on the bus, but infuriating for those already on the bus. It would also be infuriating for the vehicles behind that bus in traffic. Forced to make their way around the stopped bus until every single sainted passenger finally made it to the front door. Which is why we don’t run transit that way. If you miss the bus, you wait for the next one, or you get a cab.
But we do run traffic that way. Our cars, trucks and motorcycles sit patiently for us the majority of every day, machines in repose, waiting for the 5% of their existence that they are actually used as advertised. We require that they do this out of the roadway in most of America, which is why subdivision codes come with specific parking minimums that must be obeyed if the property is going to connect the road*.
Bikes and shoes wait for us on our time, too, but its hard to use those in an America built for the ecstasy of traffic.
* What a silly notion. Of course the property has to connect to the road. How else can we get to it?
I was getting off at a metro stop when it occurred to me that the modern rail transit station Is a petty good metaphor for America’s transportation choices.
Passengers arriving at a heavy rail station have to pass from the street level to the track level, often via a concourse. As these are each usually at separate levels, they have to move up or down to reach the transit platform and their desired ride. For systems built before the transit revival of the 1960s, this was unanimously via stairs, with some expensive retrofits for electric elevators and escalators. For the systems planned in the 1960s and built in the 1970s and 1980s, this usually included at least some elevators and escalators. After the 1992 Americans with Disabilities act, elevators became mandatory for all new stations, and strongly suggested for all of the old ones. Most of our heavy rail stations were built before the revival, and mechanized mouthed of getting to the track platform are an afterthought.
Stairs, the oldest way of moving up and down, invented with larger architecture 8,000 years ago, call for walkers, developed 6 million years ago. Many opt to take the stairs after the escalators are full, or when they are in the mood for a bit of exercise.
Escalators, invented in 1898, can be used while walking or standing. They deliver passengers to the same places as the stairs with the same or greater speed. They make exercise at the station optional. The important distinction of the escalator is that they are continuously moving. Passengers can use them whenever they want, and then relax for five seconds while they are delivered effortlessly to the next platform. Since they are continuously moving , they use the most energy of the three, much like automobiles in traffic.
Elevators, invented in 300 BC but perfected in 1852, offer vertical movement of a flat platform from one level to the next. The important thing about them is that you have to wait for them, and then move on their schedule. Much like transit.
This was just a thought I had, I figured I’d share it with you while it was still in my head.
I’ve been traveling and posting from Asia for about a month now, and its time for that to wind down. I’m going to post one more photolog of what we saw in Delhi, though the time in Delhi was far more “family event” and far less “sightseeing. We would not have made the trip if not to wish our uncle and aunt a happy 50th anniversary, but I am not going to blog about that.
This weekend, I’ll post those pictures from Delhi, and Monday I’ll post about the spatial aspects of Washington, Atlanta, Philadelphia, Istanbul, Mumbai, and Delhi.