Basic facts:
·
Average freeway lane maxes out at 1800 vehicles
per hour at full speed.
·
An unoptimized bus route will carry 45 people on
average.
·
An optimized bus route will carry 80 people.
·
A single BART railcar will carry 56 seated
passengers, and up to 200 passengers in total with standing capacity. BART
trains are between 3 and 10 cars long.
·
A single Link train carries 600-800 passengers.
So, an average freeway lane is anticipated to carry 1800
vehicles per hour per lane at peak times. Most vehicles are empty most of the
time.
Basic equivalency
|
Passenger
|
Freeway lane
|
Optimized Bus
|
Link
|
BART
|
Passenger
|
|
|
|
|
|
Freeway lane (per hour, 1 passenger = 1 car)
|
1,800
|
|
22.5
|
2.25
|
0.9
|
Optimized Bus
|
80
|
|
|
|
|
Link
|
800
|
|
10
|
|
|
BART
|
2,000
|
|
25
|
2.5
|
|
I-80 SF westbound per hour (5 lanes)
|
9,000
|
|
112.5
|
11.25
|
4.5
|
Katy Freeway eastbound per hour (7 lanes)
|
12,600
|
|
157.5
|
15.75
|
6.3
|
Bus optimization
If I am planning for a city of 100k people, and I have a bus
depot in downtown which has 10 buses which leave downtown every 15 minutes, I
have 40 vehicles per hour. Those vehicles only need to carry 45 people each on
average to equal the carrying capacity of a highway lane.
To replace a 3-lane freeway, I need 120 buses per hour
across my metropolitan area, or 30 vehicles every 15 minutes. That is honestly
not that many buses. 20 bus routes at 10-minute intervals are the equivalent of
a 3-lane freeway, in both directions.
If each bus is running a fully optimized route and carries
75 people per vehicle, and we have only 10 routes in our city, 25-minute
headways on these optimized routes will equal the capacity of one highway which
has 6 lanes each direction.
This however is complicated because many of the less efficient routes are feeder routes to the system which serve local neighborhoods. Cutting the local can significantly reduce overall passengers carried, reducing the viability of the entire system. Transit planers cannot just put in a few high capacity low operating cost vehicles and call it a day. The lack of a network effect will all but guarantee those routes will never get to that optimal capacity. Higher order more cost efficient transit effectively subsidizes the local routes, while local routes make it so that rider can get to that higher order transit from their front doors. They work together and in larger cities they need each other.
The I-80 bridge in San Francisco has 5 lanes and carries
about 9000 people per hour. It would take 120 buses to replace the I-80 bridge.
2 buses per minute per hour leaving San Francisco will carry an equivalent
number of people.
Rail optimization
Using Link rail cars
A single Link train will carry 700 passengers per hour. I-5
is a 4-lane freeway south of the I-405 interchange in Tukwila. I-5 carries 7200
people (single individual cars) per hour at rush hour in each direction. 10
trains per hour (6-minute headways) will carry as many people as I-5.
Using BART cars
A typical BART train on the highest demand routes carries
2000 passengers at peak hours. The I-80 bridge has 5 lanes, so it carries 9000
people per hour. It takes 5 trains per hour to beat the capacity of the I-80
bridge.
The fare for BART needs to be less than the toll to go over
the Bay Bridge.
Annual ridership
So… let’s say you take the I-80 bridge between Oakland and
San Francisco, which can carry 9000 people per hour each direction. So, you
have 250 workdays per year, and you have about 6 hours per day where people are
commuting, which means that if one side of the bridge is at capacity 6 hours
per day, 250 work days per year you have a total capacity of 13.5 million
passengers per year. BART carried 26 million passengers in 2021.
Even simpler, you can expect the I-80 bridge could
comfortably carry 45,000 people per day. For comparison the Embarcadero station
on BART carried 48,000 people per day by itself in 2017, including weekends.
Montgomery Station carries another 45,000 people per day by
itself.
If the I-80 bridge were running at full capacity 12 hours
per day it would carry 90,000 people.
These two stations, Embarcadero and Montgomery, carry 93,000
people per day.
Most of the passengers of BART in downtown San Francisco are
going to the East Bay.
The Embarcadero station by itself fills 24 trains per day.
Montgomery station by itself fills another 22.5 trains per
day.
Put another way:
If the I-80 bridge was working at full capacity 12 hours a day,
it could carry 108,000 passengers per day. BART carried 145,700 passengers per
day in Q3 2022.
And yet another way:
There are four BART lines which go under the Bay. Those are
the green, blue, yellow, and red lines. Each of them runs 4 trains per hour
under the Bay, so there are 16 trains per hour running through the Transbay
Tube, or one every 3 minutes and 45 seconds, in each direction.
These four lines can carry 32,000 passengers per hour.
The I-80 bridge only carries 9,000 passengers per hour.
The four BART lines which run through the Transbay Tube
generally carry 4 times the number of passengers as the I-80 bridge.
How does BART compare to buses?
If you were to try to build a bus network to carry as many
people as are carried by BART… you would need to run 400 buses over the
TransBay bridge in the morning to match the capacity of BART. That would be one
bus every 9 seconds!
Conclusion
For high-demand routes, buses are more expensive to operate,
slower, and carry fewer people.
BART is one of the best investments ever made. If you tried
to build a system which carries as many people as BART with light rail you will
end up with significant congestion, and you would have to build another 20
lanes of highway IN EACH DIRECTION across the Bay in order to match the number
of people BART carries every day.
The two tracks of BART underneath the Bay carry as many
people as 50 highway lanes. On a cost per passenger km basis the San Francisco
Bay tunnel is one of the best investments this country has ever made.
Whenever a politician or pundit is talking about how they
need to constrain costs, you need to think about the benefit that program will
provide just as much as you are talking about how much it will cost to build
and consider how much it will cost to maintain.
The TransBay Tube was far cheaper to build and is far
cheaper to maintain than any other comparable infrastructure which would carry
the same number of people as BART carries every day.
References:
https://www.numerade.com/ask/question/a-four-lane-highway-has-a-normal-capacity-of-1800-vehicles-per-hour-per-lane-in-the-southbound-direction-a-vehicle-disablement-on-the-roadway-shoulder-occurs-at-430-pm-due-to-rubbernecking-t-70091/
https://www.codot.gov/programs/innovativemobility/assets/commuterchoices/documents/trandir_transit.pdf
https://www.soundtransit.org/system-expansion/building-system/modes-service
Appendix A: Hierarchy of transportation modes
1.
Walking. Walking is the best mode of
transportation. Every trip starts with walking, so it is best to make it so as
many trips as possible can be done by only walking. Trips under 2 miles should
be done via walking.
2.
Biking. Biking is excellent because there is no
wait time, it’s the second lowest cost after walking, highly space efficient,
and very cheap to build for. For trips up to 5 miles biking should be
preferred, and possible to use bike infrastructure for trips up to 10 miles.
3.
Buses. Buses are excellent for short, low demand
routes. They are more expensive per passenger km than higher orders of transit,
but for low demand routes buses are ideal.
4.
Streetcars are ideal for short high demand
routes. They have the cost and capacity advantages of higher order rail over
buses while being cheaper to build and operate. Build these in your city center
on routes which have high usage. The San Francisco Streetcar is a good
implementation of this useful technology.
5.
Light rail is useful for medium distance, medium
demand routes. A perfect use case of this is the Hudson Bergen Light rail.
Routes which are too high demand for a bus or a streetcar, but not enough
demand to be worth building a tunnel for rapid transit are perfect candidates
for light rail.
6.
Rapid transit. Rapid transit is only ever
realistically going to be rail. Tracks are no more expensive to build than road
outside of legal barriers (which can be overcome), they carry a lot more
people, you can go faster on tracks than is possible to go on a road, and
trains have a lower operating cost than buses per passenger km. In big cities
rapid transit is realistically only going to be done either as elevated rail
and/or tunnels. If you don’t elevate or bury you will have more delays, it
stops being rapid, and you are wasting money. It is less expensive to do it
right way the first time.
a.
Short-haul rapid transit: This is rapid transit
for moving people around 20 km or less. The New York Subway or BART is a great
example of this.
b.
Long-haul rapid transit: This is for moving
people between urban centers, even within the same metro area. Deutsche Bahn
between Bonn and Koln is a great example of this.
Appendix B:
In response to the whole "America can't have good
transit because we are too spread out!!!" argument...
San Francisco has a population density of 2,642 people per
square km, and Los Angeles has a population density of 3,206 people per square
km.
Yet San Francisco is MUCH faster and easier to get around
without a car than Los Angeles, and Los Angeles is known for being a
car-centric city.
Creating real urbanism takes more than just building density
through dense zoning laws. Los Angeles proves that you can zone for density,
but if you don't provide convenient, affordable, fast transit, walkability, and
safe places to bike, you will still be stuck in car-centric hell.
Plus, you can build bikeable cities with good transit with a
lot lower density than you might think... to a certain point. Bellingham has
half the density of San Jose, but I find Bellingham is a much easier city to
bike around than San Jose.
It's a combination of walkability, available transit, dense
neighborhoods, and safe convenient ways to bike around your city, all working
together.
Density does not lead to a car-lite or car-free lifestyle
being reasonable by itself.
Appendix C: Global mode transport data
I grabbed mode share data from Wikipedia. It includes 146
cities from every continent across a large variety of countries.
First of all, let’s describe our data:
Most common mode of all:
private
motor vehicle 106
public
transport 30
cycling 5
walking 5
74 cities saw cars compose more than 50% of all trips.
United States 23
Canada 11
Spain 10
Italy 7
Australia 6
New Zealand 3
Netherlands 2
Ireland 2
United Kingdom 2
Indonesia 1
Greece 1
Germany 1
Malaysia 1
Sweden 1
How many cities per country are car dependent, meaning that
car trips compose 50% or more of all trips?
|
Country
|
Car
dependent cities
|
Total
Cities
|
0
|
United States
|
23.0
|
24
|
1
|
Canada
|
11.0
|
11
|
2
|
Spain
|
10.0
|
14
|
3
|
Italy
|
7.0
|
8
|
4
|
Australia
|
6.0
|
6
|
5
|
New Zealand
|
3.0
|
3
|
6
|
Netherlands
|
2.0
|
5
|
7
|
Ireland
|
2.0
|
2
|
8
|
United Kingdom
|
2.0
|
4
|
9
|
Indonesia
|
1.0
|
1
|
10
|
Greece
|
1.0
|
1
|
11
|
Germany
|
1.0
|
15
|
12
|
Malaysia
|
1.0
|
1
|
13
|
Sweden
|
1.0
|
3
|
14
|
Poland
|
0
|
5
|
15
|
Switzerland
|
0
|
3
|
16
|
Japan
|
0
|
3
|
17
|
China
|
0
|
3
|
18
|
Denmark
|
0
|
2
|
19
|
India
|
0
|
2
|
20
|
Czech Republic
|
0
|
2
|
21
|
Belgium
|
0
|
2
|
22
|
South Korea
|
0
|
2
|
23
|
Brazil
|
0
|
2
|
24
|
Austria
|
0
|
2
|
25
|
Lithuania
|
0
|
1
|
26
|
Columbia
|
0
|
1
|
27
|
Norway
|
0
|
1
|
28
|
Finland
|
0
|
1
|
29
|
Serbia
|
0
|
1
|
30
|
Singapore
|
0
|
1
|
31
|
Slovakia
|
0
|
1
|
32
|
Philippines
|
0
|
1
|
33
|
Romania
|
0
|
1
|
34
|
Bulgaria
|
0
|
1
|
35
|
Ethiopia
|
0
|
1
|
36
|
Chile
|
0
|
1
|
37
|
Belarus
|
0
|
1
|
38
|
Israel
|
0
|
1
|
39
|
Hungary
|
0
|
1
|
40
|
Portugal
|
0
|
1
|
41
|
Mexico
|
0
|
1
|
42
|
Bangladesh
|
0
|
1
|
43
|
France
|
0
|
1
|
44
|
Taiwan
|
0
|
1
|
Car dependent cities appear to be a regional phenomenon
concentrated both in Southern Europe and the Anglosphere.
If we exclude Anglosphere countries and Southern Europe (commonly
referred to as the PIGS), we find that Eindhoven, Rotterdam, Essen, and
Gothenburg are the only cities left which are car dependent and in highly
developed countries.
Every other city in Northern Europe (excluding the UK and
Ireland) has car usage below 50%.
What are the reasons for this?
Well, a commonly cited reason is downtown urban freeways.
But Munich, Paris, Cologne, Bonn, Tokyo, Mumbai, Hong Kong, Budapest, Prague,
Berlin, and more all have highways within 5 km of their downtowns. Almost every
city in the world has a freeway within 5 km of downtown. Urban freeways are not
a factor.
Could it be because the urban highways are tolled? Well…
there are practically no toll roads in Germany, which does extremely well on
this list with almost all their cities dominated by non-car usage, some even as
low as 20% car usage even with urban highways in all of their cities. It’s more
complicated.
It’s not the urban highways.
Of these cities which have less than 50% car usage… what is
their dominant mode of transport?
private motor vehicle 38
public transport 30
cycling 5
walking 5
About half of them still have private motor vehicle as the
dominant mode of transport, and all but 10 of the others have public transport
as their dominant mode of transport.
109 cities in the sample see public transport with a greater
mode share than walking.
Here is a four dimensional chart, comparing which cities
have:
·
more public transport usage versus walking
·
most used mode
·
private motor vehicle usage above or below 50%
(PM>/<.5)
·
population above or under 1,000,000
(P>/<10e6
|
Public transport > walking
|
Public transport < walking
|
Private motor vehicle dominant
|
|
PM>.5
|
PM<.5
|
P>10e6
|
43
|
11
|
P<10e6
|
9
|
15
|
|
|
PM>.5
|
PM<.5
|
P>10e6
|
4
|
7
|
P<10e6
|
8
|
16
|
|
Public transport dominant
|
|
PM>.5
|
PM<.5
|
P>10e6
|
|
26
|
P<10e6
|
|
2
|
|
|
PM>.5
|
PM<.5
|
P>10e6
|
|
|
P<10e6
|
|
|
|
Cycling dominant
|
|
PM>.5
|
PM<.5
|
P>10e6
|
|
3
|
P<10e6
|
|
|
|
|
PM>.5
|
PM<.5
|
P>10e6
|
|
|
P<10e6
|
|
2
|
|
Walking dominant
|
|
PM>.5
|
PM<.5
|
P>10e6
|
|
|
P<10e6
|
|
|
|
|
PM>.5
|
PM<.5
|
P>10e6
|
|
3
|
P<10e6
|
|
2
|
|
This shows us that almost every city has private motor
vehicle as the most frequently used mode of transport. All but 10 of the cities
which don’t have public transport as their dominant mode of transport see
private motor vehicles as the dominant mode.
Most car dependent cities as earlier described are in the Anglosphere
and Southern Europe.
Cities which are not car dependent are mostly in Northwest
Europe and East Asia.
There is no correlation of geography between whether cities
are car dependent or not. Hot climate, cold climate, a city with canals, a city
on the coast, or a city in the middle of a great flat plain… this doesn’t
matter.
Having freeways within 5 km of your city center makes no
difference as to whether your city will be car dependent or not.
Whether the freeways or tolled or not does not make a
difference.
The majority of cities which do not have private motor
vehicles as the most used form of transportation have public transportation as
their most used form of transportation.
Surface parking lots and zoning codes do you no favors in
terms of having a car centric city… but then again… Jakarta with 78% private
motor vehicle usage doesn’t have surface parking lots dominate their cities,
and their traffic jams are famous. Jakarta is what happens when you don’t have
sufficient mass transit around your city and you have little surface parking
lots. Your city will still be car dependent.
There is one factor and one factor only which is by far the
DOMINANT FACTOR in whether your city has low car usage around the world. Every
single city with low car usage in this sample has this one factor in common:
This data demonstrates a good quality rapid mass transit
network is the most important factor towards reducing dependence on private
motor vehicles.