Optimizing Speed-relative Density


emathias

Adoptive Chicagoan
It seems to me that in any given region, transportation planning should be geared to optimizing speed-relative density. Ultimately, people don't just want to get places quickly to save time, they want to use increased speed to expand the range of services (which I use generically to include both things like grocery stores and jobs) they have access to.

Given a number of people, a given geography, and a survey of available technologies it should be possible to figure out what the optimum arrangement of transportation would be.

Simplified example:

Assumed average speeds:
Walking - 3 mph
Bus - 12 mph
Car - 25 mph (this does assume the existence of at least some expressways)
NOTE: I start with these three, because in a good system they all function similarly - they get you very close to your final destination in any direction. Some other possible inclusions, for example commuter rail, have limited stops and will often require modal transfers or limited geographical access which complicates the calculations.

With those speeds, if you assume that for a daily trip, the average person doesn't want to travel more than an hour from home, that gives you a daily trip area (radius squared multiplied by pi) of:
Walking - 28 square miles
Bus - 452 square miles
Car - 1,963 square miles

Then we can calculate densities in two ways:

First, densities needed to approximate the densities a walker has access to in central Paris, a place where people have excellent access to a wide range of services by foot.

Walking - 50,000 pop/sq/mi (the default rough Paris density)
Bus - 3,125 pop/sq/mi (this density can't usually support extensive bus service, but if it could, buses would give people access to similar services that a walker in Paris has)
Car - 720 pop/sq/mi (this is a density probably found in agrarian counties in the U.S.)

Second, effective access for each mode in an area with typical American suburban density - 3,500 pop/sq/mi. (density multiplied by daily trip area)

Walking - 84,000 people
Bus - 336,000 people
Car - 700,000 people

At Chicago's density of 12,000 sq/mi, the numbers become even greater - 288,000, 1,152,000, and 2,400,000, respectively.

Commentary:

Obviously this paints a pretty stark picture as to why people like having cars. Beyond pop-psychology reports of feelings of power or freedom, there is a practical effect, especially in larger metro areas, of having access to a wider range of services.

Calculating comparisons when integrating rail or for areas without comprehensive bus service increases the complexity, but I think the numbers demonstrate where the challenges are. Buses will nearly always be slower than cars, unless they have dedicated, strongly-enforced lanes and car traffic is just comparatively atrocious, and even a few mph difference can greatly increase access.

The other point is that in smaller, isolated metro areas, of 50,000 people or so over an hour away from anywhere else, just about any mode can max out the available services, so there may actually be less pressure to own a car there than in a big city -even a big city with comprehensive bus service.
 
That assumes a place is a uniform density, and that you are starting from the absolute center. I think also there is a relationship between the frequency one makes a trip, and the perceived limit on distance and time.

My own attitude is to be open to personal preference, you'd have a city where people could have a house with a yard or a apartment if they wished, leading to a variety of densities, people from the surrounding area can drive into the city easily, while people in the city could walk to nearest local favorite place but also have somewhere to keep their car to explore the rest of the metro. high density in one place would support the one-stop-shop needs of lower density populations

Ideally then I guess you'd have a city that does accomodate cars but doesn't allow them to negatively impact the urban fabric.
 
Last edited:
^^ And you've just described something similar to LA, whose nodes of density allow people in the low-density areas the choice of driving to several different areas with good concentrations of jobs, services, and retail.

The same pattern, albeit with a lower density overall, can be seen in any sort of suburbia, where regional malls have often spawned perimeter office parks and townhome neighborhoods.

There are some intangibles that lead one to take transit, however. Commuter rail to the central city, so long as the tracks are relatively direct and straight, is often faster than driving because of congestion on the expressways. Transit can turn a commute into leisure time. Finally, transit often costs less, if subsidies are to be ignored (and most people don't think about the subsidy level when they make the choice).

In the end, though, I doubt we will be able to convince a mass amount of the populace to take transit. Cars simply offer the greatest amount of flexibility and freedom in most areas, which is why their side effects of pollution and noise need to be curbed through hybrid, electric, and fuel-cell technology. I'm not a fan of intelligent highways, which seem elitist and reflect a lack of trust in the common people. I do, however, think that transit needs to be strategically improved in already-dense areas or corridors, presumably congested, where they are able to offer a faster trip than driving.
 
Last edited:
Cycling Still Offers Quickest City Commute

Brad Aaron
Streetsblog
May 29, 2008
http://www.streetsblog.org/2008/05/29/cycling-still-offers-quickest-city-commute/

race2.jpg


streetsblog said:
To the surprise of no one -- with the possible exception of Bike Snob NYC -- bike commuter and social worker Jamie Favaro won Transportation Alternatives' 7th Annual Great NYC Commuter Race this morning, completing the 4.5-mile route between Fort Greene and Union Square in 16.5 minutes. Driver Emmanuel Fuentebella came in second at 22 minutes, and transit rider April Greene made the trip in 29 minutes. Writes T.A.:

According to 2000 Census figures, New Yorkers have the longest average commute in the country, about 45 minutes. However, the average bicycle commute in New York City only takes 30 minutes. And with 75% of driving commutes in NYC under 5 miles (the distance of today's challenge), there is great potential to shift driving trips to bicycling.

Congratulations to Jamie. Stay tuned for full coverage of the race from Streetfilms.

Photo: Transportation Alternatives

---
I average about 12 mph on my daily bike commute between the Castro and the Financial District, along Market Street. It's as dense a corridor as exists this side of Chicago, and is served by every MUNI metro line as well as most of the BART lines. Yet door-to-door, my commute is shorter now on two wheels than when I took the same route by subway.
 
^ Bike is easily the fastest way to get around central Chicago too, especially since most bicyclists don't obey any traffic laws, are faster than buses by definition (since they'll ride in the few bus lanes Chicago has), and barrel through all intersections.
 
Factors which will change urban density in the US

If one considers the financial crisis which started in September, 2008, as a watershed marking the beginning of a long term reduction in per capita US wealth as I do:

A) We likely will be forced to use much of the infrastructure that exists today, i.e., the auto and truck centric network, because the money may never be available to build new systems en mass.

B) Instead, we likely will have to make millions of small changes to the transportation grid as it now exists while letting significant sections turn into urban and suburban prairie. More energy efficient cars and a more electricity based economy will preserve upper middle and upper class suburbs while less affluent suburbs will decay.

C) Where right of way is relatively inexpensive and population density is on the "bus" density or greater, TOD developments are inevitable. Some will have the design elegance of the best Japanese urban planning, combining entertainment, employment, and, shopping, with quick connections to air transportation and dwellings. Many more will be "Stanlinesque" apartment blocks located "close" to transit stations.

D) The well to do will be the last to give up cars, if they ever do. The freedom to decide where to turn, when to go, while in the privacy of one's personal space will become increasingly expensive, and, even more than now, a visible proof of wealth.

IMO the US and Canadian cities that grow over the next 20 years will have marked density increases, but, will still be car centric with the caveat that a smaller slice of the population will be able to live the "classical 20th Century lifestyle."
 
I'm going to throw this in as well - the vast majority of low-density auto-oriented suburbia are not really destinations for people to go to (other than visiting relatives during the holidays), but they necessitate people who live in them to use a car because they can access so little from their single-use zoned neighborhood. The residents of these places NEED a car because they can't walk anywhere.

Besides industrial zones, most of the stuff people want to access (jobs, stores, cultural things and parks) are more often than not clustered around at least a busy highway, a downtown, or strip mall. If you live close to those, you are less likely to need a car to access it... unless your city forgot to put in sidewalks, which happens quite a lot in the 'ol US of A it seems...
 
That assumes a place is a uniform density, and that you are starting from the absolute center. I think also there is a relationship between the frequency one makes a trip, and the perceived limit on distance and time.

It doesn't really assume uniform density or central start. The simplified example uses averages over a diverse area, however for planning purposes you'd obviously have to do a more detailed examination that took into account where people live and work and congestion at various points along their routes, etc. I think the data is there (or can be gotten) to be able to do very complex analysis of this sort of thing, I just haven't ever seen results of such analysis published if it is actually ever done.

My own attitude is to be open to personal preference, you'd have a city where people could have a house with a yard or a apartment if they wished, leading to a variety of densities, people from the surrounding area can drive into the city easily, while people in the city could walk to nearest local favorite place but also have somewhere to keep their car to explore the rest of the metro. high density in one place would support the one-stop-shop needs of lower density populations

Ideally then I guess you'd have a city that does accomodate cars but doesn't allow them to negatively impact the urban fabric.

I don't think most people (not me, at any rate) disagree with you, my point is that it should be empirically demonstrable where to prioritize transit over cars and vice-versa, and then use that empirical data to make hard policy and tell people "this is what we're doing, and why we're doing it, so make your lifestyle choices based on this and don't expect a subway in the exurbs and don't expect 50-mph car-travel times in corridors x, y and z because the data simply doesn't support making that happen.
 
I'm going to throw this in as well - the vast majority of low-density auto-oriented suburbia are not really destinations for people to go to (other than visiting relatives during the holidays), but they necessitate people who live in them to use a car because they can access so little from their single-use zoned neighborhood. The residents of these places NEED a car because they can't walk anywhere.

Besides industrial zones, most of the stuff people want to access (jobs, stores, cultural things and parks) are more often than not clustered around at least a busy highway, a downtown, or strip mall. If you live close to those, you are less likely to need a car to access it... unless your city forgot to put in sidewalks, which happens quite a lot in the 'ol US of A it seems...

"The suburbs" include commercial strips. Even old, walkable cities have residential areas and commercial areas, it's just in the suburbs everything is auto-oriented so you can't usually readily walk between the two. But part of my point is that from a planning standpoint it's valid to consider a 10-minute car-ride from house to supermarket comparable (not equivalent to, but comparable to) to a 10-minute walk to a local market.

Beyond that, we can't really provide attractive alternatives to car travel until we can really understand what it takes to provide comparable convenience. It doesn't have to be exactly the same, it just has to be reasonably similar. If we can't even measure the way people use their environment properly, we can't hardly provide them with proper carless alternatives.
 
It seems to me that in any given region, transportation planning should be geared to optimizing speed-relative density. Ultimately, people don't just want to get places quickly to save time, they want to use increased speed to expand the range of services (which I use generically to include both things like grocery stores and jobs) they have access to.

Given a number of people, a given geography, and a survey of available technologies it should be possible to figure out what the optimum arrangement of transportation would be.

Simplified example:

Assumed average speeds:
Walking - 3 mph
Bus - 12 mph
Car - 25 mph (this does assume the existence of at least some expressways)
NOTE: I start with these three, because in a good system they all function similarly - they get you very close to your final destination in any direction. Some other possible inclusions, for example commuter rail, have limited stops and will often require modal transfers or limited geographical access which complicates the calculations.

With those speeds, if you assume that for a daily trip, the average person doesn't want to travel more than an hour from home, that gives you a daily trip area (radius squared multiplied by pi) of:
Walking - 28 square miles
Bus - 452 square miles
Car - 1,963 square miles

Then we can calculate densities in two ways:

First, densities needed to approximate the densities a walker has access to in central Paris, a place where people have excellent access to a wide range of services by foot.

Walking - 50,000 pop/sq/mi (the default rough Paris density)
Bus - 3,125 pop/sq/mi (this density can't usually support extensive bus service, but if it could, buses would give people access to similar services that a walker in Paris has)
Car - 720 pop/sq/mi (this is a density probably found in agrarian counties in the U.S.)

Second, effective access for each mode in an area with typical American suburban density - 3,500 pop/sq/mi. (density multiplied by daily trip area)

Walking - 84,000 people
Bus - 336,000 people
Car - 700,000 people

At Chicago's density of 12,000 sq/mi, the numbers become even greater - 288,000, 1,152,000, and 2,400,000, respectively.

Commentary:

Obviously this paints a pretty stark picture as to why people like having cars. Beyond pop-psychology reports of feelings of power or freedom, there is a practical effect, especially in larger metro areas, of having access to a wider range of services.

Calculating comparisons when integrating rail or for areas without comprehensive bus service increases the complexity, but I think the numbers demonstrate where the challenges are. Buses will nearly always be slower than cars, unless they have dedicated, strongly-enforced lanes and car traffic is just comparatively atrocious, and even a few mph difference can greatly increase access.

The other point is that in smaller, isolated metro areas, of 50,000 people or so over an hour away from anywhere else, just about any mode can max out the available services, so there may actually be less pressure to own a car there than in a big city -even a big city with comprehensive bus service.

It's a nice attempt to explain the car's appeal, but I believe your model is flawed for the following reason. I base this on my own experience living in both automobile and pedestrian-friendly neighborhoods.

People choose a mode of transportation that reaches destinations of highest benefit at the lowest cost.
They want the highest benefit-cost ratio. In a suburb, even though walking has the lowest cost, not many destinations of value can be reached in a timely manner. A car may be the only convenient way of getting groceries, travelling to work, etc. even though it is much more expensive.

However, in a dense, walkable, urban neighborhood, there are many destinations of high value (shops, grocery stores, restaurants, offices) in easy walking distance. Even though a car can technically reach more destinations, there aren't many destinations of higher value than those within walking distance. For example, if there's a grocery store right next to my house, I'm going to go to use that one; it doesn't make my life any better if there are another ten 5 km away. For the occasional destination (sports arena) beyond convenient walking distance, people may choose to take transit, carpool or join a car-share program. In this context, the benefit-cost ratio is clearly in walking's favor.

Hopefully with this analysis, the picture isn't so stark anymore. Cars have a lot of things going against them, namely their high capital costs. And with road pricing, it will become even less desirable when compared against other modes.
 


Back
Top