Saturday, February 20, 2010

What the Savannah Square can do better than the Roundabout


An Odonomy of Savannah, Part III

I am a proponent of using roundabouts to lubricate traffic flow in many contexts; however, a new precedent has become my default alternative to signalized intersections. I often see transportation and development plans employing roundabouts where their traffic flow advantages are not really necessary. A roundabout intersection is an a auto-centric option in my book, which should be used only when vehicular through-put is an important goal. Savannah's squares, I believe, offer urban designers a relatively uninterrupted traffic flow alternative with more urbanistic and multi-modal advantages.

As I pointed out in Part II of this series, Savannah's one-way flow pattern is primarily controlled by yield signs (and no sign at all where the grid simply splits one way flow). Savannah's squares, like roundabouts, can thus provide relatively smooth traffic flow, which, while not as fast and effective as a roundabout, are quite appropriate for a highly pedestrian and bike-friendly urban context.

I hope the figures at left are useful for understanding the urbanistic advantages of Savannah's "square-flow" strategy. The figure on top represents a roundabout of 100 feet diameter, a typical size for a single lane roundabout. The second figure is a typical Savannah Square at the same scale (note the dashed 200'x200' "Portland Block" superimposed over both figures).

While the square has obvious disadvantages in terms of volume and speed of traffic flow, here are some of the advantages I note in the Savannah square-flow:

1) While the typical Savannah garden square (approximately the size of a "Portland Block") can inscribe a typical roundabout intersection, the square's T-intersections remove the long merging distances required. Ignoring for a moment how comparatively rare the radial convergence of more than three streets appears in cities, note that for a roundabout to distribute traffic effectively for eight converging street segments (as the Savannah squares do), the central island diameter would have to double in size, a need that would demand slightly more right-of-way than the Savannah square. (Actually, in the real world an eight-access point roundabout would probably effectively appropriate considerably more real estate for right-of-way due to the need to radially align the converging streets.) Moreover, just imagine how sadistically delirious an eight-access point roundabout would be for most drivers to navigate. The square actually ingeniously adapts the roundabout's radial flow for compactness, by removing the merging distances, as well as simplifying the geometry for drivers, preserving their sense of navigation by conforming to the gridiron (roundabouts, on the other hand, attract an overabundance of signage partly because they don't work all that easily with driver intuition).

2) The T-intersections of the square increase drivers' attention to bikes and pedestrians, since both pedestrian and traffic flows converge at the same location. Drivers have to yield and pay attention to all modes of intersecting traffic, as opposed to roundabouts, which divert pedestrian crossings and bike lanes away from the roundabout. Moreover, all the intersections in a Savannah square are 90-degree intersections to one-way flow, an obvious safety advantage.

3) Due to channelization standards, higher speeds, and, sometimes, merging distances needed near the roundabout, driveways and on-street parking locations will be pushed further away from the roundabout on the streets. I've seen numerous driveways access generously sized roundabouts, as at the Circle in Waco, TX, but try replicating those quirky throwbacks to bygone ways nowadays! In contrast, the Savannah square's straight segments and lack of merging/weaving needs encourages parallel parking to be safely provided on both sides of the square's streets. (Driveway locations are also possible on the square, but thankfully Savannah's alleys typically remove this need.)

4) The 90 degree turning is a natural traffic calming measure. For this reason, bikes are able to easily share the road around the square, removing the need to create bike lanes (or to demarcate "sharrows") or to put bike and pedestrian travel into conflict (a typical condition for all roundabouts but the residential "minis").

5) For sustainability purposes, this important advantage has always endeared me to Savannah's peculiar take on a gridiron fabric. The Savannah fabric creates east-west oriented development patterns, perfect for our Southeastern climate. Had Oglethorpe consciously invented it for this reason, along with all the other modern advantages it provides, I would consider him a genius pushing DaVinci level caliber. The perimeter development strategy of the roundabout option above is hard to adapt to this need. This is one of the things that makes me honestly suspect that an angelic power gave Oglethorpe his inspiration!

6) The land-use advantages are numerous. First, Savannah's squares are functional city parks, easily accessible on all sides, which by virtue of being what they are, attract strollers. Look at all the safe crossing points the square affords (as opposed to a roundabout's...uh...none). A roundabout island might be pretty, but it is a seldomly used open space resource by humans. Secondly, note the development footprints shaded in gray in the figures. In the top figure, I represent the typical development pattern we encounter nowadays, which both developers and municipal codes increasingly prefer for urban mixed use districts (if not require by law as form-based codes do): namely, the building envelope is moved to the perimeter of the block. I suppose this creates an attractive roundabout intersection, but notice that it allows the block interior behind the buildings to be typically conceded to vehicular needs. Savannah's form instead promotes on-street parking and more compact lot patterns. Because of the shallow lot depths (usually no more than 70 feet) surface lots and parking garages, where needed, can't be tucked behind other development. This allows them to become a competitive use along with all other development, making them a competitive land use resource rather than an isolated one that is insulated from potential conversion to other real estate uses.

Just some handy numbers to illustrate the land use advantage. Each of the figures above represents the same land area, which is the size of the central Savannah grid modules at 600' x 600' (8.3 acres). The total right-of-way area in the roundabout figure at the top is 34 % of the 8.3 acres. Four 250' x 250' lots abut the roundabout and the typical double-loaded corridor building footprint coverage shown is 45% of the total lot area. To achieve an overall FAR of 1.5 (a great base target for mixed-use, urban development), the four lots would need these buildings to average 3.3 levels overall. Based on Charlotte's MUDD code, the white space behind each building is, in fact, exactly enough space for a surface lot to park a 3.3 level apartment building or a 2.5 level mixed-use building with retail/office on the lower level (this means that you would need to build at least one parking deck on one of the four lots to reach an overall FAR of 1.5 if you want retail or office space as part of the mix). Alternatively, you could build townhomes or four small (relatively inefficient) "Texas Donuts" of 220' x 250'. (More than likely, developers would want larger block areas than the four 250' x 250' lots above can provide to achieve the efficiencies they need, but let's for comparative reasons entertain them here).

Savannah's development pattern is finer grained and with a much greater percentage of area dedicated to right-of-way: 45% of the total 8.3 acres! The advantage is that you accommodate a lot more on-street parking, including diagonal parking on the wider streets, reducing the amount of off-street parking needed. You also end up with smaller blocks, a highly linked street pattern and more than double the street frontage of the hypothetical MUDD code buildings surrounding the roundabout above. Admittedly, Savannah's pattern encourages less parking intensive residential uses, but you would be surprised by how many civic, office and retail uses Savannah's fabric accommodates without a seeming shortage of parking spaces. Yes, you do see a lot of surface lots punctuating the potential development coverage, but I believe the pressure to convert these lots to other uses is a healthy one and helps set off-street parking prices closer to their real value. Moreover, let's say only 1/2 of the lot areas in Savannah's 600' x 600' grid module have building coverage (the rest being open space or space dedicated to vehicle storage). An overall FAR of 1.5 can be achieved with an average of 3 level buildings throughout, something that Savannah achieves handily. The denser squares in Savannah easily exceed FAR 1.5 by more than double this factor (any seasoned planner/architect will confirm this with simple visual inspection of the Google Maps aerial). By virtue of its compact and shallow lots, Savannah far exceeds the density efficiency of the large block perimeter, auto-centric development regimes that roundabouts encourage.

In short, consider employing the Savannah square-flow strategy as smarter way to handle traffic flow while promoting a density-efficient land use mixture and bike and pedestrian friendliness. I would only use a roundabout when at least one of the intersecting streets (preferably both) is a thoroughfare or high-volume traffic street. Otherwise, I'd prefer to square it.

Thursday, February 11, 2010

VMT and Congestion geo-tagging. Bueller ...Bueller?

Over at the Bellows, Ryan Avent wonders why everyone is still so tight-ass about VMT...

I think the place to start is with congestion pricing, although I’d also like to see a share of revenues from carbon pricing devoted to transportation. But I don’t think it’s absurd to put a VMT tax on the table. Policy merits will be balanced in Washington against political acceptability, and clearly there are legislators who feel that the idea of a VMT tax — of paying for the miles you drive on public roads — might make sense to drivers. I don’t think that’s nuts.

I know that many people feel that the civil liberties issue is a dealbreaker, but I have a hard time believing that. These days, people don’t blink at the prospect of unwarranted domestic surveillance, they walk around with multiple mobile devices, and they live half their lives online. A gizmo that keeps track of how far they’ve driven isn’t likely to phase most Americans.


Not nuts...and in fact, it is only fair. Americans love fairness. And if you can measure it, and present it, consistently, air-tightly, logically and comparatively, so much the better. This is what Google does after all. We get it.

As far as the actual tracking system used for implementing a VMT system, we have a myriad ways to think about this. Getting hack-proof "gizmos" to track VMT, in itself, is not going to do it. Any implementation has to fine-tune a fine and audit system. It works for transit boarding. Primarily you have to think hard about the incentives for fraudulent behavior and learn the best ways to police against it in a fair and efficient way. No one likes deadbeats, so we will get innovative. Big deal. We do this for far more complicated things. If we're smart, we'll jigger our wireless technologies to pick up our slack ...and do it more neutrally and fairly while they are at it.

We also need a system that is graduated and offers rebate incentives (just like income taxes). VMT should be graduated on weight of vehicle anyway. And (as Ryan suggests) we can combine all tax approaches, taxing fuel inefficiency, congestion zone travel and VMT as well. We're not far from having the capability to monitor this in a smart and decentralized way. Look at the god-awful amounts of data banks have on you already.

And...I don't know why some are suggesting vehicle-installed (and "hack-proof") devices to begin with. It seems to me we need to think about a more robust monitoring system, utilizing multiple data points. You can use a wireless geo-tagging system for starters. The central monitoring interface can still be at the pump. The wonderful thing about pumps is they are already wired to exchange information and they don't move (easily). What information is exchanged at the pump and how is what we have to think harder about. How we work in incentives should also be pondered and debated, e.g. the greater the VMT per last deposit's gallons and the lesser the percentage of (time/location-dependent) "congestion geo-tags" your vehicle has acquired in that time, the greater the rebate you will see lopped off the fuel charge...right there at the pump! This way, the monitoring equipment works in the driver's favor sometimes. Convenience store owners will need to get deals themselves for relaying the data, perhaps with an opt-in voluntary data-sharing program that rewards customers with deals as well....Kinda like Google-style marketing.

C'mon, people, we gotta get all American about this.