Goodbee Square: the Quest for a Contemporary Urban Pattern

Duany Plater-Zyberk and Company was hired to turn a greenfield about 50 miles north of New Orleans into a 1,280-unit blend of town and rural living. In the process, they proposed a radical new way of looking at the street grid. Fanis Grammenos explains.

Goodbee Square, a recent project by Duany Plater-Zyberk and Company constitutes a fertile departure from previous DPZ plans, integrating novel elements of traffic flow, pedestrian movement, traffic safety, park allocation and distribution and storm water management into the regularity of a simple grid. As a change in direction, and because street patterns are the most enduring physical element of any layout, it could potentially contribute to systematic site planning and, consequently, deserves a closer look.

The street pattern

Unlike the classic street grid of Portland (Fig. 1, left), the Goodbee Square street layout (Fig. 1, center) impedes north-south vehicular and pedestrian movement, although pedestrians are given another option (Fig. 2). Though the network is entirely interconnected, north-south movement becomes circuitous, indirect, and inconvenient, making driving an unlikely choice and vividly illustrating that interconnectedness by itself is insufficient to facilitate movement. The 3-way intersections limit through traffic, a lesson incorporated in TND (Traditional Neighborhood Development) and reminded of recently at Seaside.

 Three grid patterns
Fig. 1 Three layouts and three patterns (all plans same scale based on GOODBEE SQUARE)

Were we to apply this street pattern to a town center in nearby Covington or New Orleans, it would be entirely unworkable. Drivers would have serious difficulty reaching local destinations, and pedestrians would find their walks to be disorienting and unnecessarily long. But its very unsuitability for an urban center justifies its current usage as a suburban or ex-urban pattern.

As a principle of organizing circulation, it constricts traffic and confines expansion, unlike earlier simple street grids like Portland's regular grid or Savannah's cellular grid which, can be expanded in both directions without loss of functionality. If expanded to a large urban or suburban area, the Goodbee Square plan, with the discontinuous north-south roads, would severely limit traffic dispersal, a base for advocating regular grids. The Savannah and Portland grids both allow traffic to disperse in both directions, a feature that makes them equally applicable to city centres and to suburban locations.

The Goodbee Square street pattern eliminates unsafe four-way intersections within the neighborhood. The frequency of intersections with the main artery contradicts current traffic engineering practice, which subscribes to the notion that longer blocks reduce stop-and-go inefficiency and driver frustration; provide more uninterrupted movement space for pedestrians; opportunities for commercial façade size and treatment and increase on-street parking spaces which facilitate drivers becoming pedestrians and then shoppers. Longer blocks move cars more efficiently through Main Street, accentuating its role as a busy, vibrant thoroughfare. Perry's Neighbourhood Unit, a recurrent urbanist prototype, includes such blocks.

The north-south movement constraint, the lack of traffic dispersal and the frequency of intersections on Main Street contradict the usual practice, and require a fresh look at the Goodbee Square street network as an urban pattern.

The pedestrian network

A welcome attribute of the Goodbee Square plan is its pedestrian network which rejects the notion that streets are sufficient and suitable carriers for both car and pedestrian traffic. The plan has an independent north-south path network, which compensates for the inconvenience of the street network and favors pedestrians over motorists (Fig. 2). The footpaths are almost straight and cross parks frequently. Recent research confirms that directness and pleasure, as well as path independence from roads, are important attributes for enticing and enabling pedestrian movement.

 Three grid patterns
Fig. 2 Exclusive pedestrian paths in three plans as they would function currently

The principle of providing separate pedestrian paths could transform current site layout practice, which uses streets almost exclusively as the connectors for all mobility modes.

With respect to pedestrian movement, the Goodbee Square plan improves on that of Savannah and is a dramatic departure from the Portland plan, implemented in the18th and 19th centuries respectively, when the entire street and space network was a pedestrian domain and no other modes were dominant.


The Goodbee Square plan differs from previous DPZ plans in the number and location of its many charm-infusing parks which are regularly arrayed along streets with no attempt at civic monumentality or visual significance, unlike Savannah's plan, which locates parks within an 8-block cell as a focal point for each neighborhood. Both Savannah and Goodbee Square use parks as a means to enhance the pedestrian experience by placing them along pathways. The Portland plan has no obvious park strategy.

 Three grid patterns
Fig. 3 Parks and their distribution. (the Portland parks are indicative only)

Both the Goodbee Square and Savannah plans create a delightful environment with most residents near a park or with park views. Savannah, however, does it with greater economy of means; four parks compared to nine in Goodbee Square within a similar area (Fig 3). While parks are generally welcome, land value, urban density, unit yield, unit price and municipal maintenance cost considerations would normally lead to reducing their number.

The quest beyond Goodbee Square

Can the advantages of the Goodbee Square plan be retained while alleviating its limitations? We believe that a plan combining the main characteristics of the Portland, Goodbee Square and Savannah could do just that. If feasible, such a pattern can then be applied to many 21st century site plans, much like the simple grid pattern found in hundreds of North American plans over the centuries.

The Goodbee Square plan, an offset grid closely resembling the Flemish Bond brickwork pattern, would be the starting point for a new template, meeting the following objectives through proven planning strategies:

  • Keep vehicular traffic safe with a high proportion of 3-way intersections
  • Reduce cut-through traffic by similar or other means
  • Improve traffic flow in both directions using Savannah's cellular structure
  • Improve traffic dispersal by a car-sized grid
  • Improve pedestrian mobility utilizing Goodbee Square's path separation
  • Make parks a focus as in the Savannah cell
  • Improve land use efficiency and unit density

As an experiment, we combined the Flemish bond pattern (Fig 3), with the cellular organization of the Savannah plan by imagining a two-directional Flemish Bond. This new stencil emerges as a re-invented Savannah cell with a geometry that satisfies all the requirements for vehicular circulation and pedestrian movement; Jefferson, Oglethorpe and Hippodamus meet at the square.

 Three grid patterns
Fig. 4 From a unidirectional Flemish Bond towards a contemporary network pattern

As in the Goodbee Square and Savannah plans, all intersections within the neighbourhood are 3-way, satisfying the first two objectives. (Fig 4). The cellular structure creates a car-scaled grid that moves and disperses traffic, meeting the third goal.

Every block faces a park, generating a delightful milieu. Separate, strategic through-the-block paths achieve high pedestrian connectivity in every direction, and short streets provide easy access to nearby through-routes for drivers.

Efficiency of land use is achieved by subtracting half the Goodbee Square through-the-block path segments; reducing parks from eight to four, and reducing street length in equivalent areas.

 Three grid patterns
Figure 5. Recombination of Savannah and GOODBEE SQUARE site plan elements (red lines: pedestrian paths; blue dotted lines: car lanes or greenways)

The interface with Main Street now includes two long block faces for every short face, improving traffic flow, parking and pedestrian safety and enjoyment.

The Goodbee Square plan lays the foundation for the next step in the search for a contemporary pattern which might be called a "fused grid," as it combines car dominant and pedestrian dominant paths to form a complete, amalgamated network.

Fanis Grammenos is a principal of Urban Pattern Associates and was a Senior Researcher at Canada Mortgage and Housing Corporation for over 20 years. He focused on housing affordability, building adaptability, municipal regulations, sustainable development and, recently, on street network patterns. Prior to that he was a housing developer. He holds a degree in Architecture from the U of Waterloo.

The author wishes to thank his colleagues Doug Pollard, Barry Craig and Ray Tomalty to whom the article owes its clarity and logic.



Mike Lydon's picture

It's about stormwater

The Goodbee Square plan was developed using Light Imprint urban design techniques, pioneered by DPZ's Charlotte office. Each of the squares is designed to handle stormwater in a way that reduces the overall cost of infrastructure, while providing a fantastic civic amenity, and yes interrupting the free-flow of automobiles.

Beyond stormwater

Both stormwater management and integration of urban agriculture are the key drivers for the pattern. A few other issues drive the pattern that should be taken into account before examining it as a driver of a larger urban pattern. The principal reason that blocking n-s traffic is endured is due to both natural and physical constraints that block connectivity in these directions. The grid connectivity intends to connect to the west, which is the location of the town center and a connection to the local highway. In the case of a larger adoption of such a pattern, it would be wise to interrupt with more n-s avenues, as in Savannah. This is a relatively similar approach to the Flemish bond strategy. The one thing that the bond pattern lacks, and an application of Goodbee to a larger scale, is applicability of mixed-use infill / growth and adaptability. Goodbee is paired with a more regular grid adjacent that handles mixed-use, but it can't really grow too much. Savannah on the other hand allows for mixed-use to spread both along the avenues and into the squares while maintaining orientation.

Now park frequency would certainly be a detriment if they were all to be maintained by the city or an HOA. The intention, however, of the parks is for capturing runoff and to provide space for urban agriculture. A new plan detail for the squares has just been completed that subdivides them into community gardens with the central water catchment and treatment area as a water supply for irrigation.

So generally as a departure from previous DPZ planning techniques, it is an experiment in stormwater management and urban agriculture. I think that analyzing the pattern for its advantages and disadvantages in larger application of connected grids is healthy. These other main drivers should be brought into consideration as ways to help cities reduce their stormwater infrastructure costs, upkeep and as a way to prevent degradation of riparian corridors, as well as a way to integrate community based agriculture.

Two-Directional Flemish Bond.

Goodbee Square is a very interesting design, which I think will have a lot of influence, but I have a couple of concerns about the claims for the two-directional Flemish bond pattern.

"The interface with Main Street now includes two long block faces for every short face, improving traffic flow, parking and pedestrian safety and enjoyment."

I think Main Streets actually work better if you have short blocks to make it easier for pedestrians to cross. This slows down traffic a bit, but it allows the stores on both sides of the street to support each other. If you come to a store on one side, you are more likely to walk across the street and also use stores on the other side, if you have short blocks and frequent crossings.

"short streets provide easy access to nearby through-routes for drivers."
"a 'fused grid,' ... combines car dominant and pedestrian dominant paths to form a complete, amalgamated network."

To me, this sounds too much like Perry's neighborhood unit concept, which was a destructive element of post-war planning and which this article explicitly rejects.

Perry also said that there should be local streets within the neighborhood unit that discourage driving and are pedestrian dominated, and arterial streets around the unit that are "car dominant." As we all know, the car-dominant streets are barriers to pedestrians and they encourage people to drive whenever they go outside of their own neighborhood unit, even if they are just going across the arterial street.

For this design to work for pedestrians, the "auto-dominant" streets have to be pedestrian-friendly as well as auto friendly. They have to be multi-use rather than auto-dominant.

Charles Siegel

It's a latter-day cul-de-sac model

We WANT there to be through traffic on local streets, because it means everybody isn't forced to go on the arterial. It seems to me that this plan is just a latter-day rethinking of the old cul-de-sac model. Pedestrians, you may recall, were allowed to walk between cul-de-sacs too.

Points for original thinking, but this seems like a new way of continuing a failed model.

In search of Flemmish bond, missing the point on Goodbee Square

The greens also serve as interconected stormwater facilities. (See the earlier Planetizen post)
There is only 4 feet of elevation change over the entire site.

John Anderson
Anderson Kim Architecture + Urban Design
Chico, CA

Pedestrian dangers

Much of this effort appears to rely on a misunderstanding of urban history and misconceptions about traffic.

It gets off to a bad start by accepting the myth that "the entire street and space network was a pedestrian domain and no other modes were dominant" in the 18th and 19th centuries. In cities, horses and streetcars alike were quite dominating presences. These were well captured in art of the period as well as the written descriptions of urban conditions at the time. They were inconvenient, and they were dangerous. Statistics are not readily available, but the dangers of being run over - or kicked or trampled - were high and factored highly in public concerns. The golden age of pedestrian utopia is a myth.

It appears they extrapolated from this mythical pedestrian past to conclude the dangers of today's traffic could become just as irrelevant if they simply try to make it less convenient for drivers. Nevertheless, fantastic ideas do nothing to help real drivers see pedestrians any better, or to anticipate them when they cross in unusual places.

Unless they neglected to mention that they are also employing grade separation for the pedestrian paths (and effectively stealing their ideas wholesale from earlier planning like Radburn without giving credit where credit is due...), the grid they illustrate would increase the danger to pedestrians. If these are, in fact, short blocks, and the pedestrians are actually encouraged to cross mid-block as illustrated, then the pedestrians would be crossing at the point where motorists would be completing their turns and beginning to accelerate. Moreover, the curbs a pedestrian would be leaving, particularly the one on the inside of the turn, would be difficult for a driver to observe while trying to negotiate the turn. Not to mention this is not a place where drivers would ever encounter a possible conflicting movement anywhere else in their driving experience, making it entirely unanticipated.

Encouraging pedestrians to walk out in front of drivers who don't expect them is never good planning!

Pedestrian Realms and Havens

Most thoughtful, practical, advice for improving a “system” with safety of pedestrians in mind.
Correct, Stein in Radburn got it right when he gave pedestrians their own path network that was more extensive compared to roads (within the neighbourhood). When paths did cross roads at the neighbourhood boundaries, they were grade-separated; freeing the pedestrian entirely. That idea is worth remembering and reinterpreting; reinterpreting, because Radburn may be ill-suited for Any-polis or Anytown.

Grade separation, may be unfeasible or impractical in some locations; downtown or Centertown, for example. Even so, a few cities, faced with the conflicts at grade, have introduced “Stein-like” alternatives – underground networks (e.g. Toronto, Montreal) and +15 (e.g. Calgary, St Paul); ideal they are not, as most adaptations . Few and the exception because there are many factors at play; most downtowns or peripheral neighbourhoods may not be amenable to grade separation; some forced examples are inconvenient, unpleasant or risky.

An alternative for these cases is partial independence and partial overlap, a “fused grid”. The idea is that where the two networks converge or intersect, within a neighbourhood, to give pedestrians an edge (Alexander has a pattern for this). Examples exist of a raised landscaped median exactly at the point of path and street intersection. The pavement narrows and curves and visual clues abound; pedestrian and driver have a chance to negotiate. “Median” only by name for what in fact is an “islet”.

We need to find ways to reclaim the mythical pedestrian realm that was real, is still real in some places and becoming real in more. Many of Pompeii’s streets (still visible today) never saw a cart or even a horse; they were simply too narrow. The same is true for numerous streets of ancient Athens, Rome and many medieval cities (horse ownership was a privilage of few). Mediterranean island villages perched on hillsides are living examples of a pedestrian haven; totally inaccessible except on foot. Cities have pedestrianized streets (Stroget, DM) or districts (Montpellier) and the list is growing. Separation first, careful, studied mix after, is a path to regaining ownership, and making neighbourhoods true pedestrian realms.

Prepare for the AICP* Exam

Join the thousands of students who have utilized the Planetizen AICP* Exam Preparation Class to prepare for the American Planning Association's AICP* exam.
Starting at $245

Essential Readings in Urban Planning

Planning on taking the AICP* Exam? Register for Planetizen's AICP * Exam Preparation Course to save $25.

Wear your city with style!

100% silk scarves feature detailed city maps. Choose from five cities with red or blue trim.

Brand new! Urban Grid City Collection

Each city has its own unique story. Commemorate where you came from or where you want to go.