By Michael Brown, PE, AICP
It is in vogue to convert one-way couplets back to two-way streets because many urban planners believe that one-way couplets are inherently high-speed and auto-oriented, and therefore poor choices for would-be walkable environments. I believe the opposite is true. With one-ways, it is easier to achieve driver compliance with whatever speed is desired, be it 50 mph or 15 mph. One-ways are safer, more pedestrian-friendly, and can support larger Activity Centers as well. One-ways can move more vehicles with fewer lanes, which may sound auto-oriented, but can actually favor place-making. This article discusses the virtues of one-way streets, concedes some weaknesses, and addresses common misconceptions.
Can the tortoise win the race? Watch this video before this article or read the introduction article titled Place-Making Innovative Intersections.
Please comment or view comments about one-way couplets at the end, and if you like it please share with anyone who might care. You can also contact me privately to learn more.
A Town Center Intersection is created any time one-way streets are involved. Though one-ways are not particularly “innovative,” they can achieve high efficiency because there is no need for left-turn arrows, since there is no opposing traffic.
Imagine two typical suburban “stroad” highways coming together (high-speed street/road hybrid). Normally this would result in double-left turn pockets within a single massive auto-oriented intersection, accompanied by excessive congestion. But this diagram by Calthorpe Associates shows how the two highways can each be divided into lower-speed one-way streets as they cross, resulting in four smaller, friendlier intersections, with more parcels having good access and visibility. The system of four intersections can handle more overall traffic than the double-left intersection it replaces, but importantly it feels like less traffic, because each intersection is small and each does in fact have less traffic.
This Town Center Intersection in San Marcos, CA shows how a system of four simple intersections replaces what would have emerged as a single huge intersection (16 popular corner parcels, as opposed to just 4). Even though this system can handle much more traffic than a single signal with double-lefts, the photo shows that it does so within an attractive and safe pedestrian context, creating the appearance of less traffic because there IS less traffic at each intersection.
It is currently in vogue to convert one-ways back to two-way, under the pretext that one-ways are “high-speed & auto-oriented” and therefore bad for multi-modal Place-Making. No doubt there are situations where conversion makes good sense, but I believe one-ways are unfairly taking the blame for many auto-oriented situations that have nothing to do with one-way flow.
Why do one-ways get a bad rap? Many one-way streets were introduced in the 1950’s and 60’s as traffic engineers wrestled with how to deliver huge numbers of vehicles into CBD’s using streets that had always been very narrow. In that generation, Complete Streets and traffic calming were far from anyone’s mind, so one-way couplets were designed for high speeds. Land uses and other modes were an afterthought at best.
Thus because there are many examples of older, high-speed, auto-oriented couplets, it is common for planners and the general public to unfairly malign couplets as incompatible with walkable, sustainable design. There are thousands of examples of terrible, high-speed two-way stroads also built at the same time, so shouldn’t two-way streets also be labeled as inherently auto-oriented? Speed is all about design, and a driver’s decision to hit the gas has literally nothing to do with one-way or two-way. In fact, you’re more likely to get speed-limit compliance for whatever speed you want with one-ways, as I’ll note in a moment. Consider these virtues of one-ways relative to two-way…
Pavement is half as wide; thus easier to cross
Signal cycles are usually shorter with fewer conflicts; both faster and safer to cross
Pedestrians only need gaps in one direction; thus faster and safer if crossing without a signal.
Traffic is spread across two streets, doubling the parcels with high visibility and good access, and increasing the opportunity for a larger Activity Center as opposed to a single all-important street.
Narrow streets have better pedestrian enclosure even when buildings are only one or two stories.
One-ways do not require a left-turn lane, which means more space for other uses.
Speed is fairly independent of capacity – you can move just as many vehicles at 30 mph as at 50 mph, (because space between vehicles is less at 30 mph). Thus you can redesign high-speed one-ways as low-speed, walkable boulevards without creating new congestion.
Regarding the last point, couplets encourage high-speeds only if intentionally or inadvertently designed for high speeds. The maximum speed of traffic has nothing to do with whether it is one-way or two-way, but average speed is definitely affected (drive slower, travel faster). Lane widths, on-street parking, stately trees (or lack thereof), signal spacing, speed limits, and the nature of the urban fabric have much more influence on maximum speed.
One-way streets with closely spaced signals can achieve perfect signal progression. While that may sound “fast and auto-oriented” because green-lights are good for autos, it actually means we can more easily achieve compliance with any desired speed limit. Want vehicles to travel 50 mph? Then make 12-ft lanes, no parking, and time lights to fall like dominoes at 50 mph. Want the same street to be 30 mph? Go with 10-ft lanes, on-street parking, great street trees, paver cross-walks, and set the dominoes to fall at 30 mph. Drivers will then drive exactly 30 mph, because at 35 they reach the next signal while it’s still red, but at 30 they hit green right-on. Two-way streets can’t “fall like dominoes” so drivers almost always travel 5-10 mph faster than the posted speed limit as they see no reason not to. Many streets in Portland are timed for roughly 12 mph – and drivers comply because they see it is pointless to go faster. But convert those same streets to two-way, where signals can’t be timed for any discernible speed, and maximum speeds will almost certainly increase substantially.
With many one-way to two-way conversions, it is common to hear that business has improved. Everyone’s excited because there is more walking and biking, and renewed development interest. Since the conversion grabbed the headline, people assume one-ways were holding business back. But such conversions often include significant Complete Street beautification efforts, traffic calming, etc. It is a mistake to conclude that one-way vs. two-way was the change of greatest significance when multiple changes occur. In places where a more commercial street needs additional traffic and a more residential street needs less, then conversion may in fact be an element of “breathing life back into a dying Main Street.”
But in areas where you have visions of an ever-expanding Activity Center where commercial viability and multi-family livability on both streets is important, one-ways can support larger, more walkable centers than two-ways. I passionately believe that one-ways will prove better for livable and sustainable development goals than two-ways, when given the same respect (Complete Streets, traffic calming, etc.) This is partly because one-ways don’t need left-turn storage lanes, so you get all the more room for other uses, and also due to the more walkable facts already noted above. Are you hoping to breath life back into your languishing business district served by tired, auto-oriented one-ways? Great! – Just reinvent your one-ways, but don’t “throw the baby out with the bathwater.”
One-way streets are a common thread linking America’s largest and most successful urban centers – Portland, Denver, Manhattan, New Orleans, and even Boise, Idaho. Each of these have many examples of walkable, narrow, low-speed couplets. Alternative modes are vital in all of these, but their ecosystems still require vehicle circulation for deliveries, construction, transit, and a great many private and shared autos. Their network of one-way streets ensures that no particular street has overwhelming traffic. Here are some examples.
Try converting their one-ways back to two-way, and it would immediately create gridlock, which in turn increases the sense of auto-dominance and would hinder market interest in further development. “Where did all these cars come from? Did we suddenly become auto-oriented?” – Nope, same cars, but it feels like 3-times as many because they’re stacking up for blocks, never able to get out! Congestion decreases Livability, and limits the market’s ability for continued development.
Auto-accessibility may not sound like a critical need for Transit-Oriented Development, but it is. Developers understand that even if alternative modes gain impressive shares, odds are most trips to/from their potential development will still be by vehicle. So when congestion becomes chronic, some developers will change their mind and sprawl out to the fringe. Thus one reason we sprawl outward is that congestion won’t allow us to grow inward. Creating a high-density center from humble strip-mall beginnings is only possible in places that have great auto-access.
A one-way couplet will provide more access, even at lower speeds, than the same two streets each operated as two-ways. Would-be activity centers may never reach critical mass for major transit investments until first growing as dense as auto-access can allow.
Peter Calthorpe has published a short white-paper titled “The Urban Network, a New Framework for Growth.” The primary diagram from that article is shown below. One-way couplets are a primary feature of Calthorpe’s Town Centers and Village Centers. He notes that the couplets are a good way to provide significant vehicle access to the core centers, within a “slow-but-steady” very pedestrian-friendly framework. Personally, I think this entire network concept is a stroke of pure brilliance and a very good model for any/every Greenfield area poised to urbanize in the near future. He also discusses the San Marcos Town Center, noted earlier.
Calthorpe also provided some analysis of a Town Center at Issaquah Highlands just east of Seattle, shown below. The default plan would have resulted in a single huge intersection with 166 ft for pedestrians to cross. On the couplet, one street requires a 40 ft crossing, and the other just 28 ft. The couplets also allow on-street parking, which would have been unattractive (under-utilized) and possibly disallowed in the default plan. Traffic engineers estimated 11% faster actual time with the couplet in spite of additional intersections, on-street parking, and reduced speed limit.
Huge densities are emerging here, which would have been almost impossible to support under the default “Stroad” plan.
Years ago I analyzed a Town Center system of four one-way intersections as a replacement for a single Stroad intersection. The image below is from Synchro, perhaps the most common software used by traffic engineers for determining roadway geometry needs. The stroads represent two 7-lane arterials. At the intersection there is a double-left, three through lanes, and a right-turn lane on each approach. Huge! I increased volumes until it reached Level of Service E, the failure point, which is about 1-minute of delay per vehicle. At that point there were 2,000 vehicles per hour entering the intersection from each direction, or 8000 total.
Then I did the same with the couplet. To make the experiment similar, I made three through lanes on each street, one right pocket, but only one left lane as that’s all that was needed. In this case, each intersection has 5,800 vehicles (rounded to 6,000). The sum of all four is 24,000, but really represents about 14,000 unique vehicles, as many use more than one intersection. But even at 14,000, which is 75% hire than the stroad intersection, these relatively small intersections are uncongested (Level of Service D).
Lest ye think this proves couplets are auto-oriented, consider this.
- The system can support almost twice as much traffic, which may mean up to 3-4 times the density of localized land uses (given a high share of walk, bike, transit trips within a higher density node).
- If you’re only planning for 8,000/hr anyway, then you don’t need as many lanes on the couplets as you’d need on the stroads.
Here’s another way to look at it. The cross section below represents the stroad in the previous analysis, where the intersection can serve 8,000 vehicles per hour. It’s huge! 121 feet for pedestrians to cross, intimidated by 10 lanes of traffic! With a road this huge, speeds naturally gravitate to 45-50 mph. There is no chance that a quality walkable environment can emerge here.
This is what the couplet of the previous analysis would look like at the intersection. Remember this system can support 14,000 vehicles per hour, 75% higher, and all of that at just 30 mph! But it is also vastly superior for pedestrians and mixed use development. If you allow on-street parking, then it couldn’t support 14,000, but it would still support much more than 8,000. Further, if you have additional connectivity in the area, you can downsize even more to just two through-lanes per direction, and still support impressive densities without much congestion.
Obviously creating a couplet requires a second street – challenging to retrofit if there is no second street, or if the candidate alignment has politically challenging uses. There are often more opportunities for couplets than meet the eye, and Metro Analytics is skilled at finding them. But our toolbox goes well beyond couplets and we can probably help you do something win-win for multi-modal access and economic development.
Couplets are ideal for Greenfield areas being planned as major activity centers, or have potential to evolve as centers. One drawback of a couplet is that for transit, people prefer to come back to the same place they got off. Not a big deal if separated by just a block, but becomes onerous in wider couplets. In Greenfield environments, consider creating a “Triplet” such as in this figure. As traffic approaches your soon-to-be Center, it diverts onto one-ways as in the Calthorpe design. But if you add a middle alignment (shown in green), then transit can continue on to what might be a pedestrian plaza, angle parking, etc. If you then design parking at the edges, and use frequent circulators, you may need almost no parking closer to the primary intersection. Denver’s 16th Street Mall and Boulder’s Pearl Street are good examples of triplets.
This Greenfield concept for Salt Lake’s West Bench shows a major arterial entering the activity center (black) is split to two one-way streams, and slowed to 30 mph. A center alignment (blue) is for light-rail and other pedestrian-oriented uses. It crosses several other smaller couplets, as well as many two-way streets.
Below shows ongoing plans in Logan, Utah, where City leadership is excited to convert their asphalt-dominated historic Main Street into a one-way couplet, primarily to handle growth in traffic. The conversion would allow an extra lane in one direction, but it could also fit bike lanes, additional landscape, and potentially even angle parking in some places all within the same 126 feet! This is just one of many configurations yet to be explored, but at least with a couplet they have options they didn’t have before.
Below are more renderings and analysis from our Logan study. The existing 5-lane cross section has 62-ft of pavement for traffic, and is expected to flow at just 7 mph in the PM peak in 2040. But if converted to a couplet, there might be just 33-ft for traffic, and an average of 20 mph when factoring in stop-lights at cross-streets (same 35 mph speed limit in both cases).
The couplet will also encourage their downtown to expand to the first street to the west, helping them attract new development downtown rather than losing it to suburban locations.
In summary, there is nothing inherently “auto-oriented” about one-way streets. They are not “high speed” unless they are designed to be high-speed (intentionally or unintentionally). Yes, they are more efficient at moving more vehicles with less delay, as they don’t need left-turn arrows, and have better coordinated signals even at lower speed limits. But just because something good has happened for cars need not define that something bad has happened for other modes. Indeed it can be win-win for all modes as there are usually opportunities to reclaim automobile space for other modes. When all modes are doing well, there is no drag on a Center’s ability to expand.
If “efficient vehicles” seems auto-oriented, consider that high delay creates the feel of auto-oriented, because delayed vehicles stacked up for blocks are annoying to everyone, and it pressures engineers to “go rouge.” Couplets help vehicles keep moving even at low speeds. Thus many vehicles can move through sensitive areas without it ever feeling like as many as it is.
Where two highways cross in rapidly urbanizing Greenfield areas, consider adopting Calthorpe’s Town Center design, or Metro’s Triplet design. If planning a large Greenfield area, lay out a grid of streets that could become one-ways if the future thinks it helpful, but operate as two-way until then.
|Advantages of Couplets||Disadvantages|
|Impressive vehicle capacity gains||Initial confusion for drivers|
|Less pavement = Complete Street space||Out of direction travel (offset by less delay)|
|Shorter signal cycles||Opposition if 2nd street has incompatible uses|
|Safer for autos and pedestrians||Convenience stores may have initial loss.|
|Enhances and motivates TOD||Transit boarding stop different than exit stop|
|Little additional cost if planned from start||Risk of high-speed if designed poorly|
|Expands grid, enabling larger activity center||Change affects many people (education effort)|
|If synchronized, low speed limits observed||Unwarranted stigma to overcome.|
|Narrow streets help pedestrian enclosure|
|Proven: common feature of thriving Centers|
|TownCenterIntersections.org has a collection of existing locations (not comprehensive), along with videos, articles, and other links.
Top-10 Advantages of One-Way Couplets: A short pdf I wrote years ago, with similar points
Top-10 Arguments Against One-Ways, Addressed: Rebuttals to common arguments against couplets.
Please Comment! I like to think I know everything, but I’m just an engineer trying to help broker deals between New Urbanists and Sad Reality, and we need to hear more from both Urbanists and Reality if we’re to discover how these ideas can really help.
If you think this topic is important, please Share!