Bike share is beautiful. Bicycles owned by the public, available to the public at any time for just a couple bucks. It’s a public bicycle transit system operating on a relatively shoestring budget. It’s not a system designed with hardcore cyclists in mind, it’s designed for everyone else.
Or at least it should be. With so much of our coverage of Pronto focusing on problems with management and reasons ridership fell short of projections, let’s look forward to what Pronto could be. Because while Seattle has a unique urban design and geographic challenges, bike share can open up neighborhoods and express transit to many more people if we invest to give it a real chance to succeed.
While there are many changes Seattle can make to help bike share succeed (like building the planned and funded Center City Bike Network or removing our rare adult helmet requirement), the shortest answer for why Pronto is operating over-budget is that it is just too small. With only 54 stations and 500 bikes split into two essentially distinct and even smaller systems, Pronto did not go big enough to pull a profit in its first year. This doesn’t mean we should throw in the towel, it means we should invest the money now that we should have invested at launch.
Both Portland and Vancouver, B.C. have learned from Seattle’s experience by planning bike share systems this year at a more appropriate scale for cities our size: BikeTown in Portland will launch with 1,000 smart bikes and 100 stations (though their bikes can be docked at any bike rack in the service area). Vancouver’s system will launch even bigger with 1,500 bikes at 150 stations.
Pronto, for comparison, has only 500 bikes and 54 stations, but only 42 of those stations form a centralized and connected network. As we have discussed before, that connected network mass is everything:
The good news is that the city’s plan for expansion in 2017 is a good size: Up to 1,500 at up to 150 stations, depending on the bids potential operators propose. So what should such an expansion look like? We’ll dive into that below.
First off, let’s look at the just-announced planned bike share service area in Vancouver, B.C. once their system gets up to 1,500 bikes:
Our rough estimate puts this service area at about ten square miles (depending on how you factor in the water, also an issue in Seattle).
For comparison, here’s what that service area would look like overlaid on Seattle:
Vancouver’s population is much more dense than Seattle’s, and its bike share system will be, as well. Their plan allows them to blanket the service area with stations every two or three blocks. In fact, their station density (15 stations per square mile) will be much higher than successful systems in DC (8.9), Boston (8.3) and Chicago (9.4). Vancouver also has high-quality bike lanes in its downtown and fewer very steep hills than Seattle. So long as the hardware rolls out without major issues (and they handle their helmet law well), Vancouver’s system appears ready to thrive upon launch.
Portland does not yet have a service area map, but it will likely be bigger than Vancouver’s.
To kick off our thought experiment to plan an expanded system in Seattle, let’s start with the whole city. At 84 square miles of land area, it would take a system of at least 500 stations and 50,000 bikes to cover the entire city (the city is aiming for a bare minimum of six stations per square mile). Obviously, we don’t have the funds for that right now (though how cool would that be?).
The map SDOT has been tossing around (with DRAFT in big letters) shows the following service area for a system with up to 1,500 bikes and 150 stations:
By our measure, this is around 20 square miles (again, I’m not sure how best to factor Lake Union and the Ship Canal into the square mileage). With 150 stations, the city’s proposal would clock in somewhere around 7.5 stations per square mile, which is on the low end compared to other successful cities.
The first bit of public feedback we heard was: Where’s Ballard? Well, Ballard is big and far from the rest of the service area. If we’re going to add Ballard, we gotta do it right or we’ll just repeat the mistakes we learned when launching the U District stations.
Note also that the Rainier Valley arm breaks several bike share rules. Most importantly, it is long and skinny, reducing the network reach of each station (top-performing stations are typically in the middle of a dense web of stations in all directions. See the graphs posted above). And perhaps just as importantly, there is no direct and easy bike route between Rainier Valley neighborhoods and the central core of the bike share system (Rainier Ave bike lanes would change this, but they’re not currently in the plans for the near future).
As such, we should plan on ridership being lower than average in this area and budget accordingly. This likely means purposeful public subsidies for ongoing operations of these stations. There are great opportunities in Rainier Valley for connecting more homes and businesses to high schools and high-capacity transit. And having access to bike share offers other public benefits, such as access to a healthy way to get around and making sure city investments are not disproportionately serving wealthier and whiter parts of Seattle.
Public bikes should be for everyone. Like public transit, you don’t just serve the routes that make money, you aim to serve everyone. While a for-profit business might shun providing service in Rainier Valley, as a city we can make the choice to serve the area even if it means taking a financial loss on station operations.
So with a basic understanding of the factors that make bike share work best (density of destinations, topography and popular bike routes) and maintaining the desire for a Rainier Valley launch, we thought we’d take our own stab at what an expanded bike share system in Seattle could look like. First, here’s our overly ambitious map (the circles are existing stations):
This map serves population centers and popular longer-distance recreation routes (like Lake Washington Blvd and Golden Gardens). It also includes the city’s Northgate expansion and the popular desire to reach Ballard. I used major bike route barriers like water, steep ridges and bridges to influence the service area. And I got rid of Queen Anne Hill from the city’s map because it’s just so steep (especially on bike share bikes) that it is essentially isolated from the rest of the station network.
Unfortunately, this map clocks in around 30 square miles, or only five stations per square mile. Without more money for more stations (we would need at least 50 more than planned), we’re going to need to trim it down.
Our second try turned out like this:
The first thing we cut are those long and mostly-disconnected waterside routes like Lake Washington Blvd and the route to Golden Gardens. Instead, we focus on areas with homes and commercial centers within easy bike rides of each other. The biggest challenge to adding Ballard is the mega-steep Phinney Ridge. You pretty much need to extend the Ballard service area north to 85th in order to connect it with the rest of the system (bonus points for serving the Greenwood business district).
To make up for some of this additional square mileage, we leave Northgate for a future expansion. Certainly by the time Northgate station opens there should be bike share service. But that is years away.
This map clocks in around 21 square miles, just a touch bigger than SDOT’s proposal. At 7.1 stations per square mile, the station density would still be above the city’s minimum of 6, but it would be lower than key peer cities that have proven successful.
Or if we want to focus on station density and connectivity, we could leave both Ballard and Northgate (as well as Sodo) for future expansions like so:
This map clocks in around 17 square miles for a healthy station density of 8.7 stations per square mile. The north section uses the steep Phinney Ridge as its natural western border.
Seattle is not an easy city for planning a bike share system. It’s got a lot of geographic challenges and oddly-shaped neighborhoods that resist a simple network of public bike stations.
But those same challenges also create the unique public spaces and magnificent views that make Seattle such a rewarding place to bike. And though biking between neighborhoods can sometimes be tough, that’s where the potential for pairing transit and public bikes really starts to make sense.
For example, the Rainier Valley system will probably operate as a way to connect transit and destinations on the two main busy streets: MLK and Rainier. In much of the Valley, it’s a ten-minute walk between the two streets, but a three-minute bike ride. It’s a 20-minute walk from the nearest light rail station to the Hillman City business core, but a six-minute bike ride.
Likewise, bike share on Beacon Hill will connect Beacon Hill Station to many other Beacon Hill businesses and homes. And Beacon Hill bike share stations will have reasonable connectivity to the central system stations.
North of the Ship Canal, the bike share stations would have great access to both Aurora and I-5 express buses, bringing so many businesses, homes and other destinations into reach of those fast bus routes. And, of course, the system will connect with UW Link Station and the under-construction U District and Roosevelt Stations. And by connecting the Fremont, University and Montlake Bridges, there will be many more connectivity options for users that greatly expand the reach of taking bike share.
150 stations and 1,500 bikes is not big enough to get everywhere we want to go. But it is big enough to make a much more usable and powerful network. And that’s what Seattle needs to realize the bike share vision we are so close to reaching.
I know people are frustrated by Pronto’s budget troubles and the city’s handling of the system buyout process. But don’t lose sight of what we could gain if we stabilize and expand this public bike system to the size it should be.
Comments
35 responses to “How Pronto can become a beautiful public bike system by going bigger”
Heck of a huge subsidy for magnificent views which one can get in foot!
“But those same challenges also create the unique public spaces and magnificent views that make Seattle such a rewarding place to bike. “
NACTO drew an exponential line instead of linear and gets a lot of mileage out of that. Look at those scatterplots for yourself….. Any system that is bigger will have more use. And more cost.
The y-axis is rides *per station*, not rides. Obviously total cost increases with the number of stations, but it’s not clear at all to me that cost *per station* increases with the station number — I’d guess it stays flat or falls a bit due to economies of scale with fixed costs getting divided across more stations.
See the scatter plot. The relationship is not exponential.
You may be right about the curve (I’m no statistician), but either way there is a strong upward relationship. That’s the main point.
Conversely, David Hembrow points out that one of the big obstacles towards growing cycling in London is that the ongoing subsidies to their bike share system is eating a huge chunk of the money that would have been spent on all ages & abilities infra.
Here we have a second chance at a missed opportunity, I remember back when a Seattle bike share was first being discussed there seemed to be some possibility that Vancouver would have one first, and that Seattle could learn something from them. Of course Vancouver dogged a bullet when Bixi went under and Vancouver got a chance to learn from Seattle.
If we let Pronto go and watch Vancouver for a few years (working on our infrastructure in the mean time) we could be in a better position to make reasonable decisions regarding bike share. Note that both Portland and Vancouver are using different technology than Pronto, so I don’t think even mothballing the existing equipment would necessarily be worthwhile. Also note that it seems that Vancouver has adopted the Melbourne helmet model, provide free helmets but don’t bother with the expense of cleaning them, avoiding that expense makes it easier to provide them for free, and not charging for helmets reduces the cost disincentive for the use of the PBS, though the cooties factor may trump that.
Remember back in 2014 you wrote:
“But lest you point to Melbourne’s mostly failed system as evidence that it cannot work, Pronto is going well above and beyond Melbourne’s efforts by installing a helmet vending machine at all 50 stations”
It will be interesting to see how Vancouver does, they will have the station density needed, there will largely be that one variable between them and “successful” bike shares.
There is still my idea of requiring helmets for pedestrians (including walking to your car) , so that everyone will already have their own helmet.
Agree. Work on bicycle infrastructure for a few years. Vital to improve this transportation experience for present Seattle cyclists
I like number three, 17 square miles and 8.7 stations per square mile. Combining public transportation and bike trips has long proved itself to be the most efficient way to get around this traffic plagued city. But there just isn’t enough capacity for bikes on board our buses, light rail and street cars to meet the bike mode demand that we are seeing and would like to see. This is becoming evident with the opening of two new stations on Link and the concerns for bike capacity with the anticpated ridership increases.
I’ve given up hope for West Seattle to get bike share in my lifetime. But that doesn’t mean bike share won’t be very useful to me if I can get to the bike share service area by Metro Bus/Rapid Ride/Water Taxi. We need more personal bike storage options near rapid transit in the outlying neighborhoods that won’t be getting bike share, combined with a healthy network of bike share stations in the central part of Seattle. Because not everybody has the desire, ability, or free time to do the whole trip on their personal bike.
Testimony at the City Council meeting on February 19th was overwhelmingly in favor of the city taking over Pronto. Interestingly, a large number of Pronto supporters were older women like me. Maybe we get the fact that even SLOW riding is more efficient than walking downtown. With 7 speeds, the bias against uphill Pronto trips is mostly psychological. Pronto is promoting the “Seattle Squiggles” using the gentlest routes to get up the steeper hills without having to dismount: https://www.prontocycleshare.com/squiggles . Even if you have to dismount for a block, so what? It still beats walking all the way up.
Expansion of bike share in Seattle will be the fastest way to increase the bike mode and justification for increased investment in bike infrastructure will naturally follow.
Well said Kathy, though I do think after the current proposed expansion there will be a lot more momentum for additional expansions say along the RapidRide routes that could happen before too long. The last couple of years have shown bike share networks that have 100 docks tend to start growing a lot faster, which is probably why Portland and Vancouver are starting out at that size.
Thanks, Tom, for an excellent post. I agree with most of what you say.
Nevertheless, I will once again offer up an alternative view for station placement.
What if the reason people haven’t embraced Pronto as currently deployed is because some of the fundamental bikeshare hypotheses, while appropriate for other cities, don’t hold in Seattle? I’m particularly talking about the focus on the center city and a grid arrangement.
I would describe Pronto’s original station placement priorities as follows:
1) center city
2) density
3) grid arrangement
4) U District
5) tourist destinations
…
11) proximity to light rail or BRT
12) proximity to low traffic, low speed, low gradient routes
13) avoid hills
These priorities would lead to the current arrangement of stations.
My thought experiment is the following:
What sort of arrangement of stations would we arrive at if our priorities were as follows:
1) proximity to low traffic, low speed, low gradient routes
2) proximity to light rail or BRT
3) avoid hills
4) density
5) tourist destinations
6) equity
7) center city
Gone is the a priori requirement for a grid arrangement and for uphill stations in the U District. (Though some downtown or U District stations will satisfy other priorities.)
Using this new set of priorities we would proceed as follows:
1) identify low traffic, low speed, low gradient routes: all the separated bike trails and protected bike lanes at <5% grade
2) identify where routes intersect light rail and BRT stops
3) identify where routes are close to live/work density
4) identify where routes are close to tourist destinations
This approach recognizes the following realities about Seattle:
* Seattle is geographically and topographically constrained so that route choice is more important here than in other cities with connected grids of streets. This is especially true for cyclists.
* Many people don't feel safe biking downtown.
* We have some of the best urban bike trails in the country — why not use them?
Given that bikeshare is supposed to appeal to "casual" users, I think this approach would result in a more popular system than the one we currently have.
I think this makes a whole lot of sense. For the North end that could mean staying away from Greenlake/Greenwood and instead expand through Fremont, Frelard, Ballard. This takes advantage of the Burke, a higher density commercial corridor. and matching intersection with the N/S bus routes.
The UW stations are not convenient enough for people to use them within the campus, but there are a lot of people coming to and from UW from the lower Ballard Fremont corridor.
I had an option that sort of did this, making 50th the northern limit and spreading out along the Ship Canal to the Ballard business district. But it ended up being very skinny along the Ship Canal. Maybe there’s a better option. You can use Google’s My Maps feature to draw a box that will give you square ft. Give it a shot.
Here’s an example to start: https://www.google.com/maps/d/edit?mid=z0xXa0ZjM3qM.kiClgWjHJTRU&usp=sharing
I believe you can “make a copy” to make your changes (click the three dots to pull up the menu, you’ll need to be logged into google)
Also, I included Greenlake not because of the lake, but because it’s so bikeable once you’re up the climb to 50th, anyway. There are lots of easy bike routes and transit connections (like the Green Lake Park and Ride, which has a horrible walk zone but a great bike zone).
As for Greenwood, it makes sense that if you’re going to include N 45th Street in Wallingford, you should include the northern part of Fremont, too. And if you’re going to go all the way up the hill on Fremont, you might as well include Phinney Ridge businesses and the zoo. But Phinney Ridge is so poorly connected east/west that you might as well keep going north to Greenwood, where you have more destinations and good connections to Green Lake (and Ravenna Blvd and beyond). And that’s how that map happened. My goal wasn’t to get to Greenwood, it just sort of makes sense if you’re trying to create a usable and connected network.
Really, I prefer the map I have above that includes Ballard and Greenwood. It would take some extra stations to do it. Maybe one of the companies that bids on the contract will notice the need, too, and include a way to make it work in their pitch…
I think there is a lot of merit to what you are saying, but I think you quickly get close to what Tom has made anyway. For example, I think there is too much focus on the idea of downtown being the main area we will serve. I agree that I would not want to ride a Pronto bike in the heart of downtown, but there are certainly areas where it makes a lot of sense. Pioneer Square, the I. D. and South Lake Union, for example. The Denny regrade is a bit challenging (because of traffic) but great in terms of grade. With enough stations, “greater downtown” will have plenty of riders.
Meanwhile, the areas east of downtown are very good. Yes, getting from downtown to First Hill or Capitol Hill is a challenge, but once you are on First Hill (or Capitol Hill) getting to other parts of the east side of the city is a lot easier. A ride from Seattle U to Seattle Community College makes a lot of sense on a bike, and will make a lot more sense once Link adds a station. The same with Cherry Hill, and the rest of the Central Area. So I think unless you simply avoid everything south of I-90, what Tom has made south of the ship canal makes a lot of sense.
I could definitely see skipping the upland parts of the area north of the ship canal — favoring the Burke Gilman — but you quickly start robbing Peter to pay Paul. To being with, the UW and the U-District is a given. It is pretty hard to include the bottom of the UW and not the top. It is also silly to include the lower businesses of the U-District, but not the upper ones. So you definitely extend this to the at least 55th east of I-5 (and more likely 65th).
You can cut out some of the highlands doing that. Here are a couple of very scaled down versions, one of which is really cut to the bone. Compare these to Tom’s smaller map (https://www.google.com/maps/d/viewer?mid=z0xXa0ZjM3qM.kiClgWjHJTRU). Here is the first one:
https://www.google.com/maps/d/edit?mid=z-ZcpzpzqRA0.kTe2cJ5IFa78&usp=sharing
I’ve abandoned Greenwood and much of Ballard. These are areas that make sense for bike share, in that they are fairly flat and have plenty of destinations. So this might be better (since Ballard and Frelard are obviously good) but it looks to be bigger (from what I can tell).
But this can be cut even further. Phinney Ridge is, like the top of Queen Anne, very difficult to access from the south. But unlike Queen Anne, it isn’t that hard to access from the sides. This means that if you remove it, you start isolating other areas, or you might as well cut them out. I did the latter here:
https://www.google.com/maps/d/edit?mid=z-ZcpzpzqRA0.ktX20UvURO_Y&usp=sharing
I think this map has some merit. While there are plenty of very good places that are gone (Greenwood, the very dense area north of Green Lake), you save a lot of space (I think). I used I-5 as the natural border.
It is pretty hard to say this is much better than Tom’s small map. Maybe, but it is starting to get fairly disjointed. I really have a hard time seeing any of my maps as being clearly better than Tom’s (in terms of efficiency).
It’s harder than it looks to make a good map!
In reality, a more nuanced and professional approach can probably figure out the best way to deal with all these differences (where to focus density, where density can be lower, etc).
I might be off the mark here, but I wonder if there is anything that can be learned from the huge success of Car2go in Seattle. Of course, it’s not an apples-to-apples comparison but I do think they both have a similar place in the market. Although Seattle has a very high rate of car ownership, Car2Go has fulfilled a demand for one way car trips, avoiding parking fees, increased flexibility and I am sure many other reasons. I am a big advocate for Pronto because it could also fulfill needs that are not currently being serviced. For me personally, I find myself conveniently taking the bus to work in the morning but then being unable to take it home because of inconsistency in my work schedule and the poor frequency of the bus service. I would gladly use Pronto for those trips where a bus is just not possible. I wish the city would do a better job at evaluating what Pronto could be used for by someone that already owns a bike. Like we have seen from the U District, putting bike stations in an area just because there is a already a high share of bike commuters is not a recipe for success.
Exactly.
I frequently work remotely on the Silicon Valley campus of a well-known tech company where EVERYONE uses bike share, and each bike probably turns over dozens of times a day.
They don’t just place bike share where people already ride bikes — they place it to solve currently unmet trip needs for everyone (where it’s a 3-7 minute bike ride, where the infra is safe enough that everyone can feel safe riding.)
Just to be sure, nearly none of those people ride bikes when they return to Seattle, nor would they ever think of riding them here. Bike share or not.
We have sidewalks everywhere in Downtown/Central Seattle. It may not be safe for the pedestrians to have to share with sometimes inconsiderate cyclists, but it is certainly a safe alternative for cyclists. How is it not safe to bike (slowly and carefully, which is what the Pronto bikes are designed for) downtown? If you are a fast rider, then, it is probably not safe for you down there. But that is true no matter where you ride. Speed kills. The perception that it is unsafe to ride bikes in Seattle is just that, a perception. Blown way out of proportion.
Allowing riding on sidewalks would be a disaster. Shared use trails (which the sidewalks would become) are quite dangerous for people biking and terribly uncomfortable for those walking. Downtown is busy and congested. The answer is more protected bike lanes, not forcing people walking to share space with bikes.
Riding bicycles on sidewalks is already allowed throughout the city.
I’ve lived and worked in Vancouver BC, and much of their strategy is placement of stations for tourists and respective desired destinations. Yes, this will be available to locals (duh), but the city wants tourism to drive the usage, not locals. I think Seattle is noble in wanting to provide rideshare for “locals”, but in fact, unless tourist destinations are captured, that will likely not be the case.
The goals that Vancouver’s system lists here are:
http://vancouver.ca/streets-transportation/public-bike-share-system.aspx
-Extend the reach of transit and walking trips
-Reduce the need for personal vehicle trips
-Trigger greater interest in cycling, and increase cycling ridership
I’m sure there will be destinations that serve both tourists and residents, but they’ve noted that most bike share systems get about 20% tourist use. Commuters and short errands will drive the usage in this kind of system.
Vancouver’s also smart in not relying on a “soak the tourists” strategy to break even (which can discourage casual users from even trying the system). It looks like they’re treating bike share as an investment in transit.
http://www.theprovince.com/travel/vancouver+bike+share+program+about+investment+being+cash+city+staff/11741961/story.html
[…] on the Network today: Seattle Bike Blog says the city should fix its struggling Pronto bike-share system by expanding it. Bike Portland […]
I really wish Seattle would take the lead and abolish its helmet law. It would make it much easier for British Columbia to follow suit, and ensure the success of Vancouver’s upcoming bike share.
Vice versa! :-)
And versa vice or something.
It’s going to take a lot to get rid of the provincial helmet law I think. It’s the right thing to do of course but it won’t be easy. The municipal law won’t be so difficult I predict. (And I really don’t know to be honest…)
There are choices to be made and funding more into this system, at this time is not a good choice.
The analysis and operations of the existing bike share are flawed and biased. The director of SDOT has personal ties to the program, ties which go beyond simple professional support. His inappropriate payment of over $300,000 into the failed operations should be called out. Projections which were way off the mark should not be simply excused away. This is not a completely new idea, never done before and totally experimental. It has been done, it has both failed and succeeded elsewhere and the promoters in Seattle created unrealistic expectations, desperate to sell the idea. They in-turn are not the people who can or should be trusted to establish a successful program.That money along with the original investment is now gone and the city is no better for it. The promoted concept is bankrupt. The faster it is stopped, the sooner we can get on with more demanding priorities. Let it go and revisit the idea with truthful and achievable expectations in the future.
Are there any examples of successful bike share systems in cities with helmet laws? If so, I’d love to see them. Mexico City repealed their helmet law before bike share started, and they have a fantastically successful system. The two cities I know if that have helmet laws and bike share (Melbourne & Seattle) have both been seen as failures.
[…] Better expansion: Seattle Bike Blog examines how Seattle’s bikeshare could be improved through targeted expansions. […]
[…] Bikeshare networks depend on station density. Best for short point-to-point rides, it is absolutely critical to minimize out-of-direction travel to return a bike. Tom from Seattle Bike Blog has already done some great analysis on station density, and I won’t repeat his good work. Go read it. […]
[…] I apologize in advance, but I’m going to use a baseball metaphor. Ditching Pronto now would be like laying down a bunt on the first pitch that goes foul. Then instead of trying again, you quit because you didn’t hit a home run. As we have reported, Pronto’s limited size and split system was doomed to come up short. The answer isn’t to give up, we need to give it the effort it needed from the start. […]
[…] Bike Blog recently took a look at different expansion configurations to show how the Pronto could transform into a robust public bike system by going […]
Cool, Older people Cycling at all good!
[…] the system, more employers join the corporate membership program, and more sponsors sign on. And as we’ve reported previously, revenue could become a lot healthier with more stations reaching more […]