Bikes London

Can Bike Sharing Work in Cities With Monofunctional Job Centers?

» London’s experience may provide a useful example for American cities looking to introduce large bike sharing systems.

Bike sharing is growing rapidly as the transportation mode du jour; not only have the standardized bikes and their docking stations invaded most major cities across Europe, but they’re now headed towards introduction in a number of American cities as well. Before investing full-scale in the purchase of thousands of new bikes and the installation of hundreds of docks, U.S. planners should be looking closely at previous experience to determine best practices in system design.

Last month, I laid out my concerns that Washington, D.C.’s new Capital Bikeshare doesn’t plot its stations close enough together for the system to be effective, at least based on the manner in which Montréal and Paris have implemented their networks. The lack of station density could prevent easy use by day-to-day users because of difficulties related to finding stations in some neighborhoods.

London, which just introduced its Barclays Cycle Hire system using 6,000 Montréal Bixi bikes and 400 docking stations spread out across 17 square miles of the center city, does not have the same problem, since its stations are tightly packed in a circumscribed area. One difficulty it might have, however, could potentially be even more problematic: Because of London’s land use geography, commuting patterns are overwhelmingly unidirectional, towards the center in the mornings and away from it in the afternoons. This may put a strain on bike sharing, since to work, the concept requires a relatively even pattern of bike pick-ups and drop-offs at every station.

American cities, which feature similar concentrations of office jobs in the inner-city core and distributions of residential areas in peripheral zones, must evaluate how London is handling this problem and develop their own coping techniques before moving forward with a major spending program.

Consider the images below of usage distribution of London’s bike share, products of a mapping system developed by Oliver O’Brien. In the mornings, thousands of people bike from the outside of the Cycle Hire zone into its interior; by the afternoon, this produces a situation in which the majority of stations in the center are full (red) and the majority of those along the edge are empty (blue). In the evening, on the other hand, the movement of commuters from the core and into the periphery produces the opposite situation, where the stations in the center are empty and those on the periphery are full.

Afternoon – 1:35 PM London Time
Evening – 8:55 PM London Time
Red dot: full station | Blue dot: empty station

For commuters intending to use the bikes during off-hours, this is extremely problematic. If you want to ride from the London jobs center to the outside of the Cycle Hire zone at 9 PM, for instance, it may be virtually impossible to find a bike; even if you do, you might have a difficult time finding a station at which to dock your bike. The same can be said for a commuter attempting to make the reverse commute at 11 AM.

Perhaps more important, this situation is difficult to handle from an organizational standpoint. Because of the fact that the managers of the system want to alleviate these problems, they have 14 trucks (one of which is pictured at the top) which transfer bikes from full stations to empty ones. Other cities with bike sharing have a similar transportation method, but London’s may be particularly overcharged because of the monofunctionality of many of the city’s neighborhoods.

The worse-case situation seems to be occurring at the bike share docks adjacent to the Kings Cross and Waterloo intercity stations. There, the Cycle Hire management company Serco is simply leaving dozens of non-docked bikes in front of full stations, cluttering up the sidewalks sometimes for hours in anticipation of them being moved elsewhere. There are a few solutions that could be implemented relatively easily, including the hiring of more trucks to move bikes around and the creation of more docking points at places with heavy demand for parking.

But both of these would require a ramp-up in operations costs. One of the great benefits of a well-designed bike sharing system is that the riders can do the moving for you, thereby reducing the onus on the operator to make sure there are an adequate number both of bikes and of empty docks at every station.

Some cities, like Paris and Barcelona, have it a bit more easy, simply because office and residential uses in those cities are not nearly as segregated as they are in London, making the flow of bikes in the sharing system multidirectional. In other words, a mixed-use city is most appropriate for the implementation of a bike share system. It is indicative that the one place in Paris where there is a massive concentration of jobs but few residences — at La Défense, just outside of the city limits– has virtually no access to the Vélib bike sharing network. The city’s planners likely understood that the result of putting docks there would be the same problems as are now experienced by London, and have resisted expanding the system into that business district.

But most American cities have no choice but to include their primary, monofunctional, business districts in their bike sharing plans simply because those business districts are in the center of the city. It will be interesting to watch Washington, D.C. and other cities attempt to cope with the problem of the unidirectional commute as their inhabitants get used to biking to and from work, but London’s experience makes clear what they’re likely to experience.

Images above: London Cycle Hire bikes being moved about the city, by Flickr user Tom Anderson (cc); and Status of London’s Cycle Hire stations (17 and 18 August 2010), from (cc)

24 replies on “Can Bike Sharing Work in Cities With Monofunctional Job Centers?”

This brings to mind the Japanese system of undergound bike parking. Even something less technically intricate, like a two-level system that puts the top, full tier up when someone wants to find an empty space to park, and brings the full top tier down when the bottom tier is empty and someone wants a bike, would double the buffer capacity at high use, highly directional stations.

But the long term solution would seem to be “stop doing that” … development of multiple use and a more diverse pattern of use across the day through activity specific block zoning that spills the dominant activity across to neighboring blocks and retains space for complementary activities has benefits across the board, and not just in a better fit to bike sharing schemes.

It occurs to me that a two-level system would also make it less time-consuming to move bikes around, if the top level can come off and slot into a flat bed trailer. For things to be self-organizing, there’d be target locations spread around at multi-modal transit interchanges where you are guaranteed a bike and a space, and people needing a City of London bike at 9pm or an outer ring bike at noon would know to go to those hubs.

Another solution (and the best one, to my economist’s eye): London should vary the pricing of bikes at different stations at different times of day based on demand. If you wanted a bike at King’s Cross at 8:30 am, you might find that the price is twice as high as it would be at other stations – and decide to walk a bit to where there are more bikes available and prices are lower.

For those people who really need a bike and are willing to pay extra for it at a high-demand station, they would be more likely to find one.

This is, of course, the same principle that San Francisco is trying to apply to on-street car parking spaces.

There will have to be some sort of in-built extra capacity and re-distribution (during mid-day or overnight).

As much as people like to think that Montreal is European, it has a very North American style downtown with a high concentration of employment relative even to the surrounding neighbourhoods and certainly to the Bixi territory. As far as I know, this hasn’t been a huge issue there, though I hear that they do re-distribute the bikes throughout the day to even things out.

When a downtown/central area is large enough, it would seem that this is less of a problem; hopefully everyone’s ‘last miles’ overlap in multiple directions and bike usage/parking evens out spontaneously. You could have issues if there are really one or two very dense blocks, with little nearby, but my feeling is that there would have to be a very drastic drop-off in employment/activity for this to be problematic.

I like C Neal’s idea, but it seems like a lot of math to do while on the fly (and possibly with a fixed deadline/appointment approaching); how would I know how far to walk? Just keep going until I find a station with the ‘right’ price? Interesting…

Smartphone apps that tell you in real time which stations are full and which have empty spots are common for existing bikeshare networks. Adding price info would be pretty trivial. For non-smartphone users, each station could include a display of nearby stations and their inventory/price levels.

Sounds like London just needs more bikes. If they are using all the bikes each way, they should just keeping adding bikes until they have surplus bikes at the origin. They would also need to add more bikes parking spots at the destination.

Although I applaud blunt-force economics a la congestion pricing, I would amend C Neal’s proposal slightly: instead of charging by station and time, perhaps we could charge by the number of bikes available at any given station anytime of the day- that way we avoid rigid and confusing zones and fare tables, and don’t penalize users traveling from periphery station to periphery station during the peak hours.

So, for instance, what if we charged a 25¢ fee for taking one of the last bikes from any station that is 90% empty (the last bike in a station with 10 docks) AND for depositing a bike in a station that’s 90% full, but give a 25¢ credit to anyone depositing a bike in a station that is 90% empty and for taking a bike from a station that is 90% full. That way we are a) rewarding folks who voluntarily restock the system, and b) charging for peak-hour AND peak-direction travel. This would also be uniform throughout the system, so all regular customers would know the rule of 90s, or whatever, and wouldn’t have to learn any particular times/locations.

So, if I take a bike from Dupont (in DC) at 8:30 Monday morning and deposit it downtown, I get charged 50¢. If there is another station downtown that isn’t usually full, I can drop it off there and only get charged 25¢. On the other hand, if I pick up a bike in Dupont at 8:30 Monday morning and drop it off in U Street, I get charged 25¢ and credited for 25¢ (effectively gaining back the fee for taking the bike from Dupont). If a group of enterprising teenagers take bikes uphill after school everyday, they can even earn some quick cash for their public service.

That may be too complicated, and I can imagine the fights that could ensue over the 2nd-to-last bike, and we aren’t even sure if this will be a problem in DC, but who knows?

Bikes and bike racks are relatively cheap. The infrastructure to implement variable pricing may cost more than having more bikes and bike racks.

I was thinking along those same lines, of charging based on the number of bikes available at a station. The thing that could make it work is open data on real-time availability at the various stations. People could then build apps to provide expectations of cost, and to highlight trips that pay. All you’d have to do is make the rewards high enough for the system to be self-balancing, which I suspect is still less expensive than trucking bikes all over the place.

By analogy with Donald Shoup’s recommendations for on-street parking pricing, it would make sense to aim for a certain proportion or absolute number of bikes being available at each station.

In Montreal there is the Mountain, which means that bikes, like water, go downhill and that they need to be continously redistributed up the mountain.

The very same thing happens in Paris. People ride downhill from Menilmontant and other locales in the east and north, and then they take the Metro back up the hill. Bikes just have to be trucked back up, but that’s just the price of doing business.

As far as La Defense not having Velib access, it also has hardly any bike access. La Defense is built for transit and for automobiles. I once biked from Paris proper to La Defense, and it was harrowing to say the least. Maybe things have improved over the last two years.

I’ve heard an excellent method to attempt countering that: let people take a bike for free, or even credit them, if they are taking it up a hill. Extra points for taking it from an employment center to a residential area at the top of a hill.

Are these bike sharing systems recovering the all the costs (capital and operations) through subscriptions and advertising? Covering just the operating costs? Are they near covering operating costs or far from covering operating costs? I realize there is a range of systems around the world but I’m looking just to get a ballpark idea.

They certainly make sense in large world class cities but what are the thoughts about bike sharing in small walkable cities that dont have much of a transit system, and with maybe a young somewhat hipster population, a college town, a touristy town… to throw some cities out say, Santa Barbara CA, Newport RI, Ann Arbor MI, Dubuque IA, Burlington VT, Boulder CO, Santa Fe NM, Savannah GA, etc.

The Paris bike share was originally to be entirely funded by the advertiser, but with high rates of vandalism, the city now pays $500 per bike replaced due to vandalism.

The Paris system also has a bonus for returning a bike to a station higher than 60m in altitude, in the form of 15 free minutes, so the next free rental period to subscribers extends from 30 minutes to 45 minutes if they return a bike to an “uphill” station.

This article is interesting, and it underlines the fact that bike sharing isn’t really efficient for coping large flows…

Anyway, the reason why there is no Velib stations in La Défense, in Paris, is quite different from what you think.

The Velib’ system has been planned and financed by the Ville de Paris, which only governs the inner city, and not La Défense.

The Velib system has been authorized to spread to neighbours town shortly after the first implementation, but La Defense is not even a neighbour of the Ville de Paris, since the town of Neuilly is located inbetween…

This is just another proof of the need for a Greater Paris, which feeds quite a number of debates here in France…

Anyway, and regarding the subject of your article, you may see, during your next trip here, that the problem you are pointing at is happening in Paris too.
Indeed, the velib stations in Paris real Central Business District (where most of the banks are), located roughly between the Opera and the Champs Elysées, are fully packed in the morning, this being the case as well in other business districts, like the surroundings of the Gare de Lyon.

One thing to consider about the redistribution efficiency in London is that, at present, only registered users can use the bikes. These early adopters are undoubtedly almost all commuters, which distorts the model. Once the system is thrown open to casual users, the impact of the one-way commute should be mitigated by tourists, shoppers, etc who will have more dispersed travel patterns. We’ll need to wait until then before we can really evaluate the system.

London’s bikeshare still isn’t open to casual users (but should be from next month), yet they’ve already announced a major expansion. By the 2012 Olympics the area covered will increase by 50% and the total number of bikes in the scheme will be around 8000, according to the press release.

Completely OT but much more exciting is the idea of a gondola over the Thames. [returning to topic] But I wonder if you can bring a bike aboard…

Great analysis Yonah.

It’s amazing to see how fast patterns are emerging in bike share systems that you can’t necessarily predict before you put the system in. Luckily, unilke, say, a multi billion dollar metro project, it’s pretty damn easy to add more racks, move them around, add capacity, etc etc. Flexibility by design makes problems like this interesting, instead of devastating.

This problem is no different than car parking is for city centers. While bikes are definitely more space efficient than cars, there is nothing more space efficient than a vehicle that drops you off…aka public transit.

Having more distributed urban structures could help to ease the problem, but transit is a better solution. Not saying it should be done away with, and if there is space, then increasing rack size is an apt solution…but transit is still a better solution for the transportation problem.

Mass transit will always outperform individual transit (including bikes; walking doesn’t count) at moving a large number of people to one location. Tying bike sharing closer to transit could improve the distribution, by keeping bikes within neighborhoods and encouraging transit for downtown trips: Montreal offers pretty substantial discounts to people holding passes for both transit & bike share.

DC has a geographically dispersed CBD by American standards, but even with the limited SmartBikeDC pilot system I’ve found myself searching around for an open dock downtown during the day. Minneapolis’ Nice Ride launched with plenty of excess capacity on the UMn campus (America’s 4th largest), but it’ll be interesting to see what happens once classes start in September.

A few bike-share systems at present do provide incentives for riders to ferry the bike themselves — the Paris uphill credit is the widest known, but a recently launched system in China (I think in suburban Beijing, will check my notes from Velo-City) also offers users credit for redistributing bikes away from peak flows.

Why not make the bike stalls so that they are on wheels, making redistribution easier? Perhaps a fifth wheel type flat bed? One worker with a pick up and a cell phone could even things out even in a large Metro area.

Given its configuration, in Barcelona there is a similar problem. Most people use the bycicles to go from the hilly neighborhoods to the beach, leaving uptown stations without bycicles and having problems to leave the bycicle downtown. What is worst, most people prefer to return by subway or bus, so that means that several times during a day vans must take the bycicles uptown.

One simple way of avoiding this problem could be giving credit (for instance 30 or 50 cents), for those journeys departing from a station downtown (for instance, those close to the beaches)and ending in stations located above 50m over the sea level. This credit will be reedemable when you’ll have to pay your next annual fee.

BTW, regarding costs and revenues, the City Council has to pump every year circa €7 million for the system. Part of the reason is that the the annual fee (36€ per year) it’s extremely cheap for people using the system on a regular basis.

Bike sharing should never be considered an alternative to transit. It’s a compliment to transit. In fact, in Paris, my ‘Velib’ pass and my RATP ‘Navigo’ card are one and the same. Bike sharing works to enhance the transit options that are available. For example, Metro Line 14 is the fastest way to get thru the heart of Paris; but it’s southern terminus, “Olympiads” is too far from my apartment for me to walk to routinely. Therefore Line 14 by itself is of little use to me. However “Velib” puts Olympiads Station within easy reach, and thereby makes Line 14 accessible to me when otherwise I would have to make several Metro connections to get me where Line 14 gets me to easily. The same can be said for my other transit situations in other directions. Bike sharing, when done well is completely transformative.

Leave a Reply