2014 itdp bikeshare planning guide

8/20/2019 2014 ITDP Bikeshare Planning Guide

1/152

1Introduction Sub

THE BIKE-

SHAREPLANNINGGUIDE


2/152

2Introduction Sub


3/152

3Introduction Sub

THE BIKE-SHAREPLANNINGGUIDE


4/152

4Introduction Sub

9 East 19th Street, 7th Floor, New York, NY, 10003tel +1 212 629 8001www.itdp.org

The Bike-share Planning GuideCover Photo: Mexico City's Ecobici has helped to increase

cycling mode share in Mexico City.Cover Photo By: Udayalaksmanakartiyasa Halim


5/152

5Introduction Sub

Authors and AcknowledgementsThe writing of this report was a collaborative effort across ITDP andour partners. Contributing authors include: Aimee Gauthier, ColinHughes, Christopher Kost, Shanshan Li, Clarisse Linke, StephanieLotshaw, Jacob Mason, Carlosfelipe Pardo, Clara Rasore, BradleySchroeder, and Xavier Treviño. The authors would also like to thankChristopher Van Eyken, Jemilah Magnusson, and Gabriel Lewensteinfor their support in the creation of the guide.

ITDP is especially grateful to the following people for providingcomments on and contributions to sections of this report:

Alison Cohen, Director of Bike Share Services, Toole Design Group(with many thanks to Shomik Mehndiratta and the World Bank fortheir support of Ms. Cohen’s research)

Dani Simons, Director of Marketing, NYC Bike Share

Matteo Martignoni, International Human Powered Vehicle Associationand former ITDP board member

Jeff Olson, Alta Planning and Design

Chris Holben, former Project Manager for Capital Bikeshare DistrictDepartment of Transportation.


6/152

6Introduction Sub

1 INTRODUCTION 81.1 The Benets of Bike-share 141.2 History of Bike-share 191.3 New Developments and Trends 251.4 Building Political Will 261.5 Elements of Bike-share 27

2 THE PLANNING PROCESS 28 AND FEASIBILITY STUDY

2.1 Overview of Planning Process 302.2 Feasibility Study 322.3 Bike-Share Metrics 40

2.3.1 Basic Context Data and System Metrics 402.3.2 Performance Metrics 41

2.4 Coverage Area 43

2.5 System Sizing: Three Basic 44 Planning Parameters2.6 Financial Analysis 48

3 DETAILED PLANNING AND DESIGN 523.1 Station Location 573.2 Station Sizing 633.3 Station Type and Design 64

3.3.1 Manual vs. Automated 65 3.3.2 Modular vs. Permanent 68 3.3.3 Docking Styles 71

3.4 Information Technology Systems 74 and Payment Mechanisms3.5 Bikes 763.6 Marketing 82

3.6.1 System Identity 83 3.6.2 Internal Marketing 83 3.6.3 External Marketing 83

4 BUSINESS MODEL 864.1 Organizational Structure 90

4.1.1 Implementing Agency 90

4.1.2 Operator 914.2 Asset Ownership 944.3 Contracting Structure 95

4.3.1 Publicly Owned and Operated 97 4.3.2 Publicly Owned and Privately Operated 97 4.3.3 Privately Owned and Operated 98 4.3.4 Types of Operators 101

4.4 Managing Contracts Through Service Levels 102

Contents


7/152

7Introduction Sub

5 FINANCIAL MODEL 1065.1 Capital Costs and Financing 109

5.1.1 Bicycles 110 5.1.2 Stations 110 5.1.3 Software 111 5.1.4 Control Center, Depot, and Maintenance 112 and Redistribution Units

5.2 Operating Costs 114 5.2.1 Stafng 115 5.2.2 Redistribution 116 5.2.3 Maintenance 117 5.2.4 Control and Customer Service Center 118 5.2.5 Marketing and Customer Information 119 5.2.6 Insurance (Anti-Theft, Accidents, Vandalism) 120

5.3 Revenue Streams 125

5.3.1 Government Funding 126 5.3.2 Loan Financing 126 5.3.3 Sponsorship 127 5.3.4 Private Investment 127 5.3.5 User Fees 127 5.3.6 Advertising Revenue 129

6 IMPLEMENTATION 132

7 CONCLUSION 138

APPENDIX A:Key Resources and Publications 144

APPENDIX B:Bike-share System General Information Metrics 148

APPENDIX C:Bike-share System Performance Metrics 150

Fig. 1 Growth of Bike-share Worldwide 13Fig.2 Bike-Share System Performance 39Fig. 3 Bike-share market penetration and usage 43Fig. 4 A Comparison of Systems: Bikes-per-Population 44

and System PerformanceFig. 5 A Comparision of Systems: Operating Cost 48

per Bike and System PerformanceFig. 6 Cycling Infrastructure Implemented 61

Alongside Bike-share SystemsFig. 7 Conceptual Organigram of Communications 73

System between User, Control Center,and Station

Fig. 8 Table of Names of Bike-share Systems 83Fig. 9 Bike-share System Implementing Agencies 90

and Operators

Fig. 10 Comparison of Strengths and Weaknesses 99 of Types of OperatorsFig. 11 Bike-share System Costs 107Fig. 12 Bike-share System Annual Operating Cost 112

Per TripFig. 13 Comparison of Subscription Fees 128

List of Figures


8/152

8Introduction Sub

INTRODUCTIONsection one


9/152

9Introduction Sub

Former Mayor Adrian Fentytakes part in the launch ofthe Washington, D.C., CapitalBikeshare system. Photo byDDOT DC.DDOT (CREATIVE COMMONS)


10/152

10Introduction

Bike-share has taken many forms overthe course of its development, from freebikes left for a community to use at will to

more technologically advanced and securesystems. In every iteration, the essence ofbike-share remains simple: anyone canpick up a bike in one place and return it toanother, making point-to-point, human-powered transportation feasible.

Today, more than 600 cities around theglobe have their own bike-share systems,and more programs are starting everyyear. The largest systems are in China, incities such as Hangzhou and Shanghai.In Paris, London, and Washington, D.C.,

highly successful systems have helped topromote cycling as a viable and valuedtransport option.

Each city has made bike-share its own,adapting it to the local context, includingthe city’s density, topography, weather,infrastructure, and culture. Althoughother cities’ examples can serve asuseful guides, there is no single modelof bike-share.


11/152

11Introduction

Vélib’, in Paris, France,is one of the largest andmost successful publicbike-share systems inthe world.LUC NADAL


12/152

12Introduction

However, many of the most successfulsystems share certain common features:

•A dense network of stations across the coveragearea, with an average spacing of 300 meters betweenstations

• Comfortable, commuter-style bicycles with speciallydesigned parts and sizes that discourage theft andresale

• A fully automated locking system that allows users tocheck bicycles easily in or out of bike-share stations

• A wireless tracking system, such as radio-frequencyidentication devices (RFIDs), that locates where abicycle is picked up and returned and identies theuser

• Real-time monitoring of station occupancy ratesthrough wireless communications, such as generalpacket radio service (GPRS)

• Real-time user information through variousplatforms, including the web, mobile phones and/oron-site terminals

• Pricing structures that incentivize short trips helpingto maximize the number of trips per bicycle per day


13/152

13Introduction

This guide is meant to bridge the dividebetween developing and developedcountries’ experiences with bike-share.It should be useful in helping to plan and

implement a bike-share system regardlessof the location, size, or density of your city.

When the rst bike-shareopened in the 1960s, bike-share growth worldwidewas relatively modest.It wasn’t until after theturn of the century andthe launch of Velo’v inLyon, France, in 2005 andVélib’ in Paris in 2007that growth in bike-shareexploded.CASA BIKE SHARE MAP BY OLIVERO'BRIEN, SYSTEM WEBSITES,PUBLICBIKE.NET

Fig. 1: Growth of Bike-share Worldwide (January 2000 –July 2013 )

700,000

600,000

500,000

400,000

300,000

200,000

100,000

Year 2 0

0 0 2 0

0 1 2 0

0 2 2 0

0 3 2 0

0 4 2 0

0 5 2 0

0 6 2 0

0 7 2 0

0 8 2 0

0 9 2 0

1 0 2 0

1 1 2 0

1 2 2 0

1 3


14/152

14Introduction

. The Benets of Bike-share

Bike-share systems can benet a city in anumber of ways:

• Reduce congestion and improve air quality Bike-share offers an alternative means oftransport for short trips that might otherwisehave been made by car. As of November2011, Washington, D.C.’s 22,000 bike-sharemembers had reduced the number of milesdriven per year by nearly 4.4 million (LDAConsulting 2012).

• Increase accessibility Implementing a bike-share system gives

local users greater access to places that arebeyond their reach on foot.

• Increase the reach of transit Bike-share lls that critical gap between thestation or stop and the nal destination forthe passenger. Since cycling is more efcientthan walking, bike-share enhances mobilityand is much less expensive to the city thanextending public transport service.

The reasons for implementing a bike-share program are often centered on goals of increasingcycling, reducing congestion, improving air quality, and offering residents an active mobilityoption. Bike-share has two key advantages when compared to other transportation projects:implementation costs are comparatively low and the timeline is short. It is possible to plan andimplement a system in one mayoral term (i.e., two to four years), which means that benets to thepublic accrue more immediately than in most transportation projects.

Bike-share has becomea signicant trendworldwide, including inSeville, Spain.CARLOSFELIPE PARDO

• Improve the image of cycling Bike-share systems project a hip, modernimage and can help transform the cyclingculture in a city.

• Provide complementary services to publictransport Bike-share offers an alternative for short tripsthat people would have otherwise made ontransit.

• Improve the health of the residents Bike-share offers an active transport choice,providing both physical and mental healthbenets. Studies have shown that spendingtwenty minutes every day on a bike has asignicant positive impact on mental health(Obis 2011, p. 41).


15/152

15Introduction

topWashington, D.C.designed its CapitalBikeshare system to beeasily used by touristsseeing the sites as well aseveryday use by residents.KEVIN KOVALESKI, DDOT DC(CREATIVE COMMONS)

bottom leftLyon’s Velo’v provideseasy transportation withinthe city for students,residents and tourists.KARL FJELLSTROM

bottom right Buenos Aires, Argentina,has implemented a bike-share system that hasstations near mass transitlines, increasing thecoverage of both systems.CARLOSFELIPE PARDO


16/152

16Introduction

• Attract new cyclists Bike-share offers an easy way into cyclingfor people who may have been preventedfrom cycling by a lack of access to a bike orbike parking. Lyon, France, saw a 44 percentincrease in cycling within the rst year ofopening Velo’v, its bike-share system. In asurvey of members of Capital Bikeshare—Washington, D.C.’s bike-share system—80percent of respondents said that they cyclemore often now than they did before joiningthe program, and 70 percent said that CapitalBikeshare had been important in helping orencouraging them to ride more often (LDAConsulting 2012).

• Improve a city’s image and branding Cycling is a sustainable transportationoption, and a city that implements a bike-share system may strengthen its image asa “green” or innovative city. In 2007, Paris’Vélib’ won the British Guild of Travel Writers’Best Worldwide Tourism project.

• Generate investment in local industry Bike-share has the potential to spurdevelopment of new products and servicesthrough demand for hardware and software,as well as provision of the operations.

Bike-share can also attract existing ridersthrough its convenience and practicality. ITDP

China conducted a survey of bike-share usersin Guangzhou, China, that found that sixteenpercent of the users were previously privatebicycle users. By broadening the bicycleuser base and raising the prole of cyclingin a city, bike-share can build a constituencyfor improved bicycle infrastructure, whichbenets all cyclists, rich and poor alike. Citiesthat have implemented bike-share systemshave found that the benets are felt by awide variety of users—spanning generations,classes, ethnicities, and genders—in a variety

of seasons (New York City Department of CityPlanning 2009).


17/152


18/152

Sub

CARLOSFELIPE PARD

In 2001, newly-elected mayor Bertrand Delanoë set out to transformParis into more sustainable city. Under his low-carbon transport plan, hisadministration added 271 kilometers of bike lanes. However, the lanes werenot well used, and the city determined that the biggest deterrent was thelack of bicycle parking—most apartments were too small to store a bicycle,and people did not feel safe parking their bikes on the street overnight.Parking was also a problem once cyclists reached their destinations,where, again, there were often no safe or legal ways to park their bikes. Inresponse, the city implemented a bike-share system, which addressed theneed for bike parking and increased cycling (Spitz 2008).

How Vélib’ Came to Be


19/152

Introduction Sub 19Introduction

• Docking spaces are the places at the station wherebikes are parked and locked.

• Stations are composed of docking spaces, terminals,

and bicycles. Bikes are parked there for users to checkout, and spaces should be available for users to returnthe bikes. Users can get information and pay for usingthe system. Stations can be manual or automated, orsome variation in between. They can also be modular indesign or xed and permanent (i.e. built into the street).

• Terminals are places where users can get informationabout the system and check in and out bicycles. Theycan be self-service dynamic interfaces for the customeror static information systems that tell users how to

check in or out a bike. They can serve as the nexus ofcommunication between the bikes, the docking spaces,and the control center, as well as be the place forpayment. Terminals usually serve the function of helpingusers locate a station on the street — a visual totem thatis consistently branded. Terminals are also known askiosks, but in this guide we refer to themas terminals.

Bike-share has evolved signicantly since its inception in 1965, whenAmsterdam city councilman Luud Schimmelpennink proposed the world’srst public bike-share system as a way to reduce automobile trafc inthe city center. He proposed that 20,000 bicycles be painted white anddistributed for pick-up and drop-off anywhere in the city center, free ofcharge. When the city council rejected the proposal, Schimmelpennink’ssupporters distributed fty donated white bikes for free use around thetown. The police, however, impounded the bikes, claiming that unlockedbikes incited theft (Schimmelpennink 2012). Though a large-scale freebike program such as the one Schimmelpennink originally imagined hasnever been implemented, smaller-scale free bike systems in Madison,Wisconsin, and Portland, Oregon, have been implemented.

The next attempt at a bike-share system occurred in La Rochelle,France, in 1993, which offered a free, but more regulated, program thatallowed the public to check out bicycles for two hours. Cambridge,England, implemented a similar system in 1993. This type of free bicyclerental system, also known as a “bicycle library,” reduced problems withtheft and vandalism, since users were required to show identicationand leave a deposit in order to use the bicycles. However, these bicyclelibraries also required the user to return the bike to same place fromwhich it had been checked out, limiting the usefulness of the systemas a point-to-point transit option.

. History of Bike-Share

The Bike-Share Lexicon

Bike-share systems go by a variety of names around the world: “bicycle sharing” or simply “bike-share” in North America, “cycle hire” in the United Kingdom, “cycle sharing” in South Asia and“public bike” in China. In this report, we will use the term “bike-share.” Other key bike-sharedenitions used throughout the guide include:


20/152


21/152

21Introduction

top Copenhagen’s ByCyklenis an example of a second-generation bike-shareprogram.ELSAMU (CREATIVE COMMONS)

bottom The system for theProvidencia neighborhoodin Santiago, Chile,is complemented bysegregated bike lanes.CARLOSFELIPE PARDO


22/152

22Introduction

users must either put down a deposit on asmart card or provide a local identicationcard in order to check out a bike. If the bikeis not returned, the user loses the deposit,or he or she can be found and ned throughthe ID card. In Hangzhou, users are requiredto keep deposits on their smart-card accounts,and if they fail to return a bicycle, they forfeitthe deposit.


23/152

23Introduction

Bike share has become asignicant trend in variouscities in the developedand developing world,including Seville, Spain.Photo by CarlosfelipePardo.

The Vélib’ system inParis is the prototypicalexample of a third-generation bike-sharesystem.KARL FJELLSTRO


24/152

24Introduction


25/152

25Introduction

• Universal cards:Bikes can be integrated into other publictransport systems through the use of arechargeable smart card that can cover arange of payments and trips. Many cities inChina already have this kind of integration.In Hangzhou and Guangzhou, for example,the card used for the local bike-share systemcan also be used on the bus, bus rapid transit(BRT), and metro systems. The use of theseuniversal cards is now spreading to other

countries and cities.

• Modular, movable stations:These stations do not require excavation andtrenching, which reduces implementationtime and costs. Also, because the stations areeasily movable, the system can be optimizedonce demand patterns reveal themselvesthrough usage. They can also be removedduring winter months.

. New Developments and Trends

left Stuttgart, Germany,implemented a bike-sharesystem that includeselectric bikes withchargers at stations andGPS location devices.CARLOSFELIPE PARDO

opposite Montreal’s Bixi was therst bike-share systemto use solar-powered,modular stations.MAX HEPP BUCHANAN

• Solar cells:Solar cells can power stations and wirelesscommunications. Solar cells make modularstations feasible, as they eliminate theneed for excavation to connect the stationto underground power lines. The systems inBoston, Washington, D.C., London, Montreal,and Rio de Janeiro have stations that arepowered entirely by solar energy and arecompletely wireless.

The future of bike-share will probably includeoffering cargo bikes for large purchases,electric-assist bikes, and bikes for children.

Many new systems incorporate innovative characteristics that some believe represent a fourthgeneration of bike-share, including:


26/152

26Introduction

. Building Political Will

Successful implementation of a bike-share system requires strong political support to ensurefunding, land use rights, and coordination between various city agencies. Involving more than onepolitical party is critical to ensuring support for bike-share over several years and multiple electioncycles.

Building political will begins with educating political leaders on the benets of bike-share. Thiscan include presentations on and site visits to successful projects. Persuading decision-makersto travel to other cities to actually see and use successful bike-share programs, and to speak toother implementers, builds the necessary political will to make bike-share a reality. These decision-makers become champions for the new system in their own cities.

London Mayor Boris Johnson’s strong support for the city’s bike-share system earned that systemthe nickname “Boris Bikes.” His determination to increase the use of bikes in London by improvinginfrastructure and setting bike-share as a top priority created the context for a successful andinnovative system in one of the world’s most famous cities. While the London system is overseenby the city’s transport department, Transport for London, and operated by Serco under a six-yearcontract, the support of the mayor’s ofce was the key to the system’s success. Johnson personallypromoted the bike-share system to residents of the boroughs, whose support and cooperation wasnecessary to the success of the project (Mulholland 2008).

New York City Departmentof TransportationCommissioner JanetteSadik-Khan and other cityofcials test bicycles in forNew York City’s bike-shareprogram.NYC DOT (CREATIVE COMMONS)


27/152

27Introduction

Prior to entering the planning phase, the agency implementing the bike-share system must have a basic knowledge of the essential elementsof bike-share so that it can space stations appropriately and create abusiness model and a nancial model. These elements include bikes,stations, software and other technology needs, as well as personnel/stafng objectives. These elements will impact both the business andnancial models.

Many bike-share systemsare found in areas withlots of activity, makingit convenient for peopleto pick up and drop offa bike, like in Paris' citycenter.LUC NADAL

. Elements of Bike-share


28/152

28Introduction Sub

THE PLANNINGPROCESS ANDFEASIBILITYSTUDY

section two

Mexico City introducedsegregated bike lanesalong Reforma Avenue,which are frequently usedby Ecobici users.BERNARDO BARANDA


29/152


30/152


31/152

31The Planning Process and Feasibility Study

New York City installedits modular bike-sharestations fairly quickly—over the course ofone month.LUC NADAL


32/152


. Feasibility Study The feasibility study establishes the critical parameters that will guide the planning and designprocess—specically the coverage area and size of the system—and then analyzes whetherthe proposal will be nancially feasible and under what conditions. The feasibility study shouldrecommend investment and revenue sources, a contracting model, and an organizational structure,as the agency or department conducting the feasibility study may or may not be the implementingagency. Finally, the feasibility study will also need to review the local context and identify anyspecic local obstacles to implementation, including weather, cycling infrastructure, culture, andpolitical and legal realities. Much of the feasibility study can be done by drawing on other systems’experiences and adapting them to the local context.

The rst step is to outline the city’s objectives for a bike-share system. Bike-share systems areoften implemented as part of a general sustainable transport initiative to reduce pollution andimprove mobility options. Strategic objectives for bike-share may include solving the “last mile”problem for transit passengers who still need to travel from the station to their destination (asin the San Francisco Bay Area in California), avoiding capital investments in order to increase thecapacity of overcrowded mass transit (as in Guangzhou, China), meeting targeted city modal splitsor pollution targets (as in Paris), developing tourism (as in Hangzhou, China, and Paris), and evengenerating employment (as in Hangzhou). These locally dened objectives will inform the rest ofthe feasibility study.

Guangzhou, China’s, bike-share stations are near

BRT stations, benettingusers of both systems.KARL FJELLSTROM


33/152


1. Demand analysis The demand analysis identies the potential number ofsystem users and forms the basis for all other analysis. Itrequires the following steps.

• Dene the proposed coverage area. Usually, citieschoose as a rst phase the areas where there will bethe most demand for bike-share. Residential populationdensity is often used as a proxy to identify those placeswhere there will be greater demand. (See section 2.3.3for more detail about the coverage area.)

• Dene targets for key performance metrics. This shouldinclude both the two key performance metrics discussedin section 2.3.1 and the indicators for evaluating howwell the system is meeting its objectives in section 2.3.2.

• Create a demand prole. Review existing demandand conditions for cycling, taking into account thepopulation of the coverage area, the number ofcommuters, current modal split, existing transit, bicycleand pedestrian networks, and existing major attractionsthat will draw people to the area. Sometimes it is usefulto create proles of potential bike-share users to geta sense of who will use it and at what scale, but it has

generally been found that people of all incomes andbackgrounds use bike-share.

• Create estimations of demand. One way to do this isto create a Price-Elasticity of Demand (PED) analysisaccording to various customer types. Another, lessrigorous, way is to create an estimation of demandbased on a percentage of the population, known as theuptake rate. After Vélib’ opened, Paris saw a 6 percentuptake, meaning that 6 percent of the populationused the system (Nadal 2007). New York City ran threescenarios: a 3 percent uptake by the existing population,

a 6 percent uptake and a 9 percent uptake. The cityultimately used 6 percent for nancial estimations (New York City Department of City Planning 2009).

• Size the system by dening station density, bike density,and bikes per station. These basic planning parametersare discussed further in section 2.3.4.

2. High-level nancial feasibility analysis Based on the demand analysis and size of the system,preliminary numbers can be used to estimate how muchthe system will cost, including both capital costs andoperational costs. This is a high-level estimation used toguide decisions, not a detailed budget, which should bedone later. This analysis includes the following steps:

• Propose options for station type, bicycles, andtechnology to create a capital cost estimate.

• Estimate operational costs based on the size ofthe system. This should include maintenance andredistribution, as well as replacement costs for bikes.

• Propose nancing options to identify the mostappropriate combination of user-generated revenues(per-use and membership fees), government funds,corporate sponsorship, street advertising contracts, etc.

• Analyze estimated costs against nancing options toensure that the proposal is nancially feasible.

• Recommend a business model that establishes anorganizational structure and contracting model.

After dening the objectives of the bike-share system, the feasibility study should include thefollowing three main components:


34/152


3. Analysis of risks and barriers Identifying possible barriers and risks willhelp planners mitigate those challenges asthey go into detailed planning and design.This analysis includes the following steps:

• Review possible barriers to implementationand propose mitigation measures. Suchbarriers may include access to credit cardsby the users, advertisement regulationsand existing advertising contracts, helmetrequirements, trafc laws, safety concerns,institutional constraints, etc.

• Identify risks to project implementation andpropose mitigation measures. These risksmay include institutional inghting andlack of cooperation, NIMBY-ism and protestby the community, and the absence of apolitical champion for the system.

These three components are an iterativeprocess whereby decisions about the coveragearea and system size may change based on thenancial feasibility. This study becomes thebasis for the next steps: detailed planning anddesign, the creation of business and nancialmodels, and tendering and contracting. Withthe guidelines determined in the feasibilitystudy, the government organizing team canmove into the planning phase.

New York City’s Feasibility Study

New York City’s feasibility study determinedthat the rst phase would focus on thecity’s medium- to high-density areas, likeHerald Square, Midtown, lower Manhattanand parts of Brooklyn. The recommendedbusiness model was to contract operationsto a private company, with the assets ownedby the city and operational costs covered bymembership fees and sponsorship.

NYCSTREETS (CREATIVE COMMONS)


35/152



36/152


Most bike-share stations are rolled out inphases, with the most successful systems, likeParis, Lyon, and Hangzhou, beginning with arobust citywide network of bike-share stations.The feasibility study can help determine a

phased implementation plan. This can beespecially useful if the eventual goal is to createa system on a large regional scale that mightbe challenging to implement all at once. Initialphases should focus on covering as much ofthe city as possible, focusing on areas that arethe densest in terms of demand, have strongbicycle infrastructure, and would have goodpublic support for bike-share. Areas that arenancially more difcult or constrained byinfrastructure challenges should be prioritizedfor future phases.

Generally, the rst phase needs to be bothlarge enough to connect meaningful originsand destinations and dense enough to ensureconvenience and reliability for the user. Smallerpilots are not ideal for bike-share, as that scalecan limit the usability of the system due to poorcoverage or bike availability, which ultimatelydamages the public perception of bike-share asa viable mode of transport. Smaller pilots haveoften not been successful, as was the case with

Washington, D.C.’s original Smartbike systemand Rio de Janeiro’s original Samba bike-share.Both cities went on to relaunch their bike-shares based on lessons learned from thoseexperiences.

Paris launched Vélib’ in 2007 with 7,000bikes at 750 stations across the city. Thesystem immediately began attracting tens ofthousands of riders each day and averaged75,000 trips per day in its rst year, with peakdays exceeding well over 100,000 riders (New York City Department of City Planning 2009).The successful launch also generated publicsupport for the system and brought the cityinternational acclaim. The following year, Vélib’grew to 16,000 bicycles in 1,200 stations inthe city, and it is now planning to have more

than 20,000 bicycles at over 1,450 stationsboth within Paris and within twenty-nine othercommunities on the periphery of the city.

Paris rolled-out Vélib’ tohigh demand areas, likeabove in Les Halles, wherethere were more peopleto use the system, asevidenced by this emptystation.KARL FJELLSTROM


37/152

Washington, D.C.’s rstbike-share system, calledSmartbike (above), andits current system, calledCapital Bikeshare (left).CARLOSFELIPE PARDO

The Smartbike system, launched in August

2008, was the country’s rst fully automatedbike-share system. In this public-privatepartnership between Clear Channel Outdoorand the District of Columbia Departmentof Transport, Clear Channel receivedoutdoor advertising rights to the city’s busshelters, while the District received all usersubscriptions fees to operate the system.The city dened the system as a pilot project,with ten stations and 120 bikes. Due to the

small number of bikes and stations, as well as the great distances betweenstations and limited operating hours, the program was poorly utilized andthus largely unsuccessful (Silverman 2008 and DePillis 2010).The District of Columbia chose to nish the bike-share portion of itscontract with Clear Channel and then totally revamp the system.

In September 2010, Smartbike was replaced by Capital Bikeshare, a fullyautomated system with 1,100 bicycles and 116 stations, but now availabletwenty-four hours a day, seven days a week. The new system is operatedby Alta Bicycle Share, a company that specializes in operating bike-sharesystems and has experience in other leading cities around the world.

Why Did Washington, D.C.Relaunch Its Bike-share?


38/152


Bike-Share Essentials

In order for a bike-share system to

be well-used and efcient, it must beproperly planned and designed. Basedon the performance of existing systemsacross the globe, ITDP has developed thefollowing planning and design guidelinesthat are characteristic of the best-usedand most efcient systems. More detailabout each recommendation can befound in the guide.

PLANNING GUIDELINES• Minimum System Coverage Area: 10 km 2

• Station Density: 10–16 stations per km 2

• Bikes/Resident: 10–30 bikes for every 1,000 residents(within coverage area)

• Docks per Bike Ratio: 2–2.5 docking spacesfor every bike


39/152


40/152


2.3.1 Basic Context Data and System MetricsIn order to complete a feasibility study, a range of local data must be collected and analyzed.This data will help to determine the appropriate size and scale of the bike-share system to bestmeet the goals for the system. The following two data sets are critical to establishing the basicframework for the feasibility study—dening the physical size of the area and the the potentialsize of the users:

• System Coverage Area:Dened as the contiguous area, in squarekilometers, in which bike-share stations arelocated. The coverage area includes a 500meter radius around each station located onthe edge of the area.

• Population in System Coverage Area:Dened as the number of people that live inthe system coverage area. This gure can bequickly obtained by multiplying the systemcoverage area by the population density(i.e., the number of residents per kilometerin that area). The more specic the data isto the coverage area, the more accurate theplanning will be.

. Bike-Share Metrics

The planning of a bike-share system is based on a simple analysis of readily available data. Thisallows planners to design a system that is the right size and scale to meet their performance andnancial goals for the system.

For comparisons in this guide, the average population density of the entire city was applied tothe system coverage area to nd the population in the system coverage area. This likely under-estimates the population in many coverage areas because bike-share systems are generallyimplemented in areas with higher-than-average population density and high concentrations ofworkers commuting in.

At its most basic level, a bike-share system is comprised of a certain number of bikes, docks,and stations, which will serve a given market. These basic data points are described below:

• Number of bikes Dened as the number of bikes in activecirculation in a system (in a dock or in use).This is not the total number of bikes ownedby a system (which may include bikes that arebeing repaired or are part of the contingencyeet), which is less relevant to measuring theperformance of the system.

• Number of docksDened as the number of functional parkinglocations where a single bike can be checkedin or out. Some systems allow bikes to bechecked in and out without the use of docks,which may skew comparisons.

• Number of stations Dened as the number of specic locationswhere a bike can be checked in or out. Eachstation consists of multiple docks.

For planning purposes, two basic types of users are dened. This distinction is used to understandusage and dene fees. These are:

• Casual usersDened as users who pay for subscriptions ofseven days or less. Casual members typicallycan purchase these short-term subscriptionsthe day of use.

• Long-term usersDened as users who subscribe for amonth or longer. The registration processfor annual members typically takes a dayor more and often includes a registrationtoken, such as a key fob or a membershipcard, to provide access to the system.


41/152


2.3.2 Performance MetricsAn efcient, reliable and cost-effective system will maximize two critical performance metrics:

• Average number of daily uses per public bike Ideally, four to eight daily uses per bike.Turnover is critical to a successful bike-share system, and this is a measure of theefciency of the system. Fewer than fourdaily uses per bike can result in a very lowcost-benet ratio, while more than eightdaily uses can begin to limit bike availability,especially during peak hours. In 2010, Parisaveraged more than four daily uses per bikefor the whole year, including winter, when the

usage is lower.

• Average daily trips per resident Ideally one daily trip per twenty to fortyresidents. This is a metric of marketpenetration. High quantity of uses amongthe population of the coverage area is key toachieving the primary objectives of a bike-share system, including increased bicyclemode share, decreased congestion of vehicleand transit networks, and promotion of safe,clean, healthy modes of transport. Lyon, forexample, has one daily trip per twenty-ve

residents.

These two metrics have an inverse relationship. Many systems have a high average daily useper bike because they actually have too few bicycles in circulation, and this means that marketpenetration (expressed here as average daily trips per resident) will be very low. Other systemsmay have high market penetration, but very few uses per bike, indicating inefcient usage ofinfrastructure and low cost-benet, likely due to a surplus of bikes. The planning of a bike-sharesystem must be carefully calibrated to ensure performance is within the optimum range forboth metrics.

Fig. 2 : Bike-Share System Performance:Trips per Bike vs Trips per 1,000 Residents

T r i p s p e r 1 , 0 0 0

R e s i

d e n t s

( M e a s u r i n g

M a r k e t

P e n e t r a t i o n

)

Trips per Bike (Measuring Infrastructure Usage)

0 1 3 5 7 9

70

50

30

20

Boulder

Minneapolis

DC Chicago

San Antonio

Boston

Denver

Madison

London

Rio de Janeiro

Paris

NYC

Lyon

Moderate Performance:High market penetration butlow infrastructure usage

Moderate Performance:High infrastructure usagebut low market penetration

Mexico City

(187) (114)

Montreal(11)Barcelona

Low Performance:

Low market penetration andlow infrastructure usage

High Performance:High market penetration andhigh infrastructure usage


42/152


A system that has a very high number of dailyuses per bike may have too few bikes to meetdemand. This results in low market penetration,and a smaller impact on the city’s objectives.The early years of the Barcelona system serveas a prime example, as there were, on average,nearly ten daily uses per bike, but the numberof people using the system relative to thecity’s population was very low. According toSertel, the operator of Rio de Janeiro’s bike-share system, BikeRio has about ten to twelvedaily uses per bike in 2013. This may be due inpart because of the limited numbers of bikesavailable. If bikes are not readily available,the system will not be viewed as a reliablemode that could replace or compete with otheroptions, such as transit or private cars.

Conversely, a system with too many bikesand a relatively low number of users couldresult in the perception that bike-share is aninvestment with a low return. An indicator ofsuch a situation is the average number of tripsper bike. Bike-share systems should striveto maintain an average of four daily uses perbike to maximize the public cost-benet of thesystem.

What to Do if a SystemIs Too Popular?

Barcelona’s Bicing was more popularthan anticipated. Within its rst twomonths, 30,000 people signed up formemberships—number that had beenforecast for the entire year. While thecity originally wanted to include touristsas part of the membership base, thatoption was removed to address the highdemand and to avoid competition withthe existing bike-rental companies thatcater to tourists. However, short-term

memberships, which tourists typically buy,can be a signicant revenue generator,since cities usually give discounts forannual memberships. Now, with Barcelonafacing a nancial crisis, city services arebeing cut across the board. As a result, thecity is proposing to increase Bicing fees by116 percent, which has sparked a publicoutcry (Baquero 2012).

The gure on the previous page shows howsixteen bike-share systems perform basedon these two critical performance measures.The systems in the green area of the charthave the highest overall performance, asthey are achieving optimum levels of bothmarket penetration (expressed as daily tripsper resident) and efciency of the system (asexpressed by daily uses per bike). Systems thatfall in the orange zone are achieving high usageper bike, which reects good cost-benet ofthe system; however, they are not achieving ahigh market penetration, which indicates thatthese systems may warrant expansion. Systemsin the yellow zone are achieving high marketpenetration and usage among residents, buthave low usage per bike, which indicates thatthat they may have a surplus of bicycles. Bike-share systems in the red zone are not achievinggood usage levels on a per capita or per bikebasis, meaning they most likely need to expandtheir size and adjust other factors such asstation placement or pricing.

MICHAEL KODRANSKY


43/152


. Coverage AreaWhen beginning to plan a system, identifying a coverage area (the physical area that the bike-share system will cover) and saturating it with the appropriate number of stations are the mostcritical factors in creating a successful system with high ridership. The coverage area must belarge enough to contain a signicant set of users’ origins and destinations. If it is too small toconnect meaningfully to other places, the system will have a lower chance of success because itsconvenience will be compromised. While many people attribute Melbourne’s low ridership to the

city’s mandatory helmet law (Preiss 2011), the city’s bike-share operator, Alta, attributes it to thesmall coverage area, which was the smallest of the three options recommended in the feasibilitystudy (Alta Planning 2012).

Dense, mixed-use areas with a high trip-generation capacity (generally city centers) are likely tosee the most demand for bike-share, as they will be both the origin and destination points of manytrips and are usually the best places to start. When dening the coverage area, the city will haveto balance demand with costs. The identication of the appropriate coverage area is best carriedout by qualied planning institutions through surveying and statistical data analysis.The coveragearea must be determined in tandem with the system size to ensure that the system is both largeand dense enough to encourage high ridership due to its convenience, reliability and ubiquity.

Every dockingstation in 19 ofthe world's bikesharing programs;each city mappedat the same scale.

DAVID YANOFSKY, QUARTZ, QZ.COM


44/152


45/152


Bikes Per Population

Ideal ratio of bikes per population isbetween 10 to 30 bikes per 1,000 residents.

Bike-share Usage:Trips per Bike vs Station Density

Station Density (stations per km 2 )

12

10

8

6

4

2

0.0

B i k e U s a g e :

D a

i l y

T r i p s p e r

B i k e

R2=0.26063

Fig. 3: Station Density and PerformanceBike-share Market Penetration:

Trips per 1,000 Residents vs Station Density


180

160

140

120

100

80

60

40

20

0

M a r k e t

P e n e t r a t i o n :

T r i p s p e r

1 , 0

0 0

R e s i

d e n t s

Boulder

Boston

DenverMadison

Mexico City

Montreal

R2=0.43291

London

Lyon NYC

BarcelonaRio

Paris

Minneapolis

Buenos AiresDC

San Antonio0 2 4 6 8 10 12 14 16

0 2 4 6 8 10 12 14

Boulder

Boston

DenverDC

Madison

Mexico City

Montreal

London

Lyon

Rio de Janeiro

NYC

Barcelona

Paris

MinneapolisSan Antonio

Station Density

Ideal station density is between 10 and 16stations per square kilometer.Fourteen stations per square kilometer is

equivalent to:• One station every 300 meters• Thirty-six stations per square mile

16

A direct correlation existsbetween a higher stationdensity and a highermarket penetration. Acorrelation, albeit weaker,also exists between ahigher station density anda higher system efciency.SOURCE: ITDP DATA


46/152


A higher number ofbicycles per residentwill increase marketpenetration. Providingmore than thirty bicyclesper resident, however, maylead to too few daily usesper bike. While the data

show that the number ofdaily uses per bike tendsto decrease as the numberof bikes per residentincreases, the statisicalrelationship is too weakto accurately predictvalues here. These graphsillustrate the inverserelationship between themetrics.SOURCE: ITDP DATA

Bike-share Usage:Trips per Bike vs Bikes per 1,000 Residents

Bikes per 1,000 Residents

12

10

8

6

4

2

0

B i k e U

s a g e :

D a

i l y

T r i p s p e r

B i k e

Boulder

Boston

DenverMadison

Mexico City

Montreal

R2=0.10766

London

Lyon NYC

Barcelona

Rio

Minneapolis

Paris

DC

0 5 10 15 20 25 30 35 40


200

180

160

140

120

100

80

60

40

20

0

M a r k e t

P e n e t r a t i o n :

T r i p s p e r

1 , 0

0 0

R e s i

d e n t s

BoulderBoston

Denver Madison

Mexico City

Shanghai

Montreal

R2=0.43291

LondonLyon

NYC

Barcelona

Rio Paris

MinneapolisBuenos Aires

DC

San Antonio

0 5 10 15 20 25 30 35 40

Fig. 4 :Bike-share Market Penetration:

Trips per 1,000 Residents vs Bikes per 1,000 Residents


47/152


After the system opens, another parameter thatwill be useful in evaluating performance is thenumber of annual members per bike in service.This metric is another way to measure theamount of use that can be regularly expected.Many practitioners in the eld recommend a10-to-1 ratio of annual members to bikes tocreate a well-functioning system (Cohen 2013).Systems below this ratio will need to recruitmore members through better promotions,better bicycle facilities, better system service,etc. Systems above the 10-to-1 ratio will likely

need to expand to accommodate demand. Forexample, New York City surpassed a 16-to-1ratio in its rst two months of operation andis experiencing difculty meeting demand atmany locations. This signals that the systemneeds to expand to meet ever-growing demand.

The parameters used for the feasibility studyand as the framework for planning do not addressspecic station locations or the exact number ofbikes and docking spaces at each station, as thatis determined in the planning and detailed designphase.

Hangzhou has stationswith attendants whomonitor bike check-inand check-out.KARL FJELLSTROM


48/152


. Financial Analysis

Once the system size has been decided, aninitial nancial analysis can be undertaken.This analysis usually considers the estimatedcapital outlay, projected revenue, and estimatedoperational cost. It should also consider theadvantages and disadvantages of differentnancing mechanisms.

An estimation of capital costs and operatingcosts can be calculated by multiplying thenumber of bikes, docks, and stations againstan average cost. The capital and operatingcosts are a function of system technologyand are straightforward to determine, but therevenue depends on usage levels and can onlybe fully estimated in the detailed planningstage. Usually the revenue scenarios arebased on expectations of demand using both aconservative estimate (in which demand, andtherefore revenue, is low) and an optimisticscenario in which demand projections arehigher, resulting in higher projected revenue.

Capital costs are often expressed in termsof the “cost per bike,” dened as the totalcost of the system—including stations, bikes,redistribution equipment, the control center,and other equipment—divided by the totalnumber of bikes in the system.

Operating costs vary widely from systemto system and from city to city due to manyfactors, such as the cost of labor, accountingpractices, and, of course, system planningand infrastructure. Common operating costsare expressed in an annual-per-bike amountand can range drastically depending onredistribution mechanisms and needs, laborcosts and service level delivery. In Zhuzhou,China, for example, annual operating costsare ¥1,200 (US$191) per bike but similar 3rdgeneration systems in the West can have

operating costs of upward of US$1,970 -4,200per bike (Midgley 2011).

Using cost-per-bike may be useful in theplanning stage to size the system nancially,but in analyzing system performance after itopens, a per-bike analysis is not recommended,because bike eets vary from day to day (Cohen2013). Some have used the per-dock basis foranalyzing annual operating costs as a morestable, and therefore, more comparable basis

Bike-share systems canrange from very simple(such as EnCicla in

Medellín, top) to morecomplex (such as Citi Bikein New York City). Thelevel of complexity willhave a direct impact onthe nancial model.CARLOSFELIPE PARDO

(Cohen 2013). However, this guide recommendsevaluating the cost efciency of a system afterit opens by looking at operating costs per trip.

For example, Mexico City and Washington,D.C.have similar operating costs per trip ($468 and$556 respectively), while operating costs perbicycle are very different ($2,594 and $1,255respectively). Mexico City’s per bike costs areabout double that of Washington, D.C.’s butits per trip costs are lower. Like other transitsystems, the goal of bike-share systemsis to attract and move as many people asefciently as possible, and a system’s operatingexpenditure should be based on the numberof people, as expressed in the number of trips,

using it. Most transit systems express theircosts in a similar way.

To estimate revenue, multiply the demandestimations for usage against the proposedrevenue structure. Demand is often estimatedusing what is called an uptake rate, which is anassumption of the likely usage as a percentageof the residential population of the coveragearea. As previously discussed, London used anuptake rate of nine percent, based on market


49/152


studies. In Paris, after Vélib’ opened, thesystem saw a six percent uptake rate. New YorkCity looked at three scenarios: a conservative

estimate based on a three percent uptake rate,a middle estimate using a six percent uptakerate, and an optimistic scenario of a ninepercent. Ultimately, the city decided to use thesix percent rate for projections (New York CityDepartment of City Planning 2009).

Another measure of the nancial health ofa system is the percentage of operating coststhat are covered by membership and userfees. This metric, known as farebox recovery,measures the degree to which a bike-sharesystem is self-sustaining. Most systems do not

meet their operating costs through membershipand user fees alone, although some do comeclose. This metric can be used to determine thedegree to which other revenue sources, such asadvertising revenue, government subsidies, andsystem sponsorship, will be needed to coveroperating costs.

A nancial analysis of a bike-share systemshould consider what percentage of total tripswill be taken by long-term members, and what

Many Chinese systems,including in Zhuzhou, optfor simpler, cheaper bikes.LI SHANSHAN

percentage by casual members. This metriccan reveal which of the two user groupswill generate the majority of the system’s

revenue. In most systems, casual users arecharged a higher price per day than annualusers, and casual users are the source ofmore revenue, even if in numbers they arenot the largest user group. Casual usersare less familiar with the bike-share systemin a city and are therefore more likely tobe charged fees for exceeding time limits.However, systems with high percentagesof casual users are more susceptibleto changes in tourism and subsequentuctuations in revenue. Systems with a

high percentage of casual users may relyon overtime fees for revenue, leading tounhappy customers, who had inadvertantlyaccrued these fees. Typically, as a systemgrows, the percentage of casual usersdeclines, as some casual users purchaseannual memberships.


50/152


SOURCE: ITDP DATA

Operating Cost per Trip vs. Bikes per 1,000 Residents


$6.00

$5.00

$4.00

$3.00

$2.00

$1.00

$0.00

O p e r a t i n g

C o s t p e r

T r i p ( U S D )

Boston Denver

Mexico City

London

BarcelonaMinneapolis Was hington, D C

0 2 4 6 8 10 12 14 16

Fig. 5: Bike-share Economic Performance:

Operating Cost per Trip vs. Station Density

The different systems show

a degree of variation in costper trip, with no denitecorrelation. Logically,larger, denser systemsshould benet fromeconomies of scale, andeach additional trip shouldcost less money. However,the limited amount of dataavailable does not yetconrm this notion. SOURCE: ITDP DATA

Montreal

Lyon


$6.00

$5.00

$4.00

$3.00

$2.00

$1.00

$0.00

O p e r a t i n g

C o s t p e r

T r i p ( U S D )

Boston Denver

Mexico City

London

Barcelona

MinneapolisWashington, DC

0 5 10 15 20 25 30 35 40

Montreal

Lyon


51/152


An analysis of US systems illustrates thisshift from casual users to annual. Washington,D.C.'sCapital Bikeshare annual membersaccount for 80 percent of the system’s trips,whereas in Boston, the rst year saw only56 percent of rides by annual members (theremaining by casual) and that number grew to69 percent in the second year. Most of the othersystems have a split, such that approximately

60 percent of the rides are from annualmembers and 40 percent of the rides are fromcasual members. Madison B-Cycle system saw57 percent of its rst year trips taken by casualusers, but only 34 percent of rides were bycasuals in the second year (Cohen 2013).

When projectingusage, London ofcialsconsidered density andlocation of stations.KARL FJELLSTROM


52/152

52Introduction Sub


53/152

53Introduction Sub

DETAILEDPLANNING

ANDDESIGN

section three


54/152

54Detailed Planning and Design

Detailed system design and planningapplies the parameters discussedpreviously to determine the exact locations

and sizes of stations. During this phase, thecity should also decide on the hardwareand software of the system, includingvehicle type, station design, and IT systems.Finally, during the planning stages, thecity needs to develop a communicationsplan and marketing strategy, including thebrand for the system.

Stations should be roughly uniformdistance from one another. The sizeof the station will be a function of theanticipated demand and the attractionsof a particular area, and station’s locationwill depend on the actual environment.The station density that was decided inthe feasibility stage should be more orless adhered to, although some factorsmay inuence that. For example, areasthat are more densely populated mayrequire more stations than the statedparameter, while other areas, due to landuse and existing conditions such as largeparks or industrial areas, may require less.


55/152


56/152


of docks per station in very high-densityareas with high peak-hour ows. Vélib’ranges from twelve docks per station in

lower-trafc areas to seventy docks perstation in central tourist areas, whilestations in Hangzhou and Shanghai canaccommodate hundreds of bikes at asingle central location.

Existing trip patterns can be researchedto help determine demand and stationlocations. Because most transport modelsuse zone structures that are too large tobe of much use in deciding on the sizeand location of bike-share stations, mostcities use local knowledge to determinethese elements. To get an idea of populardestinations in the area, origin-destination(OD) surveys can be conducted at majorlocal public transport terminals andstations, focusing on passengers whotransfer to taxis or buses to completetheir journeys. This can help to determinewhere the system is most likely to succeedand to anticipate demand.


57/152


. Station Location

Choosing good station locations is critical to ensuring that the systemwill have high usage and turnover. Stations should be situated such thatthat they can be found at regular and convenient intervals throughout thearea and are in desirable locations that generate usage throughout theday. General guidelines for locating stations are as follows:

The bike-share systemin Brussels has manystations in key activityareas around the city.KARL FJELLSTROM

• The station-density parameter, such as 1station every 300 meters, that was denedin the feasibility study (see section 2.3.4)

should be the basis to ensure mostly uniformcoverage.

• Stations should be adjacent to masstransit stops and stations, as bike-share iscomplementary, helping passengers connectmore easily and quickly to their destinations.

• Whenever possible, stations should belocated along existing bike lanes or onstreets that are safe and accessible for bikes.

• Stations are best situated on or near corners,so that users can access and egress frommultiple directions.

• Stations are ideally located between multipleuses that generate activity at different timesof the day. This ensures that bikes will be

used from morning to night. For example,a station that is situated between an ofcecomplex and bars/restaurants means thatthe bikes are used by commuters during themorning and evening and by the restaurantand bar customers during the middle ofthe day and night. Proximity to places thatattract lots of different types of activity overthe course of the day increases safety for theusers.

• Stations should not be placed next to barriers

like train tracks, or single-use areas such asa large gated park or factory. Barriers reducethe area that the bikes can reach, reducingtheir effectiveness. Stations in single-useareas have lower usage because there arefewer activities to attract a variety of users.Underused areas, like underpasses, whileinteresting in terms of having space to placethe station, should be carefully consideredfor potential safety concerns.


58/152


The city should specify which guidelines itwants to follow as a framework for the nextstep of determining the exact location of eachstation. Determining ideal station location is atwo-step process:

1. Create a rst draft of all station locations

2. Finalize the positions through site visits andstakeholder engagement

Creating a rst draft of station locations can bedone in one of two ways: The rst draft can bemapped out remotely using a grid approach andthen veried by a site visit, or it can be donein the eld and then analyzed remotely andadjusted where there is too much coverage ortoo little. Either way, the idea is to have roughlyeven distribution of stations while workingwithin the constraints of the environment.

To map locations remotely, draw a 1 x 1 kmgrid over the map of the coverage area usinga computer program such as Google Maps orGIS, or simply using a paper map, marker, andruler. The grid provides a simple foundation forrational and even distribution of stations. Themap should show transit stations and bicyclelanes, as well as any other important demand

generators or facilities. Then, applying thestation density parameter and station locationguidelines, calculate the number of locationsper grid square. This ensures that stationsare spaced evenly throughout the coveragearea. If the desired station density is fourteenstations per square kilometer, fourteen stationsshould be placed more or less evenly in eachbox on the map’s grid. The grid can be altered,subdivided or zoned into high-station-densityzones and low-station-density zones if desired,though a uniformly high-density approach is

recommended for most situations.If you start in the eld, you will need to

analyze the results to ensure continuouscoverage by drawing coverage areas of eachstation (using a radius of 150 or 200 meters).The areas left without coverage will, then, needto be analyzed to see if a station should beadded, and, if so, where. While the goal is touse the station-density parameter to ensureuniform coverage, rarely is this achieved 100

(NEW YORK CITY DEPARTMENT OF CITY PLANNING 2009)

New York City’s Station Location Guidelines

The following are the general guidelines for the location of New YorkCity’s bike-share stations:

• On wide sidewalks or in the roadbed. Bike stations should notimpede pedestrian or vehicular trafc

• With enough frequency to ensure program visibility and use(approximately 28–30 stations per square mile)

• Along existing or proposed bike lanes whenever possible

• Near subway stations, major bus stops, the Staten Island FerryTerminal and other ferry landings

• Near major cultural and tourist attractions

• Adjacent to major public spaces and parks

percent in practical terms. This is becauseexisting infrastructure and space often dictatehow many and what size of stations are needed.

Exact positioning of the station will requirea site visit. Using a bicycle to conduct thesite visit is recommended because it will givethe planners a sense of the coverage areafrom the perspective of a cyclist and will bemore efcient for siting many stations. A tapemeasure and a GPS or smart phone will alsobe needed. If using a map, visit each generallocation marked on the map grid and examinethe area to determine the specic site to bestaccommodate the bike-share station and seeif there is sufcient space for the station. Thespace needed per station will depend on howmany bikes are at that station. Depending onthe bike docking design, each bike will need aspace that is approximately 2 meters long and0.7–1.5 meters wide.

Once this draft has been nalized, it needsto be vetted by stakeholders. Engagingstakeholders in the station location processis a good way to build support for the projectand gain an understanding of the demand forparticular stations. Community workshops topresent these plans provide an opportunity


59/152


to disseminate information about bike-shareto people living in neighborhoods where thesystem will be introduced, and can be valuablein choosing where to place docking stations andunderstanding demand. Before holding publicmeetings, the government should establishcriteria for approving or denying stationrequests. Another, increasingly popular methodis to crowdsource the station locations throughwebsites. This can also serve to identify highdemand areas. This, however, will not identifythe exact location of the station, just proximate

areas that need to be served by the system.Exact locations for stations need to be donethrough analysis and selection by the planningteam.

New York City crowdsourced stationlocation ideas by organizing an websitewhere people could request a bike-sharestation or vote for one that had already beenidentied. The city received more than 10,000station location ideas and 55,000 votes for

suggested stations. This helped show publicsupport for this initiative. The city also held159 workshops with the community to renestation locations. For each possible station,the Department of Transport identied up tove potential locations that were then broughtto the community to nalize the location.Crowdsourcing and community workshops alsocan help provide some political cover whenthe system is being implemented if a backlashoccurs when the stations show up on thestreets, as happened in New York City

(Miller 2013).Once the specic location for a bike-share

station is established, it should be placemarkedusing a GPS system (or bookmarked on a smartphone), a photo should be taken and precisedetails noted about the station positioning.These coordinates, notes, and photos shouldthen be given to the station installationcontractor to safeguard against location/positioning errors, which are common.

No stationsare located here becauseof a variety of reasons:a dense urban village,

private land (universities)and a park where it is

prohibited to ridea bicycle

Proposed Station Placement for Guangzhou, China’s, Bike-share

The station-density parameter is a guideline that, on the ground, may need to be adjusted,as is seen with the rst phase of the bike-share system in Guangzhou. The bike-share systemwas implemented in conjunction with the BRT corridor. Bike-share stations are needed on bothsides of the corridor because the only way for users to cross the corridor is through pedestrianbridges, which would require carrying the bicycle up ights of stairs to get to the other side.Therefore, the density of stations along the corridor is much higher than the recommendedratio. Conversely, in areas where bicycling is forbidden, such as a large park, there areno stations.

ITDP CHINA


60/152


Some systems includeapplications for smartphones that show stationlocations.CARLOSFELIPE PARDO

Care must be used in matching stationlocation to the cityscape. Stations are betterlocated in sunny spaces when possible, ratherthan under trees, so that the bikes dry off morequickly after it rains. This is also important ifthe system is solar-powered. Locations willneed to balance visibility of the system withintegration into the street environment. Often,larger stations in prominent areas are designedto stand out against their landscape, whilestations in residential areas are meant to blendin to the streetscape. Stations should not beplaced on footpaths unless there is sufcientclear space for walking beside the station. Ingeneral, a width of two meters of clear space forwalking is recommended in all locations, andmore space should be provided where there ishigher pedestrian trafc. At intersections, spaceis often more readily available on the minorstreet than on the main thoroughfare.

A variety of options for station locationsshould be considered:

• On-street parking spaces: Car parkinglocations are an ideal location for bicycleparking stations. In Paris, more than 1,450on-street parking spaces were removed tocreate space for 4,000 bicycles in the Vélib’system (Kodransky 2011). Similarly, Barcelonaconverted nearly 1,200 parking spaces for useby the city’s Bicing bike-share system.

• Space between landscaped areas or adjacentto other infrastructure: Space that is not usedoften by pedestrians such as in betweentrees or planter boxes or next to otherinfrastructure such as pedestrian bridges orutility installations can be used for bike-sharing stations without impeding pedestrianow.

• Dead spaces: Areas beneath yovers andbridges, which are often not utilized, can begood locations. These spaces may raise some

safety concerns, but those concerns can beresolved by proper lighting and good stationdesign. A bike-share station can transforma previously desolate space into somethingmore lively.

• Private property near large commercial andhousing developments: Bike-sharing stationscreate destinations, so private propertyowners can be convinced to give up privateland in exchange for the benets of having astation near their premises.


61/152


top Barcelona replacedparking spaces with bike-share stations.DUAN XIAOMEI

bottom Stations located onsidewalks are ne, as longas there is also plenty ofspace for pedestrians,such as at the station onAvenue Paseo de Reformain Mexico City.AIMEE GAUTHIER


62/152


63/152


Once the station locations have been chosen,the next decision will be how big those stationsshould be, including the number of bikesand the number of docking stations. This willdepend on the demand of the area, which canbe determined by several different methods:

• Conduct surveys at the transit stations to seewhere people are going and if they woulduse a bike to get there if the option were

available.

• Look at existing mode splits and majorattractions or points of interest that maycreate a higher demand.

• Crowdsource station locations to get an ideaof the demand for a particular area. This canbe done online by asking people where theywould like to see a station. It can also bedone in public areas in an installation wherepeople can mark on a map where they would

like to see a station.

• Hold community workshops to test stationlocation and get a sense from the communityof the local demand.

Fig. 6: Cycling Infrastructure Implemented Alongside Bike-share Systems

City Bike Infrastructure With Bike-Share System

Guangzhou, China 46 kilometers of segregated bike lanes

Paris, France 68 kilometers of segregated bike lanes, inaddition to 371 existing kilometers

London, United Kingdom 37.8 kilometers of 4 cycle superhighwaysBarcelona, Spain 150 kilometers of segregated bike lanes

Boston, United States 80 kilometers of segregated bike lanes(Kaiser 2012)

Rio de Janeiro, Brazil 300 kilometers of physically segregated(ciclovias), painted lanes (ciclofaixas) andsignalized shared routes (either with trafcor pedestrians)

To simplify the planning process, stations canbe dened into few key sizes, such as small,medium and large, so that each station sizeis not overly deterministic. Once demand isdetermined, the station size will then be thenumber of bikes per station multiplied by thedocking-space-per-bike ratio to determine thenumber of docking spaces at each station. Forexample, if the docking-spaces-per-bike ratiois 1.7 docking spaces per bike, a station that

needs ten bikes will need seventeen dockingspaces.

Using modular stations mitigates some of therisk of wrongly sizing stations, as it is easierto add or remove docking spaces once thesystem opens. See the next section for moreinformation.

. Station Sizing


64/152


65/152


Bike-share stationsshould provide customerswith information abouthow to use the system.In Rio, a placard offers

instructions, as well as amap of the system.AIMEE GAUTHIER

3.3.1 Manual vs. AutomatedSystems can be either manual or automated.In a manual system, an attendant records theuser’s information and helps with checkingin or out the bike, including payment. Thisinformation can be recorded on paper orelectronically. Automated systems are wherethe user checks in or out the bikes and makespayments electronically either at the terminalor kiosk or directly at the docking station. Thesetypes of systems often use specialized keycards for the users. The key difference is havingan attendant at the station who checks in andchecks out the bikes for the user. Some systemshave a mix, where larger or higher demandstations have an attendant.

Manual systems entail reduced initial capitalcosts compared to automated systems, butlong-term operating costs are higher, andsystem reliability suffers. Proponents of manual

systems argue that having staff at stationsgenerates better service, reduces theft andvandalism, and requires less technologicalcomplexity at stations. Manual stations areused in systems of various sizes in places likeBuenos Aires (although the city is moving toan automated system, keeping some largerstations as manual), Santiago, and Medellín.These are very basic types of stations that needonly a simple locking mechanism for the bicycle(if the bikes are locked at all) and dependexclusively on an attendant. Their designs

can be simple: some, such as the system inBuenos Aires, use specially designed freightcontainers; others, like Santiago's system,have no signicant infrastructure other thana large horizontal pole to hang the bicycles.These basic stations are obviously the easiestto maintain and least expensive.

Automated stations are more complex indesign, installation, and maintenance thanmanual stations. The capital costs will be

higher than those of manual stations, butthe operating costs over time will be lower.Automated stations are more secure and donot need staff at stations. Their design is moresophisticated in that it must include specicallydesigned docking infrastructure to lock thebicycles and technology that allows wirelessinformation to transfer from the docking

spaces and terminals in order to facilitatethe checking in and out of bikes. Instead ofattendants, the stations have a terminal thatgives users information, accepts payment,and allows for checking bikes in or out. Atleast initially, however, automated terminalscan be confusing, and can thus deter peoplefrom using the system. London addressedthis problem by employing staff for an initialinaugural period to instruct people on how touse the system. However, terminals are notnecessary, as the docks alone can allow for bike

check-in or return.While manual systems require an attendant

at all stations, automated systems may alsowant to have a staff person at certain largestations for customer-service reasons. Thoughmanned stations are cost-prohibitive in someeconomies, manned stations are desirable inmany developing economies because of the jobcreation, security and added customer servicethat accompany them.


66/152


67/152

Buenos Aires opened its bike-share in 2010 with 100 bikes, and by 2013had expanded to thirty stations and 1,200 bikes. The system is a manual

system using specially designed containers as stations. In rst half of 2014,the system will expand again, reaching 200 stations and 3,000 bikes, and itwill transition to a mixed system, with both manual stations and automatedstations. The city decided to make this change because larger systems aremore easily managed through automated stations. However, the stationswith the highest demand will remain manual for the time being. Automatedstations will be open 24 hours a day, while the manual ones will close atnight. (City of Buenos Aires 2013)

Buenos Aires:Moving from Manual to Automated

In the rst phase ofBuenos Aires’s bike-sharesystem, users checked out

bikes from an attendant.ITDP ARGETNINA


68/152


3.3.2 Modular vs. PermanentTwo main types of stations are modularand permanent. Modular stations are easilymoved, usually constructed on a base thatis then bolted into the concrete or asphalt.Those stations require solar power. Permanentstations require excavation and trenching toreach the power source. This requires a longertime frame to implement and may entail a moreonerous approval process.

The most exible type of automated stationis the one that was introduced in Montreal’sBixi system and is now used in other citiessuch as Washington, D.C., and Melbourne. Itconsists of a heavy base with docking locationsand a terminal for information/registration/payment, but it can also be relocated. Thestation is bolted into asphalt or concrete, butuses solar power and thus does not need to beconnected to an underground power source.Once a station is built, if its location is foundto be inadequate—as is sometimes discoveredafter some weeks of operation—the stationcan easily be relocated to a place with betterdemand. Stations like this are also more easilyscaled up or down, adding or removing dockingspaces as real usage is determined afteropening.

New York City chosemodular stations thatwere quicker and easierto install — both the docksand terminals are onplatforms that connectto each other.NYCSTREETS (CREATIVE COMMONS)


69/152



70/152


top Stations in Lyon arepermanent, meaning thatinfrastructure is installeddirectly into the ground orpavement.KARL FJELLSTROM

bottomSolar panels on Bixistations in Montreal

power the stations thusnot requiring excavationto power lines.MAX HEPP BUCHANAN

opposite topShenzhen’s bike-sharesystem uses individualdocking stations thatusers insert the wheel ofthe bike.KARL FJELLSTROM

opposite middleThe bike-share stationsin Santiago, Chile, aremanually operated,where the bikes do notlock to the rack andare monitored by anattendant.CARLOSFELIPE PARDO

opposite bottomBicycles in Beijing arekept in a secure bikeparking area to savespace.ITDP CHINA


71/152


3.3.3 Docking StylesIn automated stations, there are two basictypes of station design that accommodatecheck-in and check-out: docking spaces andcycle parking areas. Which type works bestdepends on the needs and location of thestation:

• Docking spaces:Each space docks one bicycle. The numberof spaces determines the size of the station’sfootprint, which means there is a great dealof exibility in adjusting the station size tot the existing urban landscape. This styletakes up more space per bike than cycleparking areas, but blends in better with theurban environment. Bicycles are checked outby customers either at the terminal or at theactual docking space.

• Bike parking areas: Bicycles are stored together in a securedarea, on racks. Cycle parking areas are agood option for larger stations—i.e., morethan 50 bicycles—because cycle parkingracks can hold more bikes per square meterthan docking spaces. At stations with cycleparking areas, bicycles are checked in andout through a turnstile or manually. Becausethese stations require a secure area thatis fenced or walled off, they can be moreintrusive in the urban landscape.

A station can use individual, stand-alonedocking spaces, as in Paris, or it can use adocking bar that joins the different spacestogether, as in Mexico City and Washington,D.C. Individual docking spaces are perceivedto integrate better into the urban environmentand are more porous—unlike docking bars,individual docking spaces do not create anobstruction that pedestrians must walk around.


72/152


Other considerations for docking spaces

include whether the user rolls the bike intothe docking space to lock it into the dock, orwhether the bike must be lifted up.

A system may incorporate both station types,depending on demand levels, desired streetviews and availability of space at a particularstation. While docking stations are popularfor roadside stations, bicycle parking areasare best utilized in underused spaces, such asthose beneath overpasses or in suburban areas,where land space is not as precious as it is incities. Whether a system uses docks or parking

areas, stations should always have moredocking positions or storage space than bikesin order to accommodate peak demand. Thisshould be reected in the station’s docking-space-per-bike ratio.

topNew York City chose toinstall some docks wherethe road was very narrowto save space.AIMEE GAUTHIER

bottomLyon, France, usesindividual dockingstations where the lockingmechanism is on the sideof the frame of the bicycle.LUC NADAL


73/152


topIn Rio de Janeiro’s system,bikes lock into the sideof the docking bar, ratherthan the top.AIMEE GAUTHIER

bottomThe docking system forEcobici requires the userto line up the bike so itcan be properly insertedinto the docking bar.AIMEE GAUTHIER


74/152


. InformationTechnology Systems andPayment Mechanisms

Information technology (IT) forms the nervous system for the bike-sharesystem, connecting the individual stations, users and control center usingsoftware and data-transmission mechanisms. Decisions that must bemade relating to IT include deciding how customers register and pay forthe system, how bikes are checked in and out from the docking spaces,and how information is transmitted both internally for management and

externally for the customers.The software needs to support the front end, or the public side, of the

system, including registration of new users, payment and subscriptions,general information about the system, and customer data management.The front end of the IT system can include website portals and apps forsmart phones. On the back end, where the implementing agency andoperator receive the information required to run and manage the system,the software needs to support station monitoring, redistribution of bikes,defect and maintenance issues, billing, and customer data.

Most systems use card technology (smart cards, magnetic cards,or credit cards) to check bikes in or out. The second most populartechnology is locks that use codes to release the bikes. Some systems

are manual and do not require any technology to release or return thebikes. A few systems use keys.

IT will need to serve two types of users: long-term users—who areusually registered members and use the system with some frequency—and casual users, such as tourists, who use the system infrequentlyor even just once. Long-term members can be given access cards andcan place deposits to use the system. Casual users will not be able touse the system if a special access card is needed or if there is no wayto guarantee the return of the bicy

2014 itdp bikeshare planning guide

Documents