4Demand Analysis

The essence of mathematics is not to make simple things complicated, but to make complicated things simple.Stan Gudder, mathematician

The analysis of the potential customer demand for the planned BRT system is the foundation for most of the subsequent planning, design, and financial work. Demand estimates are critical to designing the system, planning operations, and forecasting the financial viability of the new system. Knowing where and when customers require transport services will help shape a system that is based, above all, on the needs of customers.

It is often tempting for a decision maker to want to put a new BRT system on a wide road or a ring road where there is plenty of space, but where demand is limited. Other times, decision makers will choose BRT corridors for political reasons, like putting one BRT corridor in each district, regardless of the relative importance of the corridor to riders, or locating the BRT system where its benefits would accrue to politically powerful people. In the United States, sometimes planners put BRT in places where they hope it will stimulate economic development, but where this development may be a decade away and current ridership is too low to justify the investment. While such factors will inevitably be a part of the decision-making process, BRT planners generally suggest putting a BRT system in a location that will benefit the most customers in the best way possible as quickly as possible, most directly through time savings.

The first use of demand analysis, discussed in Chapter 5, is generally to find out where the existing public transport demand is concentrated, and from this to extrapolate where the greatest potential BRT demand may be. The second use is to rank the existing public transport demand on all the main public transport corridors so that a full network of BRT corridors can be identified and a rough phase-in plan can be proposed.

Note that as a practical matter, many cities go into a BRT planning process with one or a few possible BRT corridors already identified. In this case, carefully reviewing the municipality’s process of determining the proposed corridors can save BRT planners a lot of time and money because they can focus on a few select corridors, rather than all the main public transport corridors in the city, unless all of the proposed corridor options are poor.

Demand estimates play several key roles in the design of a good BRT system. First, the system needs to be designed with enough capacity to handle future demand while maintaining high vehicle speeds. Second, the demand estimate is also needed for creating the service plan and optimizing operations. Third, the demand estimate is critical for financial projections. For this, the demand estimates have to err on the conservative side to be credible to banks and investors. The critical factor is that the banks and investors trust the estimates, and for this the greater the accuracy of the projection, and the more methodologically credible, the better.

This projected future demand should start with an analysis of existing public transport demand, and then be expanded with reasonable expectations about customer growth. To be conservative, the system needs to be designed with sufficient extra capacity to ensure good performance when the system opens and sufficient additional capacity for at least a decade of future demand growth.

As discussed in this document, achieving capacity while maintaining high speeds depends on three main factors: the design of the infrastructure, the type and number of vehicles, and the organization of the services. It is easier to increase capacity by adding new vehicles and changing their schedules than by modifying the infrastructure, including right-of-way, junctions, stations, and terminals. If the system is designed with more capacity than it needs, it will be needlessly expensive and consume too much right-of-way that might otherwise be used for footpaths, cycle ways, public space, parking, or private vehicles. Alternatively, if capacity is too low, BRT stations will be overcrowded, and vehicle speeds might even be slower than current speeds, thus alienating customers. Any of these mistakes will significantly compromise the success and quality of service, the profitability of the system, and its ability to meet future demand levels.

This chapter outlines a step-by-step approach that provides a gradually better demand analysis as the process evolves. The topics include: overview of demand analysis; data collection; basic methods for estimating public transport demand; estimating demand with a public transport model; estimating demand with a full transport model; and risk and uncertainty.

Section 4.1 provides an overview of demand analysis—what is required and what is to be taken into consideration to properly accomplish it. It also stresses the need to gain a thorough understanding of current operations before attempting to improve on them.

Section 4.2 discusses the data collection effort required to support demand estimation, as it is necessary to gain a solid base for forecasting.

Section 4.3 describes a methodology to provide a rapid demand assessment. Rapid demand assessment will provide an approximate idea of likely BRT demand on major corridors using only traffic counts and occupancy surveys in key locations, accompanied by boarding and alighting and bus-speed surveys. With this information alone, a skilled BRT planner may be able to develop a reasonable demand estimate, but the BRT planner’s previous experience is paramount.

Section 4.4 discusses how to estimate demand more accurately as a transfer from other public transport modes; this is achieved using a public transport model. The public transport model simulates only the public transport system, and can be strengthened with the addition of a customer origin and destination survey. With a basic public transport model, most critical decisions about the BRT system and many critical operational decisions can be made, but the public transport model can only roughly estimate impacts of the system on mixed traffic and on modal shift. Most BRT planners, including the team that designed TransMilenio, primarily use a public transport model.

Section 4.5 discusses the basics of developing a multimodal transport demand model for BRT. Such a model will provide full flexibility for testing multiple routing and pricing scenarios, a more robust estimate of plausible modal shift, emissions impacts, bus route optimization, and a host of other useful tools.

Finally, all forecasts are subject to uncertainty and risk, and decision makers must understand these when planning a BRT system. The two main risks in demand estimation are the overestimation of demand, thus requiring more infrastructure and fleet than necessary and exaggerating revenues; and the underestimation of demand, thus limiting the performance and success of the BRT system. Section 4.6 discusses these issues and suggests a few ways to handle them.

Contributors: Walter Hook, BRT Planning International; Karl Fjellstrom, Far East BRT; Luis (Pilo) Willumsen, consultant; Arthur Szász, Protocubo; Ulises Navarro, ITDP Latin America; Pranjali Deshpande, ITDP India