Public Comment

Commentary: An Analysis of Bus Rapid Transit

By Wolfgang Homburger
Friday September 28, 2007

Bus Rapid Transit (BRT) has been argued and debated ever since AC Transit unveiled a proposal for a BRT project between Berkeley, Oakland and San Leandro. The subject has polarized the community into pro-BRT and anti-BRT factions—and, of course, those who have never heard of it. It is therefore timely to provide some guidance on how to analyze this proposal—and others like it.  

This article consists of two parts: the first defines and describes BRT from the points of view of the passenger, the bus operating agency, and the general public. For more details on BRT generally, look up “Bus Rapid Transit” at The second part considers the AC Transit proposal specifically. For more details about this project see the draft environmental impact statement (DEIS) at The DEIS is dated May 2007. 


Minimizing Travel Time—That’s What It’s All About 

The passenger: When we travel, the trip is not the purpose; getting to the destination is what matters, and we try to minimize the time it takes. We also often make allowances for being delayed if the highway might be congested, the bus might be late, or other possible events make the trip time estimate unreliable. In urban transit systems, “trip time” consists of: 

• The access time—how long it takes us to get to the bus stop and, at the other end, to walk from the bus stop to our destination.  

• The waiting time at the stop for the bus arrival. 

• The in-vehicle time—the number of minutes we are on the bus.  

Reliability is a measure of how well the bus can keep to its schedule; high reliability means that passengers can arrive at the stop just before the bus gets there; low reliability increases waiting time because the bus may be late. 

BRT can improve both in-vehicle time and reliability by providing the bus some or all of the following: exclusive lanes or roadways, fewer stops, priority at signals, fast boarding and alighting via platforms with a height that is even with the bus floors and/or use of low-floor buses, and fare payment off the vehicle. But BRT is likely to add to passengers’ access times, because the longer spacing between stops imposes longer walking distances to and from them. The result, as we learned from studying trips made on BART with its long station spacing and high speed and reliability, is that persons with long trips are attracted, but those making short trips remain on the local service or in their cars. 

The bus operator: The cost of operating a bus route consists of drivers’ wages and benefits, fuel, maintenance, and insurance. The driver typically accounts for 75 percent of the total. Drivers are paid by the hour, not by the number of miles they drive. The faster a bus can travel, the more productive the driver will be. (It costs somewhere upwards of $100,000 per year to operate, maintain, and insure a bus.) It is therefore in the interest of the bus operator to increase speeds and minimize time spent at stops. The latter can be achieved by allowing passengers to enter and exit through any door, paying their fares—if they do not have monthly passes—at ticket vending machines at the stops. 

If reliability is low, some buses will arrive at the end of the line late. To avoid these vehicles leaving late on the return journey, the operator schedules “recovery time” at the terminal. This will often require extra buses to be assigned and increase operating costs substantially. 

It is therefore in the bus operator’s interest to welcome BRT installations. Even accounting for minor items, such as the costs of maintaining bus stops, ticket vending machines, and electronic next-bus-arrival signs, the operating costs will be reduced. (Modifying and maintaining the new traffic signal hardware—described below—may be paid by the city.) 

But then there are one-time capital costs. BRT infrastructure can range from relatively minor investments (San Pablo Avenue locally, Wilshire Boulevard in Los Angeles) to hundreds of millions of dollars (bus freeways in Pittsburgh, Penn., transitways in Ottawa, exclusive roadways with guide rails in Adelaide). Capital funds will be needed for new buses, and for expansion of maintenance and overnight bus parking facilities. 

The general public: Little needs to be said here; the ongoing debate has already brought many issues to the surface. On the plus side there is the new option of a fast, reliable service that might entice some present motor vehicle drivers to switch to public transit. On the negative side may be an adverse impact on street traffic, on parking, and perhaps on retail establishments that depend on nearby parking. Another negative factor could be the postponement of other transportation improvement projects, if capital funds were diverted from them to BRT.  


The Route 1R BRT Project:  

The AC Transit Board of Directors has before it the BRT project draft environmental impact statement (DEIS), which offers four alternatives. The board will select one of these as the locally preferred alternative (LPA) next April. The numerical data quoted here is the range of values over all four alternatives. The BRT project is proposed between downtown and UC Berkeley and San Leandro or Bay Fair and includes: 

• Two center traffic lanes converted for exclusive BRT use on Telegraph Avenue and International Boulevard/East 14th Street. 

• Stops spaced a quarter-mile to a half-mile apart, each furnished with two boarding platforms, a shelter, benches, ticket vending machines, real-time next-bus-arrival signs, and security equipment (cameras?). These stops will be in the center of the street; their width will require the traffic and bike lanes to be brought to the curb, eliminating parking at these locations. 

• Preemption hardware at traffic signals and in the buses to be used in this service. This hardware consists of a radio transmitter in each bus and a receiver at each signal. When a preempt message from a bus is received, the signal either holds the green if it is showing at this moment or switches to green as quickly as possible. 

The new bus rapid route 1R will operate on this facility at headways of 3.6 to five minutes in peak hours and less frequently at other times. Two alternatives also include a local Route 1 that runs in the remaining traffic lanes; the other two do not. 

The DEIS estimates that in 2025 total passengers in the corridor would range from 42,050 to 49,230 per weekday , and that of these from 4,580 to 9,230 would be new passengers after shifts from other routes have been accounted for. 

Capital cost would be from $310 million to $400 million in 2005 dollars. The cost of new buses is included, but the need for, and cost of expanded maintenance and storage facilities is not mentioned. (Only about $100 million had been earmarked for the project by last May.) Operating and maintenance (O&M) costs would range from $35.5 millions to $39.4 millions per year in 2025. Fares would cover from 20 to 28 percent of O&M costs; the rest would have to be subsidized. 



The benefit from this project would be that the better bus service on the Telegraph/International Boulevard route, would attract the new passengers, some of whom would have previously traveled by car. But there are many troubling aspects or disbenefits.. 

• Investment Priorities. Is the selected corridor really in such bad shape that a large investment is urgently needed? BART parallels the BRT route within less than a mile throughout the corridor and, while it has fewer stops, will continue to attract long trips because of its higher speed and reliability. Are there no other corridors where BRT or other urban transit investment would be more valuable? 

• Facility Utilization: The dedicated lanes would be utterly underutilized at 1 bus every 3.6 to five minutes or 12-17 buses per hour per direction. It is highly likely that the empty lanes would soon be opened to carpools and hybrid automobiles, following the precedent that allows these vehicles on what were initially exclusive bus lanes on California freeways. 

• Costs: The capital costs estimated in the DEIS are in 2005 dollars and are certain to inflate by the time construction is started in view of continually escalating construction costs. Only about $100 million had been identified as available in the DEIS six months ago. Subsidies to cover 74-80 percent of the annual O&M costs must also be found.. 

• Patronage Forecasts: The patronage numbers shown in the DEIS are not very impressive. The ratio of passengers to dollars of investment will be the lowest of any BRT system in North America.  

• Traffic forecast. Substantial reduction in motor vehicle traffic is, I believe, wishful thinking. The city of Berkeley has for many years had a policy of encouraging commuters working at the University and downtown to use Telegraph Avenue en route from or to the Route 24 freeway and the Caldecott Tunnel. This was done to relieve congestion on alternate routes, such as College Avenue and Warring/Derby. Route 1R offers nothing to commuters living in Central Contra Costa County, unless they are willing to detour via 40th Street/Telegraph and MacArthur BART. Loss of a pair of lanes on Telegraph will increase congestion and the anger of residents on parallel streets where backups are already formidable. The DEIS identifies 27 intersections (four in Berkeley) where traffic demand will exceed capacity.  

• Traffic Flow and Safety. Restricted to one lane per direction, traffic would flow in a constant stream during peak hours with pedestrians unable to cross except at signalized intersections. Local Route 1 buses (only in two alternatives) would be in this stream and would block its progress every time they stop to discharge or pick up passengers. Left turns would have to be prohibited at most intersections to prevent turning vehicles from blocking the BRT roadway. Vehicles would infiltrate residential streets and make three right turns to go to the left. In sum, traffic would be slow—stop-and-go at times—and exhaust emissions would increase. 

There is also a potential safety problem at each BRT stop, because passengers must cross a traffic lane when walking to or from the loading platform. 

• Parking: The loss of parking spaces along the route is, perhaps, the most contentious issue of all. Merchants always express concern when a single space at the curb is lost. The space needed for the remaining traffic lane and probably the bike lane to pass by each stop platform can only be gained by removing parking. The DEIS shows that from 945 to 1,300 spaces would have to be moved or lost. A proposal to replace them, together with their meters, on side streets, most of which are residential, is understandably unpopular. 



What, then, might be done in this corridor? Modest investment to further improve Route 1R is warranted; the 72R Project on San Pablo Avenue is a good guide to follow. Use low-floor buses exclusively and equip them and traffic signal controllers with preemption hardware. Make such other traffic engineering improvements as will assist buses to pass chronic bottlenecks—this will also benefit local buses. Install next-bus-arrival signs at selected stops. Introduce Proof-of-Payment fares, so that all doors of buses can be used by entering as well as exiting passengers, thus reducing time stopped at each bus station. And delete the exclusive roadway and its center-of-the-road stations from the project. 

If, after this limited BRT program has been fully implemented, the experience with Route 1R is satisfactory, some of the money saved can then be used for other projects; e.g., BRT in the MacArthur Boulevard corridor, an extension of Route 1R to Amtrak at Third Street and University Avenue. Such a program would fit the most cost-effective components of BRT to the demand and geography of the inner East Bay, would cause minimum disruption to commerce and traffic in the corridors that BRT traverses, would be much more likely to be fully funded, and would meet most of the goals that the project addressed at the outset. And—perhaps most important of all—there would be no reason for opposing it.  


Wolfgang Homburger was a research engineer at UC Berkeley’s Institute of Transportation Studies from 1955 to 1990, and assistant director from 1984-to 1990. He specialized in traffic engineering and public transportation systems.