By Patti Meagher  |  Photos By Courtesy UC Transportation Center

The state’s first carpool lane was installed on the San Francisco–Oakland Bay Bridge in 1971 to encourage carpooling and move more people more quickly. But Bay Area traffic congestion, up by 6 percent last year according to transit officials, still gets worse every year; and carpoolers complain that life in the fast lane is not fast enough, due to pokey hybrids and scofflaws.

Two Berkeley Engineering professors are delving deep into freeway physics to find out how carpool lanes—also known as HOV (high-occupancy vehicle) or diamond lanes—affect traffic flow. Both researchers studied six of the Bay Area’s most notorious stretches along Interstates 80, 237, 880 and 101 and came down squarely on opposite sides.

“As currently operated,” says electrical engineering and computer sciences professor Pravin Varaiya, “the HOV system does not meet its goals.” There are exceptions, he admits, such as the heavily metered chokepoints on the Bay Bridge, where private cars carrying three-plus passengers and transit vehicles can zip by toll free. Otherwise, his study concludes, carpool lanes so insignificantly reduce congestion and travel time that they do not motivate solo drivers to carpool.

Mike Cassidy, civil and environmental engineering professor and acting director of the Institute for Transportation Studies, disagrees. While he did identify some HOV lanes that are underutilized or poorly located, Cassidy concluded that carpool lanes are effective overall in reducing travel time in carpool lanes while not significantly increasing travel time in regular lanes. Where the two researchers differed, Cassidy says, is the type of data they captured.

“Traffic is a spatial and temporal phenomenon,” he explains. “If you’re going to analyze problems, you have to look at it spatiotemporally, over several-mile-long stretches.”

Both researchers used the California Freeway Performance Measurement System—a public database of input from 26,000 loop detectors, or sensor points, throughout the state’s freeways—to analyze traffic flow, speed and other factors. Focusing on specific points, Varaiya found that carpool lanes are choked by slow-moving “snails” and violators darting in and out to dodge backups. He advocates converting HOV to “HOT” lanes, high-occupancy toll lanes for transit vehicles that private drivers can access by paying a toll.

Cassidy examined consecutive points along stretches of up to 15 freeway miles to pinpoint where and how congestion first manifests and used video to analyze bottlenecked traffic. He found that backups in carpool lanes were a natural consequence of increased traffic volume at peak hours or were caused by accidents or construction up ahead. Cassidy’s remedy is to more carefully study proposed HOV installation sites based on specific traffic patterns in each area.

At stake is a $21 billion Caltrans plan to double the number of HOV-lane miles statewide by 2020. Caltrans claims that HOV lanes do work, that carpooling is up, and that the state’s ever-growing population inevitably causes increased congestion. Despite these claims, the Federal Highway Administration cited California for not complying with federal guidelines requiring a minimum average speed in carpool lanes of 45 mph during 90 percent of peak hours. In response, Caltrans is taking measures like limiting HOV access in severely congested stretches and beefing up enforcement of HOV violators.

Both Berkeley researchers agree that HOV lanes merit further study. “The persistence of debate,” Cassidy says, “suggests that the impacts of HOV lanes are not fully understood.”