San Francisco Studies New PAX Commuter Rail Tunnel

San Francisco County Transportation Authority (SFCTA) is studying the feasibility of tunneling to extend commuter rail service underground into downtown San Francisco. Delve Underground led a team of consultants conducting the study from 2020 to 2022. The Salesforce Transit Center (STC) was completed in 2018 and will serve as a new transit hub, including a commuter rail station after two future connecting tunnel segments are completed. The northern segment, known as the Caltrain Downtown Extension (DTX), will extend the rail line from its current terminus at 4th and King Streets to the STC, and has advanced into preliminary engineering. The segment south of DTX is a 1.6-mile (2.6 km) tunnel known as the Pennsylvania Avenue Extension (PAX), which is currently in the pre-environmental planning stage.

The existing Caltrain surface rail in the PAX area creates a physical and visual barrier that adversely impacts neighborhood connectivity, including local street bifurcations and two at-grade rail crossings. Prior studies recommended undergrounding this section of track to improve neighborhood connectivity and public safety.

All PAX tunnel alignments studied will be in mixed ground ranging from soil to bedrock, all below the groundwater table. Pressurized-face tunnel boring machines (TBMs) and the sequential excavation method (SEM) were considered feasible tunneling methods.

Three shortlisted alignment alternatives were selected:

Alternative A: Long Alignment

• 1.5-mile (2.4 km) TBM tunnel(s), single-or twin-bore configuration.
• Bypasses existing 22nd Street Station. New underground station, or no station at 22nd Street.

Alternative B: Mid-Length Alignment

• 0.9-mile (1.4 km) TBM tunnel(s), single- or twin-bore configuration, with short SEM connections to existing track in Caltrain Tunnel 1.
• Connecting TBM bored tunnel to existing track just north of the 22nd Street Station allows use of reconfigured 22nd Street Station.

Alternative C: Short Alignment

• Hybrid alignment: northbound track in open-cut trench; southbound track in 1-mile (1.6 km) TBM tunnel.
• Allows continued use of existing 22nd Street Station with minor modifications.

Several significant challenges and constraints impact PAX:

The vertical alignment is controlled by the track connection points with the planned DTX grade line at the north end of PAX, the existing Caltrain tracks near Cesar Chavez Street at the south end, and the 2% maximum grade in between the two portals. The preferred minimum ground cover above the tunnel crown is one tunnel diameter. There are several locations where this minimum cannot be achieved and ground modifications prior to tunneling is one solution under consideration. Complicating matters on the vertical alignment is the need for PAX to dive beneath existing deep, large sewers as described below.

Construction sequencing at the north end of the project is especially challenging. The DTX section will be brought into revenue service while PAX is under construction. This will require ramping the DTX tunnels up to the surface to connect to existing tracks south of DTX while the adjacent PAX tunnels are under construction.

For Alternatives B and C, SEM mining will connect the new PAX tunnel into the existing Caltrain Tunnel 1 just north of the 22nd Street Station as depicted in the figure below. With SEM or TBM mining breaking into the existing tunnel immediately adjacent to live track, it is anticipated that this work will be performed on nights and weekends during off-service hours.

At the south end of the PAX alignment, Alternative A will exit out of a new portal or portals just west of an existing abandoned tunnel. Alternative B will make use of a rehabilitated Tunnel 2 before this southernmost tie-in. Exact conditions in the existing tunnel are not known at this time, and it would need to be protected from damage during the new construction.

For twin bore Alternatives A and B, TBM mining of parallel tunnels under 7th Street will have to be performed with a separation between tunnels (pillar width) as little as a few feet because of right-of-way (ROW) boundary restrictions and existing deep foundations that flank the twin tunnel corridor. Ground modifications to enhance strength and stiffness of soil within the zone of the pillar will be required.

Under 7th Street, the PAX tunnels must pass below a 40-foot-wide (12 m), pile-supported box sewer and a future 12-foot-inside-diameter (3.7 m) stormwater tunnel while maintaining the maximum 2% grade in rising up toward the DTX interface. Existing I-280 is an elevated viaduct with deep foundation elements along most of 7th Street where the tunnel is planned. All of the alternatives narrowly avoid the I-280 foundations.

The cost range of the alternatives falls between $1.67 billion (Alt C) and $1.99 billion for (Alt A), with an estimated duration of 9.5 to 11.5 years. The alternatives were evaluated using a framework matrix to qualitatively assess a broad set of criteria. This framework yielded a close range in overall weighted scoring. Whether a station should be included within the PAX project area and, if so, where it should be located, is one of the more significant aspects of forthcoming PAX work. Cost will be a major factor.

It is anticipated that the next four years will be spent on alignment and station selection, advancement of preliminary engineering, securing project financing, identifying procurement and delivery methods, ROW acquisitions, and environmental clearance.