Seismic retrofit for BART
Seismic retrofit for BART Mar 2004
Shani Wallis, TunnelTalk
Seismic preparedness in California has found urgent concern for BART, the Bay Area Rapid Transit system in Northern California. Several vulnerabilities to the region's ever-threatening seismic activity have been found with concern in particular for the system's Transbay Tube. The consequences of any damage repair outage or rupture of the structure would have by far, the greatest impact. TunnelTalk reports the causes of concern and the program being developed to mitigate potential seismic effects.

Plan of the BART system and (below) section of its San Francisco Transbay crossing


At 3.6 miles (5.75km) long, the steel-hulled, reinforced concrete-lined crossing of San Francisco Bay remains one of the longest immersed tube tunnels in the world and one of its longest continuous rigid structures. Built in the 1970s, the Transbay crossing comprises 57 elements of about 330ft (100m) long, all welded together and attached either side to the San Francisco and Oakland vent structures. Inside there are two running track galleries and a central emergency / service gallery. Designed originally by Parsons Brinckerhoff/Tudor /Bechtel Joint Venture (PBTB), the tube lies up to 100ft (30m) below the' Bay's mean sea level and was built according to state-of-the-art seismic understanding of the time.
Since then several potentially disastrous seismic concerns have been identified. Among them, the granular backfill around and beneath the tube is now known to be susceptible to earthquake-induced liquefaction that would cause serious instability. Another is diminishing displacement capacity of the innovative 'seismic joints' incorporated into the structure. These joints, for which Parsons Brinckerhoff retains a design patent, allow 6 degree-of-freedom movements (three translations and three rotations) between the long rigid immersed tube tunnel and the fixed ventilation structures. Coupling sets of restraining cables, 60 in each of the tube's four joints, allow maximum design movement of 4.25 inches of longitudinal (push/pull) motion and 6.75 inches for lateral (right/left) and vertical (up/ down) motions (Fig 3).
Over time some of the joints have moved and are reaching their full design capacity. Most serious is the San Francisco Bay-side joint which has only about 1.5 inches remaining in the push direction and 1.25 inches in the downward direction.There is little latitude left for further strain.
The US Geological Survey how-ever, predicts a 62% probability of a major earthquake in the Bay Area before 2030.It is this that has concentrated the efforts of most public infrastructure owners in California to assets, the consequences of any damage outages and the plan for prudent retrofit.
For BART,its Earthquake Safty Program started in September 2000 with a Seismic Vulnerability Study, and has moved recently into work funded in part by the Federal Highways Administration (FHWA) to strengthen aerial trackway bridges over major regional streets and highways.
For the Earthquake Safety Program, BART has engaged Bechtel as general engineering consultant with HNTB the principal subconsultant, and a team of 10 additional sub consult ants including ICEC, MGE, Geomatrix, Parikh Geotechnical and the Allen Group to provide engineering, geotechnical, seismic survey, soil engineering, community outreach, vibration monitoring, architectural, landscaping and other necessary specialist services.
Under the proposed retrofit program the original 72-mile, 34-station BART network with its subsequent 30-mile, 9-station ex-tensions, is divided into sections for at -grade and aerial routes and the underground sections.
One underground section is the 3.2 mile (5km) long twin-tube Berkeley Hills bored tunnel in Alameda County that crosses the Hayward Fault. This is considered one of the most likely sources of earthquake activity in the Bay area over coming years and like the Claremont water supply tunnel, about 750ft (230m) to the north, the BART tunnels could be ruptured or seriously damaged by a large event and offset of the Fault.

Cross section of the patented seismic joint either side of the two ventilation structures of the Transbay Tube that allows for limited translation and rotational movement

Top priority however is the central alignment from the west portal of the Berkeley Hills Tunnel, through the TransbayTube, to the Montgomery Street underground station in central San Francisco. "This is the vital segment of the core system," said Tom Horton, BART's Group Manager for the Seismic Retrofit Capital Program. "More than 150,000 passengers, or nearly half the total BART ridership, use the Transbay crossing each day. At peak periods, there can be as many as six 1 O-car trains in the tube at the same time. With 100 people/car x three lO-car trains/ direction - that's about 6,000 people in the tunnel at once. Should there be a large magnitude event in the area the crossing would be a potential safety issue and at least closed for two years optimistically, to twoand-a-half years for predicted damage repair." Rupture would be of greater consequence."We must avoid causing alarm or panic among the riders of the network," continued Horton, "but at the same time we must stress to our Board and other officials that we do have cause for major concern and that we must act to address this concern without delay."
Causes of concern
Liquefaction of the backfill material is a concern in that it would cause sections of the tube to become buoyant and to rise. Reduced friction between tIle tube and backfill material would also cause added movement at the four seismic joints. An additional concern is instability and potential movement of the San Francisco shoreline slope behind the vent structure. Another is the opening and closing of joints under the impact of traveling seismic waves with near-fault large velocity pulses - "an area that is only in the early stages of understanding and study", said Wen S. Tseng, President of ICEC (International Civil Engineering Consultants Inc) for the design team.
"Spatial variation of seismic ground motions along the tube length will cause deformations of the tube and movement of the tube ends relative to the surrounding free-field soil,"said Wen. "Such movements are especially critical in the tube's longitudinal direction due to its high axial rigidity. Studies show large relative longitudinal movements of up to 9 inches at the San Francisco end of the tube."
The liquefaction fear for -the backfill was explained by Ching Wu, Bechtel's Project Manager for the Earthquake Safety Program when TunnelTalk visited with the team managers in January in Oakland. "All backfill materials are granular and were placed underwater without any subsequent compaction," he said. "Should these expectedly loose materials experience liquefaction, support and stability of the tube would be seriously undermined."
The vent structure at the foot of the San Francisco shoreline slope is a large 120ft (36.5m) tall, steelplate-wrapped, reinforced concrete rectangular caisson sunk more than 75ft (23m) into deep, soft, backfill soils and founded on relatively weak Bay Mud. A deepseated soil slope failure with associated down-slope movement would exert large lateral soil pressure on the structure and would further stress the seismic joints either side. "Furthermore, studies reveal that nonlinear base sliding and uplifting displacements of the structure relative to the surrounding soils could take place during an earthquake adding demands on the seismic joint again exceeding its ultimate capacities," saidWu.
Retrofit concepts
Judging the probable effects of a potential earthquake and calculating the type and amount of preemptive retrofit work necessary is a subjective process. "Of course it is not feasible to do all we would like to," said Horton. "Our proposal therefore is to develop prudent measures to mitigate the worst case scenarios and to implement cost effective retrofits that would allow restoration of service in a timely manner." To help assess the need and balance need to cost, BART Directors instructed the Program team to solicit advice, comment and opinion from as many of the current experts in the field as necessary. In addition panels of national and international experts have been established to review all aspects ofthe project (Table 1).
Table 1. Design Review Board Ben Gerwick
Thomas O'Rouke Professor, Cornell University  
Ray Seed Professor,Peer Review Panel  
Jack Moehle University of California, Director PEER Center  
Craig Comartln Structural Engineer  
Norm Abrahamson Seismologist  
Po Lam Geotechnical Engineer  
Roy Imbsen Bridge Design Consultant  
Stephen Thoman Bridge Design Consultant  
Stephen Mahin Professor, University of California Berkeley  
William Holmes Structural Engineer  
Daniel Shapiro Liaison California Seismic Safety Commission  
Fadel Alameddine Liaison Caltrans  
Peer Review Panel members added to address the Transbay Tube  
Ed Idriss Professor, University of California Davis  
James Mitchell Dean, Virginia Tech University  
Having completed the Vulnerability Study, the team has adopted a Systemwide Safety, Core System Operability Program that specifies the minimum retrofit improvements required to allow resumption of BART operations in the most critical portions of the system as quickly as possible after a predicted design basis event. A similar retrofit program for the two much smaller immersed tube road tunnels under the Oakland Estuary is providing additional practical information and experience. Parsons Brinckerhoff, Quade & Douglas is leading that vibro-replacement measures design team for owner Caltrans.The first tube is also a rigid welded steel hulled structure built in 1928 while the second, built in 1958, is fitted with flexible construction joints.
All proposed retrofit work on the BART structures must be carried out with out interrupting normal service schedules and within the 4h of engineering time each night. For the immersed tube, this limits possibilities to those that can be applied to the outside off barges, and from within the central access gallery. In addition, any retrofit work must not place any extra loading on the structures as they exist at present. This is particularly important for the immersed tube. All options currently being considered would also have to gain environmental clearance before progressing into final detailed design.
Along the length of the tube, two options are being considered to improve its stability and resist the negative effects of liquefaction. First, micropile tied owns could be installed through the invert of the tunnel from within the central access gallery.Secondly, vibro-replacement measures could be undertaken from barges could be undertaken from barges and install stone columns on both sides of the tube to increase the relative density of the fill to 60-70%. Other alternatives considered and rejected include exterior tube tie-downs, heavy riprap over the existing fills, and chemical or jet grouting of the backfill. All were potentially costly, would be difficult to confirm effectiveness and would cause greater environmental concerns than the micropile and vibro-replacement options.
'Io stabilize the San Francisco shoreline slope, an array of large 8ft (2.5m) diameter steel piles would be installed between the vent structure and the shoreline to about 200ft (61m) deep to reduce movement of the lower soils. A rejected alternative was installation of shear walls in the soils by means of deep soil mixing or slurry wall constructions. A further set of about eight 10-12ft diameter steel pipe piles holding a prestressed concrete perimeter frame would be installed mound the 68ft x 122ft (21m x 37m) vent structure to resist downslope forces and dynamic movements acting on the structure and prevent differential settlement due to movement of the Bay Mud beneath. Finally, several sets of 'stitching piles' comprising a combination of four to six large diameter piles with a precast concrete cap over the top of the tube, would be installed to the Bay side of the ventilation structure to resist traveling wave movements. Similar sets of stitching pile may be added on the Oakland side to the primary concept of using vibro-replacement to increase density of the backfill.
All work at the San Francisco end would have to contend with the fact that the city's Ferry Plaza.
Platform on the reclaimed Embarcadero waterfront has been built over the top of the immersed tube and around the vent structure making the location a complex working site. Concepts rejected for stabilizing the slope and soils at this site included a sheet pile barrier wall, a soil-cement mix barrier and adding extra larger-diameter piles to the Ferry Plaza Platform. Logistics on the Oakland side are less complicated although the city's shipping container port adjacent to the tube's vent structure is a larger and much busier facility than it was during the tube's construction.

Fig 3. Detail of the movement

Inserting a new secondary seismic joint about 200ft (61m) from the existing joint was considered as an alternative to accommodate future joint movements but this was ruled due to the high costs and risks to BAU operations during construction. A permanent cofferdam surrounding the vent structures and the existing seismic joints was also considered as an interim safety measure and long-term redundant protection of the tube to limit water ingress should excessive movement lead leaks at the joints. Difficulty sealing the cofferdam across the tube and the risk of causing damaging during construction caused rejection of the proposal.
The retrofit measure being considered is to install a steel linerplate sleeve across the more critical San Francisco Bay side joint. The panels will be bolted together and cover all internal surfaces, including under the tracks. The sleeve will extend from just east of the seismic joint to the exterior of the ventilation structure with a flexible sealing compound or a series of rubber '0' rings installed between the liner plate panels and the concrete surfaces to provide a positive seal against water leakage. Removable access panels would allow for future maintenance of the joint couplings
Program funding
To date, the seismic preparedness retrofit program has been financed internally by BAHT resources, as well as by some State and some FHWA funds. Further funding however is required to take the program forward. The estimated $1.6 billion program is caught between funding categories, said Horton. "The project," he said, "does not fit under the usual Federal transit formulas for funding since it is not a new transit project. It also doesn't qualify for Federal emergency funding since the earthquake event hasn't happened yet. The State of California has declared the program a project to prevent a potential emergency. A proposition on the November 2002 ballot to increase property taxes to fund this specific program gained 64% against the two-thirds majority required." The local 20% Caltrans funding (matching 80% FHWA funding) of the current work to strengthen BART aerial bridges over State roads and highways, terminates after completion of preliminary engineering.
The next opportunity to secure funding for the rest of the program was March 2, 2004. Voters decided on a measure on California's Presidential Primary ballot to increase the toll on the seven bridges operated by Cal trans in the Bay Area from $2 to $3 to fund urgent transportation projects. To pass, the measure required a simple 50% majority and it succeeded. Some $143 million of the revenue is allocated to retrofit of BART's Transbay Tube.
Funding BAHT's Earthquake Safety Program via a Cal trans road bridge toll increase is appropriate said Horton, since "the existence of BART and the Transbay services contribute substantially to relieving congestion on the already overtaxed Bay Bridge. Without BART, conditions on existing highways and bridges mound and across the Bay would be intolerable." Funding through an increase in BART fares was also said to be not feasible. BART fares increased 22% during the 1990s to fund renovations, and another 10% increase is likely to cover increasing maintenance costs. A single journey on BART across the Bay at peak is at present $6 one way.
With the ballot measure approved, "we now await word as to when our portion of the revenue can be made available," said Horton. The project will then move ahead into final design for bids for conventional design-bid-build contracts for some of the most urgent Transbay Tube retro works to be invited, awarded and progressed as the start of an envisaged 10-year implementation schedule.


Add your comment

Thank you for taking the time to share your thoughts and comments. You share in the wider tunnelling community, so please keep your comments smart and civil. Don't attack other readers personally, and keep your language professional.
In case of an error submitting Feedback, copy and send the text to
Name :

Date :

Email :

Phone No :

   Security Image Refresh
Enter the security code :
No spaces, case-sensitive