Construction Begins on San Pablo Dam's Latest Seismic Upgrade

civil_engineering_news_san_pablo


Download article PDF


Jay Landers - Civil Engineering

Sukut Construction, project’s prime contractor, upgrading the San Pablo Dam.

Situated in northern California a mere 1.8 mi (3 km) from the Hayward Fault, the San Pablo Dam has twice been upgraded to resist the strong earthquakes that are all but guaranteed to strike the area in the future. Construction work on yet another seismic upgrade to the dam began in August, and the structure’s owner, the East Bay Municipal Utility District (EBMUD), is no doubt fervently hoping that the third time will prove to be the proverbial charm.

Located in Contra Costa County roughly 3 mi (4.8 km) east of the city of Richmond, the San Pablo Dam stands 170 ft (52 m) high, 125 ft (38 m) wide, and 1,200 ft (366 m) long. When full, the reservoir formed by the dam has a surface area of 834 acres (338 ha) and boasts 14 mi (22.5 km) of shoreline. Used for recreation and flood control, the reservoir also provides raw water to two of the EBMUD’S water treatment plants.

Completed in 1921, the San Pablo Dam was built by means of a hydraulic fill construction method that used weathered sandstone and shale from nearby hillsides. With the exception of a clay-filled cutoff trench that extends from the dam’s core to the underlying bedrock, the structure is founded on alluvial sediments deposited by San Pablo Creek. To improve the dam’s seismic stability, a downstream buttress made of compacted clayey soils was added in 1967. A similar approach was used in 1979 to construct an upstream buttress, but that one connects directly to bedrock.

In the wake of advances that have led to a better understanding of dam stability during and after earthquakes, the Division of Safety of Dams, part of California’s Department of Water Resources, directed the EBMUD to reexamine the seismic stability of the San Pablo Dam. A study completed by Geomatrix Consultants, of Oakland, California, in October 2004 indicated that the hydraulic fill material in the dam’s embankment and the alluvium foundation beneath the embankment were susceptible to liquefaction. In the event of a maximum credible earthquake with a moment magnitude of 7.25 generated by the nearby Hayward Fault, the San Pablo Dam would slump as much as 35 ft (10.7 m), allowing water to spill over its crest. In view of this finding, the EBMUD lowered the reservoir to leave sufficient free board to prevent over topping should such a temblor occur.

The current, $76-million project aimed at bolstering the seismic stability of San Pablo Dam involves excavating the existing downstream buttress, employing a technique known as cement deep soil mixing (CDSM) to increase the strength of the underlying alluvium, and constructing a new, larger downstream buttress. The project was designed by TNM, a joint venture comprising Terra Engineers, Inc., of San Francisco, and Ninyo and Moore, of San Diego.  Terra Engineers served as the project management environmental studies, monitoring, and mitigation; and Wilson Ihrig and Associates, Inc., of Oakland, which handled services related to reducing noise during construction.

One early option evaluated by the EBMUD would have involved rebuilding the dam. It would also have meant removing “major portions” of the downstream slope and foundation material, as well as “excavating all the way down to rock on the downstream side,” says Xavier Irias, P.E., M.ASCE, the EBMUD’S director of engineering and construction. Properly compacted fill would then have been used to rebuild the foundation. However, that approach would have required draining the reservoir. “That had a lot of negative effects,” Irias notes.

For example, the EBMUD would have had to build a 5 mi (8 km) long bypass pipeline to maintain water supplies to one of its water treatment plants. The utility was also loath to lose the thousands of acre-feet of water stored in the reservoir. Finally, the approach would have entailed more untoward effects on the environment than the selected option, which is referred to as the in-place alternative.

“The in-place alternative was better across the board,” Irias says. The buttress that is to be replaced is roughly 50 ft (15 m) high and consists of approximately 160,000 cu yd (122,000 m3) of material, which will be reused in building its replacement. The new downstream buttress will stand about 85 ft (26 m) high and consist of roughly 270,000 cu yd (206,000 m3) of material. Additional material will be brought to the site from nearby borrow areas. “When we’re done, we’ll have a modern engineered buttress on each side of the dam,” Irias says.

The CDSM process relies on large diameter mixing augers equipped with paddles along their shafts and ports at their tips for injecting portland cement grout. As they penetrate the foundation, the augers release the grout and mix it with the soil, which later hardens into a column that has a strength much greater than that of soil alone. Descending to a maximum depth of 110 ft (33.5 m), the augers will reach into the underlying bedrock. The CDSM process will require between 50,000 and 75,000 tons (45,000 and 68,000 metric tons) of cement, along with roughly 15 million gal (56,775 m3) of water, Irias says.

Several findings from supplemental field and laboratory investigations enabled the designers to “optimize” the project’s design, says Robert C. Kirby, P.E., F.ASCE, a senior geotechnical engineer for Terra Engineers. For example, closer examination of the materials originally used to construct the dam indicated that the dam’s shells were stronger than expected and not susceptible to liquefaction.

Meanwhile, detailed studies of the cutoff trench beneath the dam’s core found that this feature was “very effective in limiting the amount of piezometric head and seepage from the reservoir,” Kirby says. Because of this, the alluvium constituting the downstream foundation was stronger than previously believed. On the basis of these findings, the designers were able to reduce the area requiring the CDSM treatment and decrease the size of the new downstream buttress.

The smaller buttress will mean that less new material will have to be transported to the site, decreasing the project’s overall effect on the environment. Another important decision by the design team involved using the long term strength of soils treated with the CDSM technique as the basis of the design, Kirby says. As with concrete, the strength of soil-cement increases over time. “We were able to quantify the magnitude of the strength increase with time,” Kirby says.

On that basis, the design team concluded that the long-term strength of the CDSM materials would be equal to twice the strength exhibited at 28 days. “We used this long-term strength as the basis of our design,” Kirby says. By halving the number of soil columns needed to achieve the same strength or stiffness, this conclusion “significantly reduced” the cost associated with the CDSM treatment, he notes.

The project’s prime contractor is Sukut Construction, Inc., of Santa Ana. The major subcontractor is Raito, Inc., of San Leandro, California, which is handling the CDSM process. Construction of the project is scheduled to be completed in April 2010.