LA Metro's Red Line Station

Construction of the 7.1-km-(4.4-mile-) long Phase 1 of Los Angeles Metro’s Red Line rapid transit system involved significant groundwater problems, but only in about 20 percent of its length—that portion traversing the Los Angeles River flood plain. Here, the coarse-grained fill and river alluvium combined with a high water table and potential for recharge created the need for stringent groundwater control measures. The problem was intensified by the presence of sulfides in the water that release enough hydrogen sulfide upon depressurization to be considered a noxious atmospheric contaminant.

Regulatory agencies would not permit the pumped water from excavation to be put back into the ground in a closed loop (when you take it out of the ground, it becomes yours) or discharged into the river without its being treated to remove the sulfides. The problem groundwater affected three construction contracts and had the potential to introduce some interesting contingencies into three different bids, so the owner chose to handle a large part of the pollution personally by constructing and operating a 26-million-liter (7-million-gallon) per-day water treatment plant to remove the offending substance. Each of the three contractors was permitted to use this facility to treat the products from the pumps, though each was subject to certain restrictions.

Groundwater Control Options

Of the three flood plain construction sections, contract A135 had the most obvious immediate problems to overcome. A135 was designed as a subway station to provide a connection between the new system and AMTRAK’s existing Union Station. As such, A135 would have to cross a number of active tracks and roadways and keep them functioning throughout construction. The cut-and-cover dewatering and support design was also affected by the owner’s stipulation that the A135 contractor bear costs of chemical treatment for any dewatering flows that exceeded 1.14 million liter (300,000 gallons) per day.

The bidding options were:

• Soldier piles and lagging (a porous system with obvious groundwater treatment ramifications).
• A structural slurry wall system to provide support and exclude much of the groundwater from the excavation. More than a slurry trench cutoff, this option would have to be a continuous diaphragm wall designed for structural load bearing as well as seepage control.

The contractor chose the slurry wall option and proceeded to construct what at the time (January to December, 1989) was probably a record breaker for depth in difficult ground. At an elevation of + 70 m (230 feet), the Metro Rail Union Station excavation invert was overlain by up to 18.9 m (62 feet) of overburden, including a 6.9-m- (20-foot-) thick cobbly and bouldery layer.

With conventional dewatering, the water table would have had to be drawn down from its normal elevation of +76.8 m (252 feet) and maintained at a minimum of 1.2 m (4 feet) below bottom of excavation. Complete water cutoff (and structural support for the weights and earth pressures involved) was the aim, so the slurry walls had to be excavated deep enough to be socketed a minimum of 0.9 m (3 feet) into the underlying Puente “soft bedrock,” an essentially impervious silt/clay shale with unconfined compressive strength of 1379 kPa to 2069 kPa (200 psi to 300 psi). The undulating top of rock varied from 26 m to 30 m (86 to 98 feet) below ground surface and embedments were sometimes as great as 2.7 m (9 feet), so slurry wall depths greater than 30 m (100 feet) were not uncommon.

Design and Construction

The slurry wall was designed by the contractor from criteria provided in the bid documents, which included stipulations for a construction sequence that satisfied the railroad owner’s requirements for non-interruption of track yard activities. These stipulations caused a certain amount of skipping around during emplacement of slurry wall panels, which would not have been the case in most situations. The support and maintenance of one active railroad bridge, one roadway bridge, and two passenger bridges over the subway excavation had to be provided for.

Except for the skipping around and the extraordinarily deep penetration through difficult ground, the A135 slurry wall construction proceeded in a fairly normal manner. The contractor erected a small on-site batch plant to prepare the bentonite-water slurry used to support the panel slots during excavation. This was followed by the construction of lightly reinforced cast-in-place concrete guide walls to guide the clam shells during excavation. Two specially constructed clam shell buckets were manipulated by 136-Mg (150-ton) cranes as the panel excavations proceeded downward, all the time supported by a full head of bentonite slurry. The buckets were extra robust with cutting edges configured to penetrate soft rock, 0.7-m-(29-inch-) wide and with a capacity of 10.9-Mg (12-ton).

As excavation proceeded, steel reinforcing cages were fabricated in an adjacent work area. Each consisted of a double mat of #10 reinforcing bars spaced vertically and #8 and #5 bars spaced horizontally on 0.3-m (12-inch) centers. Sleeves were provided at pre-determined locations to accommodate the later installation of tie-back anchors. The reinforcing steel cages averaged 24 m (80 feet) in depth and were normally installed in 6-m and 18-m (20-foot and 60-foot) lengths spliced together during installation.

As is typical, the 5.5-m-(18-foot-) wide by 0.8-m-(30-inch) thick slurry wall panels were constructed in an alternating pattern for a so called primary and secondary panel sequencing. The lowering of each reinforcement cage into its slot was preceded by the installation of end-stops that had a V-notch keyway set into the slurry panel to effect the interlocking of the primary and secondary panels. The 27 580 kPa (4,000 psi) slurry wall concrete was placed by the tremie method through a pipe whose tip was kept constantly immersed at least 1.5 m (5 feet) below the surface of the accumulating deposit. The bentonite slurry, which up to that time was supporting the slot excavation, was displaced by the tremie concrete and recirculated into holding tanks for future use. The aforementioned metal end-stops were gradually pulled after the cast-in-place concrete had achieved its initial set. The final wall contained 174 individual panels enclosing the 442-m- (1,450-foot-) long Metro Rail station.

The need to maintain an unencumbered work site mandated that the slurry walls be supported, not with internal struts, but with tiebacks installed as the main trench was excavated. Tieback installations contained from 6 to 13 anchor cables tensioned to 1113 kN to 1780 kN (250 kips to 400 kips), and were spaced on 4.6-m (15-foot) vertical centers. They were inclined downwards at 20 to 30 degrees from the horizontal and installed to depths of 14 m to 18 m (45 feet to 60 feet). Grouting was accomplished with a 27 580 kPa (4,000 psi) mix.

One final peculiarity of the subway necessitated some finishing touches that would not have been required in a more routine installation. It had been determined that the entire system of tunnels would be lined with a 100-mil thick layer of high density polyethylene (HDPE) to keep the emissions from gassy ground from leaking into the completed structure. It was realized, of course, that such a lining would also contribute to the water tightness of the tunnels in the post construction phases. The need to install this lining on a relatively smooth surface required some extra finishing on the interior faces of the slurry wall panels. (Without the HDPE, the final station walls would simply have been cast directly against the irregularities of the slurry panels.) Also, the protruding heads of the tiebacks provided obvious pathways for leakage and had to be covered by pre-designed HDPE “hats” or “boots” heat welded to the main membrane.

Complete groundwater control required that the contractor install four dewatering wells inside the excavation. These wells served two purposes:

• The initial lowering of the groundwater while the slurry wall was being constructed
• The final dewatering inside the box after the wall was completed.

Readings from observation wells inside and outside the completed box indicated that a greater than 6-m (20-foot) head differential was maintained through the use of the slurry wall. As of September, 1990 (when I left the project) the A135 construction area remained completely free of water.

Conclusion

Slurry walls are not appropriate for all situations requiring groundwater control because of the expense and difficulty of installation. It is often less expensive to simply pump out the water and dispose of it in a storm drain or water way. The slurry wall comes into its own, however, when it is desirable both to control groundwater and to provide a tremendously stiff support system that minimizes ground movements. For example, it is a fairly common way of avoiding expensive underpinning of adjacent structures, underpinning that might be necessitated by the use of a much more flexible soldier pile and lagging support system.

In the case of Contract A135, the slurry wall’s robustness greatly simplified the task of supporting heavily loaded passenger trains and other vehicles as they continued to cross the subway excavation on a daily basis. Considering Metro Rail’s successful co-existence with a functioning AMTRAK station and the absolute minimization of leakage, there is every indication that the A135 contractor chose the best option for his system of support and groundwater control.