California dam expansion using geosynthetic-strip reinforced MSE wall

By Fransiscus S. Hardianto, Robert Lozano, John E. Sankey, and David K. Hughes

Introduction

The Los Vaqueros Dam was originally built in 1997, creating the Los Vaqueros Reservoir, to improve the quality and reliability of water available to the Contra Costa Water District (CCWD) in Northern California.

In 2011, the CCWD expanded the reservoir from 123,000 megaliter (ML) (100,000 acre-ft) to 197,000 ML (160,000 acre-ft) in capacity by raising the dam height by 10.4m (34ft). The primary objectives of the expansion of Los Vaqueros Reservoir were to develop water supplies for environmental water management supporting fish protection, habitat management, and other environmental water needs, while increasing the supply for water providers in the San Francisco Bay area.

The principal elements of the project were to excavate the crest of the existing dam by 14.3m (47ft) and then to add 24.7m (81ft) from there to the final dam height, making a total dam height increase of 10.4m (34ft)—see Figure 1. Approximately 1 million m³ of additional engineered embankment fill was placed, most obtained from borrow areas located near the dam.

The reservoir expansion project included four MSE walls with a total area of approximately 4,000m² (42,000ft²). The walls were designed for exposure to fluctuating water conditions, and some sections required trapezoidal distribution of the soil reinforcement lengths (shorter at the bottom of the wall and increasing in length as the elevation of the wall increased) to minimize cutting into the existing dam, and to avoid conflicts with the dam clay core and the existing rock backface.

To satisfy the project specifications that required the use of nonmetallic soil reinforcement, and for practical reasons due to the above conditions, the walls were designed using high-tenacity geosynthetic strip as soil reinforcement. The project site is within a relatively high seismicity zone with a specified ground acceleration coefficient of 0.50g.

Two MSE walls with a maximum height of 15m (49ft) were utilized to widen the roadway at the crest of the dam on the upstream side. The other two walls were used on the downstream side. The upstream walls were located within the phreatic zone and subjected to submerged design conditions with 1m of rapid drawdown. The wall configurations and dimensions are summarized in Table 1 and the typical cross section for the walls located upstream is illustrated in Figure 2.

MSE wall system

Geosynthetic strip was used for this project because of the variable soil reinforcement lengths and multiple field cuts required to avoid conflicts with the clay core and the existing bedrock formation. Such length variations could be adjusted based on prevailing field conditions.

The MSE wall system used for this project consisted of 1.5m × 1.5m (5ft × 5ft) and a thickness from 0.139m to 0.177m (5.5in. to 7.0in.) precast reinforced concrete panel facing system with high-tenacity geosynthetic strips as soil reinforcements. The facing panels were designed using concrete with a minimum compressive strength of 28MPa (4ksi) and reinforcing steel with a minimum yield stress of 414MPa (60ksi). The facing panels and geosynthetic soil reinforcing strips were connected using a fully mechanical polymeric recess connection system. The proprietary synthetic connection consisted of an injection-molded, polyolefin sleeve embedded in the concrete facing panels.

While polymeric (geosynthetic) soil reinforcements have been extensively used worldwide, such as in geosynthetic-reinforced block walls, the combination of the integral fully mechanical connection with the geosynthetic strip soil reinforcements and the large size concrete facing panels is a relatively new innovation in the MSE wall industry.

The geosynthetic strip is manufactured by placing high tenacity polyester yarn in bundles and then extrusion-coating them with polyethylene. The process results in 50mm-wide by 3.5mm-thick polymeric strips as shown in Figure 3(a). The ultimate tensile strength of the strip is 50kN (11.24 kips) with 11% elongation at the ultimate strength. Figure 3(b) shows a typical stress-strain curve for the geosynthetic strip in accordance with ASTM D6637.

The general long-term performance of geosynthetic reinforcements is influenced by several conditions, which are accounted for under the current AASHTO design methodology using three reduction factors. These reduction factors and the values used for this project are summarized in Table 2.

The reduction factor for installation damage accounts for the long-term loss of tensile strength that can potentially occur during construction. Testing of the geosynthetic strip using AASHTO No. 57 gravel backfill resulted in a reduction factor of 1.01. This value is less than the minimum value of 1.10 specified by AASHTO; therefore, a reduction factor value of 1.10 was selected for the design.

The reduction factor for durability is generally affected by environment conditions such as temperature and chemical exposure. For polymeric reinforcement, the major factors that cause long-term degradation are UV radiation and chemical exposure. For this application, the effect of UV degradation is usually minimal because the geosynthetic strips’ exposure to sunlight is typically limited to only a few days during installation. The durability of polyester fiber is mainly affected by hydrolysis, where the fibers will break down under the presence of water and high pH (>9). Based on a study conducted by TAI (Terre Armée Internationale) and considering the conditions in this project, the reduction factor for durability was calculated at 1.14.

For geosynthetic materials, creep is the tendency of the material to slowly deform permanently under a constant load. It occurs as a result of long-term exposure to a high level of stress that is below the yield strength of the material. The rate of this deformation is a function of the material properties, magnitude, and duration of the applied load and exposed temperature. Based on the test results performed on polyester straps, and considering the type of the application, a conservative creep reduction factor of 1.64 was selected for this project.

In addition to the reduction factors, two factors of safety were also applied. For ultimate limit state conditions in permanent walls, a factor of safety of 1.5 was used. The other factor of safety of 1.2 was applied to account for the uncertainty of determining long-term reinforcement strength and other unknown factors during construction. Applying the reduction factors shown in Table 2, and the two factors of safety, the allowable tensile strength of the geosynthetic strip for this project was calculated as follows:

Although the allowable tensile capacity of a single strap was calculated at 13.48kN (3.03kips), additional consideration for the limiting capacity of the connectors within the concrete fascia panels further reduced the allowable tension for each strap to a more conservative value of 11.88kN (2.67kips).

MSE wall design

The MSE wall design was performed in accordance with AASHTO Standard Specifications using the conventional Allowable Stress Design (ASD) method.

The wall dimensions ensured that the minimum factors of safety were satisfied. The specified minimum factors of safety for external and internal stability are tabulated in Table 3. In addition to satisfying the factors of safety, the minimum uniform equivalent soil reinforcement length was also specified to be the greater of 70% of the wall design height or 2.4m (8ft).

The soil properties specified for the design of the MSE walls are summarized in Table 4. Based on the actual gradation of the MSE backfill, the maximum apparent coefficient of friction (f*) of 1.21 was selected. A seismic ground acceleration coefficient was specified to be 0.5g. Due to the use of highly permeable backfill for MSE reinforced backfill, the AASHTO recommended default value of 1-m (3-ft) rapid drawdown was used in the stability analyses for the submerged condition.

To analyze the external stability of the wall sections with uneven soil reinforcement lengths (trapezoidal), the wall section was adjusted and represented by a rectangular block having the same total height and the same cross-sectional area as the trapezoidal section. This approach resulted in a wall with a uniform equivalent soil reinforcement length of L₀, as shown in Figure 4(a). The minimum base length (L_B) was limited to 40% of the wall height. The above simplified approach was consistent with FHWA (2001, 2009) guidelines, and in combination with the compound stability analyses performed for this project, resulted in a conservative design.

For the internal stability analyses, the maximum tension force line was the same as in conventional walls (uniform reinforcement length) where the line was bilinear as shown in Figure 4(b). AASHTO provided no guidance regarding the potential failure surface line for MSE wall using discrete high tenacity geosynthetic strip soil reinforcement. While additional studies were being conducted, the selection of the bilinear maximum tension line for this project was considered conservative. The active earth pressure coefficient at the top of the wall assumed for inextensible reinforcement (bilinear) was more than 1.5 times that for the extensible (linear) approach. The effective length for pullout capacity calculation was determined to be the actual length of soil reinforcement within the resistance zone.

The seismic design was performed using Mononobe-Okabe’s pseudo-static approach described in the AASHTO Standard Specifications, as shown in Figure 5.

While MSE walls are known to be flexible and have effectively performed well in large seismic events, the structures are usually designed with a small tolerable displacement (FHWA 2001, AASHTO 1998, 2007). To calculate the wall acceleration coefficient (k_h) in relationship with the allowable permanent displacement, a simplified Newmark sliding block method addressed in AASHTO, and further described in Franklin and Chang (1977) and Elms and Martin (1979), suggested that for freestanding gravity walls k_h = 0.5A, provided that displacements up to 250A (in mm) are acceptable; where A is defined as horizontal peak ground acceleration. Based on the above, the acceleration coefficient used for the MSE wall design was calculated to be half of the peak ground acceleration. This resulted in the calculated A_m (maximum acceleration at the wall centroid) of 0.3g.

Taking into account the above design methods and parameters, the analyses were performed. Figure 6 illustrates the typical output from the analyses. Figure 6(a) shows that for a section with 15m (49ft) in height, the uniform equivalent soil reinforcement length required by the analyses is 10.7m (35ft). Figures 6(b), (c), and (d) show the design maximum tension within the soil reinforcements for static, submerged, and seismic conditions, respectively. These are shown in the form of lateral stress imposed at each level of soil reinforcement. It is shown that the maximum tension increases approximately linearly with depth.

MSE wall construction

The MSE walls were founded on both rock and competent compacted shell-sandstone material that is a part of the dam embankment. The bearing capacity was adequate to support the MSE wall with a narrower base width at some sections of upstream walls.

The most challenging MSE construction was found at wall 3. The wall faces the upstream slope of the dam and then its alignment turns at the tallest section where the back of the wall is facing an existing rock face.

Wall 3 was designed to be almost entirely submerged. At the tallest section of the wall, the wall height is 15m (49ft) with the design height of water of 14m (46ft). To reduce the differential hydrostatic pressure within the MSE mass during reservoir drawdown, open graded (permeable) material was specified for the MSE backfill. The material consisted of a mix of coarse sand and gravel, with subangular particle shape.

The backfill gradation (Figure 7), is classified as poorly graded gravel (GP), with C_u and C_c of 2.9-3.2 and 0.9-1.0, respectively.

The initial row of panels installed for the wall consisted of alternating half and full height panels set on a leveling pad. Once the initial facing panels were set in place and braced, the appropriate length of geosynthetic strips were threaded through the recesses. As in other MSE wall systems, panels must be initially placed with a slight batter toward the backfill to compensate for a subsequent outward movement that occurs during backfill placement and compaction.

The lateral movement of the backfill during compaction tends to push the panel to a true vertical position. The amount of batter varied and depended on the type and moisture content of the backfill, required compaction, type of compaction equipment, and length of the reinforcing strips.

A batter of 38mm (1.5in.) in 1.5m (5ft) was used as a starting point. The actual movements of panels were monitored during compaction of the MSE backfill; the amount of the batter was adjusted according to field conditions.

For the compaction of the backfill, a 12-ton vibratory drum compactor was used. During the construction of wall 3, where the wall transitioned from the dam into rock, and where the back of the wall was facing the existing shotcreted rock formation, the wall panels moved during the backfill placement. To accommodate the construction sequence, the backfill at this location was configured with a steep slope as shown in Figure 8.

While the gradation of the material was likely to be the main factor of the backfill lateral movement, the exposed face also had contributed to the difficulty in achieving the compaction of the backfill due to the absence of lateral confinement of the backfill.

Figure 8 shows the limited space of about 4.6m (15ft) available between the shotcrete face and the back of the panels. During the compaction process the combination of dry aggregate, vibrations from the compaction equipment, and the presence of a backface rock formation resulted in compaction energy rebounding into the MSE backfill.

The effect appeared to cause lateral movement in the upper portion of the MSE volume due to lack of confinement. Further displacement of the facing panels occurred with operation of the vibratory drum compactor too close to the MSE panels.

MSE wall performance

Although some initial lateral movement of the panels occurred during placement of the MSE backfill, by refining the construction procedures and adjusting the initial batter during panel erection, a satisfactory wall alignment was achieved.

Figure 9 shows wall 3 constructed to the top of wall panels. The final vertical plumbness of the walls was measured to be within 38mm (1.5in.) in 3m (10ft) of vertical height. No movements were detected after the construction of the wall.

Conclusions

MSE walls with high-tenacity geosynthetic strip reinforcement can be an effective application for a dam expansion to provide a wider crest of the dam. The fully mechanical connection between the soil reinforcements and the facing panels are required to ensure the structural integrity of the walls.

The use of the conventional AASHTO Allowable Strength Design method in the wall design was practical and conservative. This design method has historically been used in the design of MSE walls that have performed quite well under submerged conditions and in past earthquakes. The wall section geometry had a trapezoidal distribution using variable lengths of soil reinforcement to avoid conflicts with the dam clay core and the existing rock cutface.

The use of high-tenacity geosynthetic strip for the soil reinforcement was practical considering the variable soil reinforcement lengths posed by site space limitations. The use of gravel backfill, while effective in minimizing the hydrostatic pressure applied to the MSE wall, has caused backfill compaction and wall alignment challenges during construction. In addition, the presence of a rock cutface behind the wall and the nonuniform lengths of soil reinforcement contributed to the challenges to maintain a vertical wall facing.

Despite some construction challenges, once fully erected, the MSE walls were stable. Additional studies are recommended to refine the construction techniques for the MSE walls using geosynthetic strip and various types of MSE backfill.

Fransiscus S. Hardianto, M.ASCE, P.E., D.GE, The Reinforced Earth Co., San Diego, Calif. 92123

Robert Lozano, M.ASCE, P.E., The Reinforced Earth Co., Vienna, Va. 22182

John E. Sankey, M.ASCE, P.E., The Reinforced Earth Co., Vienna, Va. 22182

David K. Hughes, M.ASCE, P.E., URS Corp., Oakland, Calif. 94612

Acknowledgements
The authors would like to acknowledge Jillian Gattuso for her contributions during the design of the MSE walls and Gilberto Urbina for the preparation of the figures for this article.

References

Association of American State Highway and Transportation Officials (1996, 1998, 2002). “AASHTO Standard Specifications for Highway Bridges.”

Association of American State Highway and Transportation Officials (2004, 2007, 2012). “AASHTO LRFD Bridge Design Specifications.”

Elms, D.G. and Richards, R. (1990). “Seismic Design of Retaining Walls.” ASCE Specialty Conference: Design and Performance of Earth Retaining Structures, Cornell University, Ithaca, N.Y.: 854 -871.

Federal Highway Administration (2001). Publication No. NHI-00-043, “Mechanically Stabilized Earth Walls and Reinforced Slopes Design and Construction Guidelines,” Washington, D.C.

Federal Highway Administration (2009). Publication No. FHWA-NHI-10-024, “Design and Construction of Mechanically Stabilized Earth Walls and Reinforced Slopes,” Washington, D.C.

Franklin, A.G. and Chang, F.K. (1977). “Earthquake Resistance of Earth and Rockfill Dams.” U.S. Army Engineer Waterways Experiment Station.

Terre Armée Internationale (2006). “GeoMega™ System Technical Package.”