STABILIZATION OF A PORTION OF THE HONBY LANDSLIDE, USING A MULTI-DIRECTIONAL APPROACH
Alan W. Rasplicka, P.E., G.E., Member ASCE1,
and Dean G. Francuch, C.E.G., Member AEG, ASCE1
ABSTRACT
A proposed multi-lane cross-city corridor called for grading through an area of numerous deep-seated landslides. The grading plan design called for construction of a 1.5:1(horizontal: vertical) to 2:1, 45 m (150 ft) high cut slope descending from a multi-media receiving facility to the proposed corridor road. The proposed design required an exception to the Uniform Building Code and the City of Santa Clarita Building Code slope requirements. Relocation of the media facility was not feasible due to line-of-site transmission requirements and realignment of the roadway was fixed due to property constraints. A cost effective solution using a combination of sequenced shear keys, geogrid-reinforced buttress fills and soldier piles was designed and constructed at an accelerated schedule. Prior to and during construction, several techniques were used to minimize potential damage to any nearby structures.
INTRODUCTION
In southern California, it is becoming more commonplace to develop landslide-impacted sites that in previous years were not cost effective to mitigate. Due to the current real estate market, lack of affordable housing, and infrastructure needs, it is expected that future development will continue within landslide-impacted areas. Golden Valley Road, a major 6-lane cross-city corridor, is located in the rapidly expanding city of Santa Clarita, Los Angeles County, California (Figure 1). The project consisted of a 1.9 km (1.2 mi) right-of-way constructed concurrently with an adjacent state-of-the-art high school campus. Maximum constructed heights of cut and fill slopes were 45 m (150 ft) at up to 1.5:1 gradients. The project included compaction of 7.6x106 m3 (1.0x107 yd3) of fill from April 2000 to April 2002. This paper discusses the investigation, evaluation, and mitigation of a 7.7 ha (19 ac), 1.1x106 m3 (1.5x106 yd3) landslide designated as Qls 4, a part of the Honby Landslide complex (Francuch, 2002). Landslide Qls 4 mitigation constraints included the immovable, disruption sensitive Media One/Time Warner building complex at the landslide head, the fixed location of the road, and dual 76 cm (30 in) high-pressure natural gas pipelines along the western lateral boundary and landslide toe. The financial damages for disruption of the Media One/Time Warner facility and the gas lines were estimated at thousands of dollars per minute.
__________________________
Figure 1. Simplified regional geologic map (After Saul, 1983 and Treiman, 1986).
SITE GEOLOGY
The underlying geologic units at the site include Plio-Pleistocene age (0.5 to 3.4 million years old) sandstone, siltstone, and claystone that are highly susceptible to landsliding. There were over 50 major landslides identified within the project limits. The size of the landslides ranged from several hundred cubic meters to over a million cubic meters. Bedding-plane slip from folding and close proximity to the active San Gabriel fault is partially responsible for the abundance of landslides at the site (Figure 2). Geologic data for the investigation was compiled from bucket auger borings at depths up to 33.5 m (110 ft) as well as shallow subsurface test pits and surface outcrops.
Geologic Units
The soil and bedrock materials encountered at the site consist of artificial fill, Recent age (<10,000 years) alluvium, colluvium, landslide debris, Pleistocene age (0.01 to 1.6 million years) terrace deposits, Pleistocene age Pacoima Formation, and Plio-Pleistocene age (0.01 to 5.3 million years) bedrock of the upper Saugus Formation and Sunshine Ranch Member of the Saugus Formation.
Landslide Debris
The landslide material ranged from loose sand and gravel to dense sandstone and siltstone. Depending on the amount of movement or material, the entire landslide or the upper portions of the landslide debris were disturbed and were unsuitable for fill or structural support. With progressive depth in Qls 4, material became more competent, presumably due to decreased movement and lack of surficial weathering processes. It is common practice to completely remove smaller landslides prior to placement of overlying certified engineered fill. However, where complete landslide removal was not required to obtain adequate slope stability safety factors, samples of landslide material were collected and tested for strength and consolidation characteristics.
Figure 2. Simplified site geologic map with superimposed as-built grade.
Bedrock
Saugus Formation deposits typically consisted of light yellowish brown to pinkish gray, fine- to coarse-grained, pebble- to cobble-bearing sandstones. This unit was moderately well-bedded with high dry strength. The Sunshine Ranch Member of the Saugus Formation deposits were comprised of blue-gray, silty sandstone and claystone which grade upward into dark olive brown siltstone and sandstone, with some concretionary limestone. These units ranged from low dry strength near the surface to high dry strength at depth. Landslide Qls 4 was predominantly composed of the Sunshine Ranch Member materials.
Geologic Structure
Geology structure is dominated on the site by the northwest-trending San Gabriel fault, which forms a major structural boundary within southern California. Deformation along the fault has resulted in the formation of a zone of parallel to sub-parallel faults and fractures approximately 245 m (800 ft) wide (Frankian, R. T. and Associates, 1999). North of the fault, within the area of Qls 4, bedding forms an anticline referred to locally as the Highway Anticline (Oakeshott, 1958), which has resulted in bedding dipping to the southwest, west and northwest under Qls 4 (Figure 3a). As anticipated from the initial investigation (R.T. Frankian and Associates, 1999), a series of high-angle northeast dipping faults were found within the buttress fill backcut for the landslide removals within the head area of the landslide (Figure 3b). The faults formed the failure surface within the head area of Qls 4.
Groundwater
Groundwater was encountered in approximately 95 percent of the exploratory borings. The minimum depth to perched groundwater encountered in the borings within Qls 4 was approximately 11.6 m (38 ft); average depth was 18.6 m (61 ft). Significant perched water was expected above the base of slide planes and was uncovered within the central portion of the landslide during excavation operations (Figure 3a).
Landslide Qls 4 Failure Mechanism
The underlying geologic conditions at the site include Plio-Pleistocene age sandstone, siltstone, and claystone that are highly susceptible to landslides. Bedding-plane slip and close proximity to the San Gabriel fault are partially responsible for the abundance of landslides at the site. Subsurface data indicate that landslide Qls 4 was a combination-type failure, which incorporated a fault-controlled cross-bedding mechanism at the head and block-glide bedding-plane parallel mechanisms at its base. Landslide Qls 4 ranged in thickness from 30.5 m (100 ft) at the mid-section to 18.2 m (60 ft) near the toe. Within Qls 4, geologic structure, geotechnical sampling, laboratory testing, and slope stability analyses were evaluated by, or obtained from nine down-hole logged bucket auger borings at depths up to 33.5 m (110 ft).
STABILITY ANALYSES
Slope stability analyses were performed using computer software programs that included Tslope and Tslab (TAGA Engineering Software Services, 1984) and the program Slope/W (GEO‑SLOPE International Ltd., 1991-1995). The computer
programs utilized either Bishop’s Simplified Method or Spencer's Method. A minimum factor of safety of 1.5 was required for permanent static conditions and a factor of safety of 1.1 was required for pseudo-static conditions, which included a seismic coefficient of 0.15. For temporary stability, a factor of safety of 1.25 was required.Even though the temporary factor of safety requirements were met for the planned excavation, soldier pile shoring was recommended between the top of the proposed backcut and the Media One/Time Warner facility (Figure 3b). The decision to recommend the shoring was based upon the magnitude of the landslide mitigation, perched groundwater conditions that could have seasonal rainfall variation, the ability to maintain maximum flexibility during an accelerated construction schedule, and the consequence of distress and/or disruption to the facility. A grading sequence was recommended that provided for temporary stability requirements.
After shoring was installed, an alternative grading sequence was undertaken for construction of the buttress keyway trench located at the toe of the super pad. The evaluation of this alternative grading sequence required analyses of multiple cross sections and a three-dimensional slope stability evaluation. The alternative grading sequence consisted of excavation of the buttress keyway in four separate slots. There were requirements for backfill of each buttress keyway slot to specific elevations that provided support of the up-gradient failure surface prior to progressing to the next buttress keyway slot. Construction of the buttress keyway was performed in accordance with the alternative grading sequence with favorable results while maintaining temporary slope stability requirements.
Shear Strength Parameters
Shear strength parameters were determined based on laboratory direct shear tests performed on samples obtained from the subsurface investigation. Soil parameters used in the slope stability analyses for landslide failure planes were considered to be residual strengths, based upon performing multiple shear tests on the same samples until there is no further decrease in strength. The following shear strength parameters correlated well with on-site slopes that remained stable, as well as observed failures that occurred as a result of temporary excavations during grading operations.
TABLE A. Shear Strength Parameters
MATERIAL
COHESION
kg/m2 (psf)
SHEARING ANGLE RESISTANCE - degrees
Landslide Failure Plane and Clay Bedding Plane (Residual)
976 (200)
10
Landslide Failure Plane and Clay Bedding Plane (Peak)
1831 (375)
19
Saugus Fm. Cross-Bedding
2930 (600)
32
LOSS PREVENTION
Grading activities required soil and rock movement by heavy equipment within 4.6 m (15 ft) of the Media One/Time Warner building complex. In order to prevent damage to the existing television cable and internet distribution facilities, a program of loss prevention techniques was established. The program consisted of a pre-construction building survey, slope inclinometer installation, and vibration monitoring.
A detailed building survey was completed using photography to note any existing interior building distress. The exterior walls and concrete foundation apron were logged using paper and pencil technique to create graphic logs of all exterior defects (mortar fractures, concrete cracks, etc.). Logs were completed at a scale of 1cm=0.6m (1 in=5 ft).
A slope inclinometer was installed between the Media One/Time Warner complex and top of the proposed backcut to a depth of 27.4m (90 ft). The purpose of the slope inclinometer was to measure any slope movement of the backcut during removal operations of Qls 4 below the Media One/Time Warner building complex. The slope inclinometer was monitored at least weekly and more frequently during periods of heightened grading activity.
During grading operations around the Media One/Time Warner complex, a seismograph was placed adjacent to the building to monitor for the occurrence of vibrations capable of causing adverse impacts to operations or distress to the building complex.
CONSTRUCTION
Soldier Pile Installation
A row of 20 soldier piles, 15.2 m (50 ft) deep 2.4 m (8 ft) on center were installed approximately 4.6 m (15 ft) north of the building complex, immediately adjacent to the proposed backcut for the landslide removal. The piles were installed within a 76 cm (30 in) diameter flight auger excavation drilled to 15.8 m (52 ft) below ground surface. Each pile consisted of a 53 cm (21 in) steel "I"-beam lowered into the finished excavation, which was then backfilled with concrete by tremie pipe. The tops of the piles were tied together with a 61 cm x 61 cm (24 in x 24 in) concrete grade beam surrounding a 46 cm (18 in) steel I-beam. The grade beam was located 0.6 m (2 ft) below grade to avoid future interference with maintenance vehicles.
Grading Operations
Due to constraints within other areas of the project, it was determined that the original grading sequence for Qls 4 would require modifications. The original grading sequence consisted of excavation of the entire buttress keyway after making sufficient up-gradient removal. The alternative grading sequence called for excavation of the buttress keyway in four slots, while making less up-gradient removals. Construction of the buttress keyway was performed in accordance with the alternative grading sequence with favorable results. After construction of the buttress keyway, the alternative grading sequence was abandoned. Significant removals were made within the central portion of Qls 4 without first making the recommended up-gradient removals. Failure occurred as forewarned, and this required additional slope stability evaluation of a combination 0.75:1 to 2.2:1 52 m (170 feet) high temporary backcut excavation. It was also necessary to re-evaluate the final slope configuration which would require a variable 1:5:1 to 2:1 fill slope with geogrid reinforcement and grading within areas which were originally proposed to remain natural.
Due to the failure of the landslide mass, it was necessary to remove more of the landslide mass than originally recommended. The revised temporary backcut required abandonment of the existing slope inclinometer and installation of a new inclinometer. This mitigation and revisions to the slope configuration would not have been possible and distress to the Media One/Time Warner facility may have occurred if the soldier pile shoring had not been installed. It was also possible to make modifications to the Qls 4 mitigation that were undertaken with minimal delay. The as-built finish slope configuration and the location of the installed geogrid are shown in Figure 2.
CONCLUSIONS
Mitigation of Qls 4 was completed without injury, loss of life, or distress to the existing improvements. The constructed buttress fill has performed well with no unusual maintenance or repairs required.
REFERENCES
Francuch, D. G. (2002) “Stabilization of a portion of the Honby Landslide complex, Santa Clarita, California.” Geological Society of America Abstracts, Cordilleran Meeting, Corvallis, Oregon, USA.
Frankian, R. T. and Associates (1999) “Geotechnical and Geologic Grading Plan Review, Proposed Golden Valley Road and High School, Santa Clarita, California.” Prepared for the City of Santa Clarita, California, Vol. I-IV.
GEO-SLOPE International Ltd. (1991-1995) “SLOPE/W.” Calgary, Alberta, Canada.
Oakeshott, G. B. (1958) “Geology and Mineral Deposits of San Fernando Quadrangle, Los Angeles County, California.” California Division of Mines Bulletin 172, 147p.
Saul, R. B. and Wooten, T. M. (1983) “Geology of the south half of the Mint Canyon Quadrangle, Los Angeles County, California.” California Division of Mines Open File Report, 83-24, 139p.
TAGA Engineering Software Services (1984) “TSLOPE and TSTAB.” 2855 Telegraph Avenue, Suite 415, Berkeley, California, USA, 94705.
Treiman, J.T. (1987) “Landslide hazards in the east half of the Newhall Quadrangle, Los Angeles County, California.” California Division of Mines Open File Report, 87-16.
