Earth Retention Conference 3

This paper presents an overview of developments and trends in the practice of supported excavations since 1990. Soil mixed walls, ground improvement and hybrid support systems are more frequently used. Design has evolved such that stiffness based approaches are common in urban areas. LRFD based design is replacing traditional allowable stress design in a number of situations. Numerical analyses of support systems have become more prevalent, especially in situations where ground movements adjacent to an excavation are important. With developments in sensors and information technology, the observational approach is becoming more automated, with near real time data available to share holders. This paper summarizes these trends and speculates how they may evolve in the future.

Significant advances in the past twenty years and anticipated advances over the next twenty years of fill wall structures are summarized herein. Advancement topics are presented in categories of: market drivers, materials, systems, performance, and analysis and design. Many of the future advancement topics can or should be interactive. Examples of interactive topics are presented. The purpose of this plenary paper is to stimulate discussion and debate on the future direction of fill walls, and to encourage work on moving these topics ahead.

This paper provides a summary of major developments in the design of embedded retaining walls in heavily overconsolidated clays. These include wall installation effects, prop forces, wall and ground movements and observational method applications. Risks and opportunities are considered, examples are given of strengthening of old retaining walls and geothermal energy utilisation. Future industry challenges: better capabilities to assess/control groundwater pressures and associated clay softening; and improved training of engineers in advanced analysis are highlighted

The concept of wall deflection path and reference envelope is introduced herein for evaluating performance of diaphragm walls. It has been found that, at a given site, wall deflection paths, which are plots of maximum wall deflections versus depths of excavation, converge to a narrow band as excavation goes beyond a depth of 10m or so. The envelope of a family of wall deflection paths, i.e., the so-called reference envelope, characterizes performance of diaphragm walls in a specific set of ground conditions. Based on the data obtained for deep excavations carried out in recent years, reference envelopes are established for the T2, TK2 and K1 Zones of the Taipei Basin and they can be used to evaluate the performance of individual walls.

Innovations in shoring systems are occurring world-wide as new technologies accelerate our ability to design and construct. The paper highlights four different areas where innovation has been observed in the Puget Sound area: shoring design, shoring system components, construction methodology, and instrumentation and monitoring. These innovations are driven by constraints on the use of ground behind excavations to provide lateral restraint, the demand for deeper basements, and the continued advance of powerful numerical modeling capability. The current trend in shoring system design is clearly towards the more efficient and cost effective performance based design. The paper also addresses future potential concepts in shoring design which are being driven by public demand for more sustainable solutions based on more effective use of energy and materials.

Many structures located close to excavations are damaged by excessive ground movements during construction. This is partly due to limitations of commonly used methods to design the excavation support system. The authors propose using a displacement-based design method that focuses on designing to keep movements within allowable values. A step-by-step approach for the method is described and illustrated for a deep excavation into soft clay.

Ground movements during excavation have the potential for major impact on nearby buildings, utilities and streets. Increasingly ground movements are controlled at the source. They are assessed by linking the ground loss at the excavation wall to the volume change and displacements in the soil mass, and then to the lateral strains and angular distortion in structural bays or units, and are related to damage using a damage criterion based on the state of strain at a point. Numerical and physical models of excavation-induced building damage were used to vary parameters and develop procedures for assessing distortion and damage. Examples of building distortion and damage are presented for brick bearing wall structures of the 1800's and early 1900's, as well as later frame structures, that illustrate how geometry, era of construction, stiffness, and condition influence building response to ground movement.

Empirical Apparent Earth Pressure (AEP) diagrams, such as those proposed by Terzaghi and Peck (1967), are regularly used for assessing support loads for propped or anchored retaining structures. This approach has been used successfully on many flexible wall projects over the years in the United States. Initially developed for assessing support loads, these diagrams are now commonly being used to assess bending moments and shear forces developed in the retaining structures by application of the AEP diagrams to a beam simply supported at the position of each of the struts or tiebacks (Sabatini et al., 1999). While this approach may achieve satisfactory results for relatively flexible wall systems, many authors have recommended more sophisticated analysis methods for stiff wall systems. This paper presents the results of comparisons of the empirical AEP diagram with a beam on rigid supports (RIGID) analysis to "beam on spring" (Winkler) and numerical analysis methods for the assessment of structural forces in deep multi-propped excavations. Comparisons are provided for a stiff slurry diaphragm wall system and a flexible soldier pile and lagging wall system. For the stiff slurry diaphragm wall with tiebacks, the RIGID approach was found to be highly non-conservative for estimating bending moments in the retaining structure. The reason for this is shown to relate to the staged excavation, tieback installation sequence adopted, and locked-in deformation of the ground below the excavation level, which are not considered in the RIGID analysis. The Winkler and numerical analyses, however, both consider these effects and result in global deflection of the wall which in turn leads to large bending moments. The analyses show that the magnitude of these deformation-induced bending moments far exceed those resulting from the distributed load between rigid supports. For the flexible soldier pile with lagging wall, the RIGID analysis still estimated smaller bending moments than the Winkler or numerical approach but by a much smaller degree of difference than the stiff diaphragm walls. In view of the limitations associated with using AEP diagrams with the RIGID analysis for stiff wall systems, it is recommended that a more sophisticated analysis method be the minimum requirement for assessing structural forces in these walls.

This paper summarizes the design and construction of steel sheet pile walls for permanent building foundations and earth retention on four projects in and around Minneapolis, Minnesota USA. The buildings have one to two levels of below grade parking and up to six above grade floor levels. In each case, the below grade parking slabs and the perimeter walls and columns of the superstructure are supported by the sheet pile foundation. The sheet pile are designed to carry uniform wall loads in the range of 5 to 12 kips per linear foot and concentrated loads of up to 251 kips. The perimeter sheet pile walls are designed for both temporary (during construction) and permanent earth retention. Utilizing the perimeter earth retention system as the permanent foundation maximizes the building footprint on a site and allows building construction in a "top-down" manner. Top-down construction reduces the need for temporary internal bracing or tiebacks to support the earth retention walls by using the building floor slabs to resist lateral earth pressures. It also permits construction of the above grade levels to start before the below grade levels are completed; this can save critical time during the early stages of a project. Design of the sheet pile for both temporary and permanent earth retention follows conventional design methodology, with special attention to each stage of construction. The interior building excavation and construction sequence is staged to reduce the cantilevered length of sheet pile until permanent floor slabs are constructed, thereby limiting pile deflection. The evaluation of vertical bearing capacity of the sheet pile foundations is based on conventional analyses for friction piles. The vertical bearing capacities estimated by static methods were verified in the field using high strain dynamic testing with a Pile Driving Analyzer (PDA).

The City Creek Block 75 Project encompasses a large city block in Salt Lake City, involving excavations to 90 feet below street grade and over 40 feet below groundwater. Earth retention, in combination with dewatering systems, was installed to support five adjacent high rise structures and three major streets abutting the site. Ground conditions are comprised of dense cobbles and gravels overlying interlayered lakebed deposits with incised stream deposits which provide conduits for groundwater recharge from the ancestral City Creek. The design and construction scheme provided the flexibility required and often emphasized by Terzaghi to adjust in response to actual underground conditions using the available means and methods. This case history traces the innovative modifications to dewatering and shoring systems implemented in response to subsurface conditions identified during Block 75 shoring and dewatering construction. Initial subsurface projections allowed for a dewatered approach using a complex multi-stage wellpoint system combined with soil nail and shotcrete shoring, supplemented by soldier or secant piles and post-tensioned elements for added stiffness at critical sections. In general, the wellpoint systems reduced pore pressures within the upper aquifer, but groundwater remained perched in an intermittent zone of interlayered silts and clays along the east wall. Modified shoring design, dewatering systems and changes to construction procedures were required to complete this wall. Consequently, the north perimeter of the project was redesigned as a composite shoring and groundwater cut-off system combining secant piling, jet grouting and soil nailing. The revised shoring and dewatering configurations from the east and north walls are presented with performance data for these unique earth retention systems.

This paper discusses a case history where hard boulders embedded in a residual clay matrix prevented completion of a top-down retaining wall. Observations of soil conditions in excavations during construction allowed an indirect estimate of soil shear strength properties, which were used to analyze the stability of an approximately 27-foot deep excavation with a vertical side. This excavation was necessary to construct a concrete cantilever retaining wall as an alternative to the top-down wall. The design of the cantilever wall within the limited right-of-way is discussed. The construction monitoring program used to verify soil properties and protect workers is also discussed.

Traditionally, excavation support systems are designed solely on the basis of satisfying limit equilibrium, using apparent earth pressure diagrams. Using this approach, the support system design becomes a function of the maximum anticipated earth pressure and is governed by overall structural stability as opposed to maximum allowable horizontal or vertical deformation. This approach produces a support system that is adequate with regards to preventing structural failure, but may result in excessive wall deformations and ground movements. This paper presents a design methodology that facilitates the sizing of all components of the excavation support system in such a way that limits the maximum lateral and vertical excavation-induced deformations. Based on the fundamental approach of the presented design methodology, structural and basal stability is guaranteed.

This paper presents the development of project-specific p-y curves for the analysis and design of tangent and secant pile walls in the Marquette Interchange Project. The passive soil resistance, characterized by non-linear p-y curves, available to the embedded part of the wall is one of the key elements in the analysis for geotechnical and structural designs of the wall. In this project, p-y curves were developed using pressuremeter tests in various soil types encountered along the wall alignments. Several design categories were developed to group p-y curves with similar soil conditions and lateral resistance. An average p-y curve was developed for each design category using a hyperbolic model. These project-specific p-y curves were calibrated against the measured pile deflection profiles from three lateral pile load tests. A design example is presented to illustrate the design procedure incorporating project-specific p-y curves. This procedure is proven to work satisfactorily in full scale.

Tiedback H-beam and wood lagging walls are commonly used for temporary excavation support systems and permanent earth retaining structures. FHWA recognized that there was an opportunity to improve the design of tiedback walls and initiated a research program to study the behavior of these walls and to develop new design recommendations. As part of the study, a 25-foot high wall was constructed. Important aspects of the walls behavior are discussed. Recommendations for; apparent earth pressures, axial load in soldier beam toes, controlling wall and ground movements, using limiting equilibrium methods for design, and checking different construction stages are presented.

The City of Seattle, like many municipalities, requires that tiebacks installed in the City right-of-way be de-stressed at the conclusion of the project. For buildings where basement walls are constructed adjacent to temporary shoring walls, de-stressing tiebacks is costly and time-consuming, and often diminishes both the waterproofing and aesthetic characteristics of the basement wall. These factors raise the questions: "Is it necessary or good practice to de-stress tiebacks?" and "What are the risks of not de-stressing tiebacks?" To help answer these questions, a demonstration was conducted to assess the risk when tiebacks are de-stressed along the embedded strand rather than at the tieback head, simulating the condition where the tieback is de-stressed during future improvement projects. The demonstration was designed to evaluate the behavior of the tieback when de-stressed with an excavator and to assess the impact to a basement wall upon immediate release of tieback load. An auxiliary goal was to evaluate the portion of the anchor capacity derived from the grout located in the no-load zone of the tieback. The demonstration involved installing two sacrificial tieback anchors on soldier piles at an active project site. After excavation exposed the no-load zone of each tieback, one tieback was severed with an excavator and the other with a hydraulic shear. Video cameras recorded the demonstration, and survey measurements were made of the soldier piles before and after the tiebacks were de-stressed. Tieback load cell readings were taken at different stages of the excavation. The demonstration did not reveal significant safety concerns or risk of potential adverse impacts to existing improvements or the adjacent basement wall when the tieback was de-stressed along the embedded strand. Additionally, the load cell measurements indicate that very little capacity of the tieback was derived from the anchor no-load zone.

Results from instrumented 187 mm (7-3/8 in.)-diameter (nominal) tremie grouted multi-strand tendon anchors are presented. Unique instrumentation consisting of strain gages mounted to PVC pipe sections were used to measure strains in the grout column in the bonded and unbonded zone. This is difficult and rarely done in small diameter anchors due to the small clearance between tendons and the borehole wall. The results provide useful data indicating progression of tension and compressive strains down the anchor during loading, contribution of grout in the unbonded zone to anchor capacity, and additional support that high radial stresses (greater than overburden pressures) contribute to high anchor adhesion values.

Canton Lake Dam is a U.S. Army Corps of Engineers (Corps) flood control project located in north-central Oklahoma. When it became necessary to increase the spillway discharge capacity to pass the probable maximum flood (PMF), the Corps decided to construct an auxiliary spillway through the right abutment. The auxiliary spillway channel is 450 feet wide, with walls totaling 1,470 feet in length retaining up to 50 feet of soil and rock in the dam abutment. The spillway training walls consist of 2-foot-wide concrete diaphragm walls with up to two rows of prestressed strand anchors. This site was geotechnically challenging, with relatively weak and erodible rock and soil. This paper presents a case study of the design and construction of the training walls for the Canton Lake Dam auxiliary spillway. The walls were designed as reinforced concrete hydraulic structures in accordance with Corps of Engineers guidelines. A combination of methods was used for design, including soil-structure interaction software (WALLAP), slope stability and seepage analysis software (SLOPE/W and SEEP/W), and conventional techniques. The application of diaphragm wall construction for permanent training walls of the spillway channel provided unusual challenges, including the design of tiebacks for extreme flood water levels behind the wall. Limited guidelines and standards for the design of permanent tiebacks also made the selection of appropriate load cases and factors of safety challenging.

Since the 1970s soldier pile and tieback anchor shoring in the Seattle downtown core has supported more than 40 deep excavations and has become substantially more economical and predictable in its performance. Design and construction efficiencies have been made possible using a growing database of performance results. Recent development sites have often been partial blocks surrounded by buildings with basements, vital utility corridors, and narrow streets. These conditions have challenged development teams and led to new solutions that have made tough site development practical. This paper analyzes published and unpublished data to demonstrate how soldier pile and tieback shoring in Seattle has changed, how a greater emphasis on observed behavior and calibrated predictive tools has led to more efficient designs, and how much more cost-effectively these designs can be installed in today's market.

An access road to a new 34.5 hectare (85 acre) multi-use development in Yonkers, New York required the use of an unconventional, permanent tied-back soldier pile wall to accommodate steep cross slopes, limited Right-of-Way (R.O.W.), and unfavorable subsurface conditions. The subsurface conditions were 1.5 to 9 meters (5 to 30 feet) of uncompacted boulder-laden fill, standing at its angle of repose and underlain by thin natural soil and irregular bedrock. In many places the rock level was above final grade, thus the wall had "mixed-face" conditions. The 183 meter (600 foot) long wall varied from 0.5 to 7.5 meters (2 to 25 feet) high on its exposed face. It was constructed of drilled-in steel H-piles for the soldier piles, timber lagging for temporary soil support and reinforced concrete infill "lagging" for the permanent condition. Rock anchors were installed at steep angles to remain within the R.O.W. This paper discusses the issues leading to the wall type selection and the various challenges encountered during construction of the wall.

Soil nail testing is a crucial aspect to design and is critical to ensure that parameters assumed during design are verified. The need to define a bond stress rather than "design load" in Contract Documents is explained. The needs to test a grouted length shorter than total nail length and to create an ungrouted free length are discussed. FWHA GEC7 provides clear instructions on procedures for design. However, these procedures have conservative and unconservative aspects. Punching shear and bar yield are shown to potentially be treated unconservatively. Recommendations for proper handing of punching shear and bar yield are presented.

A new limestone unloading structure consisting of a rotary railcar dump and conveyor was planned for construction in Marissa, IL, at the largest coal fired power plant in the U.S. An earth retention system was necessary to facilitate the underground construction. The original design included a system of braced sheetpiling. The geotechnical contractor provided a soil nail and shotcrete alternative including design and construction of 161.5 meters (530 lineal feet) of soil nail wall and over 1,672 square meters (18,000 square feet) of shotcrete earth retention around the proposed limestone unloading structure and conveyor. This paper describes the design and construction of the earth retention system and quality control measures.

A permanent soil nail wall was constructed to support an existing bridge abutment of an active major interstate. During the construction of the wall, poorly graded sand with trace silt (SP) was encountered directly beneath the abutment footing. The soil did not posses sufficient "face stability" to permit soil nail wall construction to proceed as originally designed. Further investigations determined this soil condition to be present the full depth of the proposed excavation support system thus requiring an alternate solution to be implemented. Due to restrictions associated with future roadway construction, sodium silicate based chemical grouting of the excavation face along the full height of support was selected to temporarily stabilize the poorly graded sands and permit the originally designed soil nail wall to be constructed. This paper discusses the conceptual design and construction procedures utilized to carry out the grouting program and soil nail wall construction.

The French in-situ earth retaining system soil nailing began in 1970 and benefited greatly from that government's investment in the 1986 study `Clouterre'. As such, French geology strongly influenced both practice and expectations world wide over the past 4 decades. Yet, recent studies in glacial till, a non-French soil type, have shown significant strength under-estimation using conventionally accepted design approaches. This paper reconsiders skin friction expectations for soil nail installations in till. Installation at 3 till sites (1 American and 2 Irish) are examined in detail. Traditional British, French, and American design methods and parameters are applied. Conventional methods under-predicted capacity by more than 50%, thereby raising serious questions as to the appropriateness of such design guidelines in tills. New correlations based on pile installation design are proposed.

This paper describes the results of an instrumented helical soil nail wall project. Design of helical soil nails differ from grouted tendons due to the bond strength resulting from shear strength of the soil at the helices versus bond strength between the grout/soil interface. The instrumented helical soil nail walls were constructed during 2005 by the Missouri Department of Transportation (MoDOT) at the Route 169 and 68^th Street Bridge in the City of Raytown, Missouri to allow widening of 68^th Street into 4 lanes. The walls are approximately 3.7 m (12 ft) high at the center and have a 2:1 sloped backfill from the top of wall to the front of abutment. Prior to construction of the permanent facing, verification and proof testing was performed at several nail locations. The north and south walls were instrumented with load cells and inclinometers at the east, west and center of the two wall lengths. The description of the helical soil nail wall design and construction, along with the results of the initial and follow up instrument readings are presented herein.

Following record spring precipitation in April 2008, Table Rock Dam, located in Southwest Missouri, opened all ten of its flood gates to lower the water level of Table Rock Lake and prevent flooding within the basin. The subsequent flow into Lake Taneycomo below was the highest on record and resulted in damage to the lake channel and many structures downstream. Significant scour along the banks of the lake resulted in multiple slope failures. The slope below one condominium building fell away leaving a portion of the building hanging unsupported over the lake below. To save this building and protect several other buildings in the area, GeoEngineers developed an innovative solution to support the building, regain lost waterfront property and protect the bank from future flood events. Because the river is an important fish habitat and tourist destination, the solution had to be integrated into the surrounding vegetation and geology and not unduly impact the river water quality during construction.

In 2007, construction was completed on the most current segment of Provo Canyon reconstruction, US-189 Wildwood to Deer Creek, widening an 8 km stretch of existing 2-lane highway to 4-lanes. Many of the cut slopes throughout the project required either ground nail or tieback reinforcement. A neat cement grout, consisting of a 0.5 water-cement ratio (by weight), was used for both tieback and ground nail installation. The primary means of quality assurance of the cement grout, by specification, included random sampling of the wet grout for grout cube generation, and laboratory testing the compressive strength of the cubes at specified intervals of time. An alternative method was also adopted for quality assurance of the grout, measuring the specific gravity of the wet grout with a mud balance and drawing the potential compressive strength of the grout from specific gravity and compressive strength correlations. Periodic cube sampling of the grout was continued throughout to verify the strength correlations, and both test methods were evaluated simultaneously. The findings support utilizing both methods in conjunction with each other.

The Crystal Bridges Museum of American Art is currently under construction in Bentonville, Arkansas. Most of the proposed buildings involve cuts up to 15 m (49 ft) and the construction of significant temporary and permanent retaining walls. The site is within a karst region and soil nail walls were selected for excavation support. They provided sufficient flexibility to adapt to the architectural contours of the structures and the varying rock surface. They also presented significant advantages in installation through soft zones in the soil, hard rock, and significant voids within the rock mass. This paper describes the installation of the soil nail walls and issues encountered during construction. It presents a comparison between the installation efficiency of hollow core bars and pre-drilled solid bars in this karst geology. Finally, data is presented that shows that measurement of specific gravity of the fresh grout is suitable for quality control of the grout in both hollow core bar and pre-drilled solid bar nails.

The Cement Deep Soil Mixing (CDSM) method was selected to construct shoring walls for temporary support of deep excavations for the Transbay Transit Center in San Francisco. The CDSM method is also planned to be used for ground treatment to improve the strength characteristics of soils to depths up to 150 ft (45.7 m) to protect adjacent structures from the potential impacts of excavation-induced deformations. Because of the unprecedented depths of the anticipated CDSM treatment, a prototype test was undertaken to evaluate the feasibility of the CDSM method to achieve the intended depths of treatment while maintaining the required quality in terms of strength, permeability, and continuity. This paper presents the results of the field tests, including the variation of soil-cement strengths and permeability.

In order to make a 13.1 m (43 ft) deep excavation, a 20.7 m (68 ft) Cutter Soil Mixing (CSM) groundwater cutoff shoring system was constructed in Seattle. The wall alignment was pre-trenched to remove obstructions and the wall was constructed using cement-bentonite slurry technology. The lower aquifer was depressurized to stabilize the base of the excavation while the upper aquifer was monitored closely for unexpected drawdown, which could result in unacceptable settlement of adjacent structures. Underpinning micropiles were installed below the perimeter footing of an adjacent 8-story building. The southern portion of the new building is supported on drilled shafts where the site's triangular shape made excavation less efficient. Permanent tie-downs were installed to resist hydrostatic uplift forces after the dewatering wells are shut off because the water pressure on the base of the foundation is greater than the building's weight.

This paper will present several aspects of the soil mixed CSM wall installed as part of an office building project completed in 2009 near the Seattle waterfront in an area underlain by fill and loose beach deposits with a shallow groundwater. The project involved construction of a 5-story office building with below grade parking that extends below the groundwater table. Constraints included an adjacent dry cleaner with a groundwater contamination plume, an adjacent railroad track, adjacent main arterial, loose liquefiable soils, and significant long term costs associated with discharging groundwater into the City storm drain system. A perimeter Cutter Soil Mixed Wall (CSM) was proposed by the contractor and selected. The CSM wall acted as a temporary shoring wall, a temporary seepage cutoff wall and a permanent seepage cutoff wall. The paper presents the basis for the wall design, and a description of various construction aspects including the CSM wall installation, tiebacks and dewatering. Field testing, instrumentation and laboratory testing results are described that provided critical data on wall permeability, dewatering effectiveness, wall deformation, and other aspects of the performance.

The Trench Remixing Deep (TRD) method of wall construction is an innovative method of deep soil mixing developed in Japan that employs a tool resembling a "vertical chainsaw" to simultaneously cut and mix soil and grout in place without creating an open trench. Ideally suited for projects requiring a high quality wall or for conditions where typical construction methods would be difficult, TRD has been used to construct earth retention walls with great success. This paper covers the design and construction of the TRD method to construct high quality earth retention walls.

Experience indicates significant movements at tips if inclinometers were not properly anchored in competent strata and, as a result, readings often indicated false outward wall movements in later stages of excavation. On the other hand, it has been observed that the changes in lengths of struts are minimal once these struts are preloaded; therefore walls can be assumed to be steady at corresponding strut levels in the subsequent excavation. Inclinometer readings can then be calibrated accordingly. Procedures have been proposed previously based on the experience learned from deep excavations carried out in the Taipei Basin. These procedures are substantiated herein by finite element analyses using the computer program PLAXIS.

This paper presents the design and construction of an innovative shoring system at the 8^th Avenue and Virginia Street project (8V project) in Seattle, Washington. Shoring was completed using tieback soldier pile walls. A truncated no-load zone was used for tieback design because of limited right-of-way and existing foundation and utility conflicts. Easements along the east side of the site could not be obtained from the adjacent property owner because the existing foundations could not accommodate the installation of tiebacks. Therefore, the design and excavation had to be completed in two phases using two shoring walls connected with a raker. Implementation of this complex shoring system made the project possible. Instrumentation data for the shoring showed total horizontal and vertical deflection of less than 2.5 cm (1 in).

The foundation construction of Lehigh Valley Hospital New Medical Building required the installation of an underpinning and earth-retaining system along two existing buildings. Because of the drastic changes in geological conditions, a mixed face of weathered limestone rock and residual soil, a vertical wall of underpinning and an earth-supporting structure was required at the face of existing building foundations. This earth-retaining system consisted of drilled mini-piles for underpinning; a combination of mini-piles and jet-grouted walls as the main part of the earth-retaining wall system; an A-frame structure to support the wall at the shallow rock section; and jet-grouted anchors to stabilize the wall at the deep rock section. This paper will describe details of the design and construction of various structural elements for this wall system.

Several retaining walls for a road widening project had been designed as Mechanically Stabilized Earth walls. However, the excavation necessary to construct the reinforced zone of the MSE Walls would require extended road closures. One of the MSE walls was re-designed as a backfilled anchored wall with a cast-in-place concrete facing. Excavation shoring necessary for construction of the wall was also designed. The excavation shoring consists of reinforcing a temporary cut slope with up to three rows of passive ground anchors connected to steel mesh at the face of the excavated slope. The excavation shoring was designed to allow construction of the concrete wall while limiting the duration of road closures to a few hours at a time. The maximum height of the concrete wall is approximately 6 m. Micropiles were designed to support the axial load of the concrete facing, while the ground anchors were designed to support the lateral loads of the soil. The anchor tendons extend through the concrete wall facing and the zone between the concrete wall and reinforced slope is backfilled with structural fill. A field testing program confirmed the design bond strength of the ground anchors and micropiles.

The construction of new Limestone Unloading and Reclaim Structures, and associated conveyor tunnels, required excavation and retention of over 12m (40-ft) of fly ash and fill material. The presence of a high ground water table and a relatively thin layer of soil below the excavation also provided for challenging geotechnical conditions to be overcome by the design/build team. A traditional rectangular excavation with a perimeter consisting of soldier piles and lagging or sheet piles would require several levels of significant interior wales and struts which would interfere with excavation and construction of the new structures. Ground and site conditions dictated that tieback installation would be costly. To accommodate the requirement to provide a cost effective "dry and unrestricted" deep excavation, the design/build team opted to build a pair of circular cofferdams, incorporating cast-in-place concrete compression rings to provide the required lateral support. This paper presents the innovative design and construction techniques that were successfully implemented during the project, resulting in significant savings in cost and time over conventional excavation support methods.

The construction of the Water Intake Structure for a new Cooling Tower at a power plant required excavation of a 16.76m (55-ft) deep pit beside the Susquehanna River south of Harrisburg, PA. The grading for the excavation was confined by a lined fly ash landfill on one side, an unlined fly ash waste area on another and the outlet channel from the existing operating power plant on a third. With top of rock being little more than 1.52m (5-ft) beneath the normal pool in the outlet channel, and rises in pool level in the Susquehanna River expected to be as much as 5.5m (18-ft) during construction, the site presented challenging conditions to be overcome by the design/build team. In addition to providing a description of the different ground support structures used in this project, this paper will focus on the innovative design and construction techniques implemented to construct the river wall in the dry, maintain a water-tight barrier, and provide for unfettered wall removal and channel restoration.

This paper covers several aspects of the Bellevue Technology Tower Project in downtown Bellevue, completed in 2007 after a project start in 2001. The base of the excavation was completed 26.4 m below 108^th Avenue NE, with an additional 2.1 m excavation for the mat foundation which extended the total depth of the excavation to approximately 28.5 m below street grade. It is believed that this was the deepest soil nail wall, built to date in the U.S., with a top-down construction procedure. The top-down building procedure includes construction of the permanent 31 – 46 cm thick shotcrete basement wall for the building as the facing for the temporary soil nail shoring in lifts, eliminating the requirement for temporary facing. During the excavation and shotcrete placement on the bottom row of the excavation, at an average depth of 22.7 m, the 2001 Nisqually Earthquake occurred with an Mw 6.8 event. No ground loss was experienced, and only nominal permanent lateral displacements were measured after the earthquake. Recent advances that were implemented during the design and construction included the use of strut nails to support vertical loads and limit deflections in the first lift where fill is typically encountered, use of a composite soil nail wall with vertical elements to support an 8 story building along the south property line, use of splayed nails in a re-entrant corner with a 3 story building surcharge, and use of nails to carry vertical loads of the conveyor for excavated soils. Finally, an extended life study of the system was completed after the project was put on hold prior to completing the floor slabs for the garage structure. The temporary soil nail system was only designed for the life of the excavation which in this case was approximately one year. In addition, a temporary wall is not designed for earthquake resistance, although the static factor of safety does provide for some inherent resistance. The soil nails are not designed with any corrosion protection for the extended life. After evaluating the corrosion rate and seismic recurrence of critical earthquakes an extended life of 4 to 8 years was determined for the temporary soil nail retention system.

This paper presents portions of a five-part study done to further investigate the behavior of a deep braced excavation and improve upon existing techniques for estimating the magnitude of the transitory forces in such structures. This year-long project, constructed in San Francisco Bay Mud, was completely instrumented and monitored throughout the construction process. This initial part of the research provided the measured geotechnical and structural data with which to compute and compare load predictions made at various stages of construction. The second part of the study consisted of the development of a three-dimensional finite element structural model to allow for the application and study of various earth pressure theories proposed in the literature as well as a new earth pressure hypothesis developed and presented for this project. Comparisons were calculated and tabulated in the third and fourth parts of the study which included the evaluation of the relative performance of each model in predicting structural loads. Conclusions and recommendations were presented within the fifth part. With respect to the five earth pressure models selected for comparison, those with triangular shapes performed slightly better in the prediction of structural loads than did those with trapezoidal configurations. For the new model proposed and developed, the use of the Tributary Area method along with earth pressures from K_o effective stress plus best-estimate pore pressures without reshaping of the newly suggested earth pressure diagram is recommended. Bending stresses within specific strut members were found to be surprisingly high with measured values averaging approximately 85% of P/A stresses. Augmentation of the predicted compressive stresses by 150% is recommended until further studies are completed. The finite-element structural model developed was a valuable research tool but incredibly labor intensive. Continued use of the Tributary Area Method is recommended for routine design of braced excavations.

Construction of new 914 mm, 1,219 mm, and 1,676 mm diameter sewer pipes by microtunneling on the East Boston Branch Sewer Relief Project for the Massachusetts Water Resources Authority (MWRA) required installation of 19 excavation support structures at jacking and receiving shaft locations. Three methods of excavation support were used: drilled soldier piles and lagging, drilled micropiles and contact lagging, and drilled secant piles. Jet grout bottom seals were installed at two excavation support structures. Jet grouting was also utilized to seal around existing utilities at one excavation support structure. The typical depth of excavation varied from 5.8 to 13.4 m. The excavation support method used at each jacking and receiving shaft was determined after evaluation of the existing soil conditions and logistical site constraints. Since portions of the tunnel alignment intercepted made-ground (fill), the soil conditions varied considerably along the tunnel alignment. Each excavation support structure was constructed in a congested urban environment. Presence of overhead and underground utilities and proximity to existing buildings impacted the excavation support method chosen for each shaft location. This paper presents an overview of the rationale used to determine the excavation support method utilized at each drop shaft, a review of the construction techniques used to build the excavation support structures, and the major lessons learned.

This paper describes the innovative application of jet grouting to enable fast track delivery of a large-scale basement project in soft marine clay. A self-stabilized jet grout dike was formed across the width of a basement to retain a temporary excavation 12m to 17m deep. The jet grout dike was designed to act as a rigid gravity retaining structure. Lateral stability was maintained by a horizontal jet grout slab confined within the surrounding diaphragm walls. Overturning stability was provided by the foundation piles integrated within the jet grout slab, which acted like a piled raft beneath the jet grout dike. The jet grout dike was trimmed in stages as the excavation progressed. This approach allowed simultaneous excavation for the basement to proceed, while foundation works were still in progress at the ground surface on the retained side of the jet grout dike. The main advantage for adopting a jet grout dike was the elimination of the activities involving installation and removal of temporary sheet pile walls and associated diagonal struts from the construction schedule.

Deep foundation construction in urban settings poses the unique challenge of working with deep excavations in a relatively confined area while preserving the integrity of nearby structures. A 94-ft-deep vertical excavation for a 31-story residential tower adjacent to an existing building in downtown Seattle, Washington, presented an opportunity to monitor shoring system performance in real time. Additionally, the analysis of monitoring data allowed for validation of the original design parameters. The shoring system consisted of a soldier pile and tieback walls. Underpinning piles were designed to support the adjacent building while minimizing deflections. The shoring system was designed to limit total lateral deflection adjacent to the existing structure to less than 1/2-inch to minimize the potential for damage. An instrumentation program was designed to monitor loads and deflections in the shoring system. The instrumentation consisted of strain gages, load cells, vertical inclinometers, and optical survey monitoring. Where the adjacent building prevented the use of an inclinometer, two underpinning piles were instrumented with vibrating wire strain gage pairs. Strain gage pairs were installed to: 1) monitor total axial load, 2) monitor the axial load in pile versus depth, and 3) monitor combined axial loading and bending in the pile. Load cells were installed on tiebacks on instrumented underpinning piles to monitor changes in the load during the excavation. The underpinning pile data was monitored remotely. This paper provides a discussion of the design and construction of the shoring system and an analysis of monitoring data collected from the instrumentation system.

In recent years jet grouting has been implemented in innovative ways to solve earth retention, underpinning and ground water control issue on very complex sites in difficult soil conditions. This paper will prove case histories of the innovative applications and discuss why the jet grouting technique was chosen for each case. The examples chosen span soil conditions from soft clay to coarse gravel and all contended with relatively high water tables. Three of the case histories combined jet grouting with other geotechnical techniques to enhance the load carrying capabilities of the jet grout installations.

Many will argue that fill walls, particularly mechanically stabilized earth (MSE) walls, represent a boom to the U.S. geotechnical practice and are a testament to the valuable and innovative contributions by geotechnical practioners. Many examples can be cited to demonstrate the profession's ability to push the limits in terms of wall height and creative applications. These success stories notwithstanding, there have been several reported MSE wall failures that should give the profession pause for concern. There are indications today that the profession's lack of attention and focus has started to reverse the impressive trends of innovative practice. These failures should remind us that we have to remain ever mindful of the basic tennants of good geotechnical engineering practice and that we cannot afford to lose sight of important geotechnical considerations and perspective regarding the design and construction of MSE walls and slopes. Unfortunately, the lessons that have been identified by others from past failures have apparently fallen on deaf ears....because we continue to experience failures. This paper strives to once again highlight important lessons regarding the design and construction of MSE walls from both big and small projects. Most importantly, the authors offer specific recommendations to halt this disturbing trend before it has potentially severe consequences.

The paper summarizes current recommendations for anticipated and specified maximum horizontal deformation of geosynthetic reinforced soil wall facings found in a number of codes of practice. Recommended limits on verticality are compared to a database of wall performance data collected by the writers. In most cases, end of construction (EOC) measured facing deformations for walls on firm foundations are within recommended limits. Anticipated deformations for walls extrapolated out to a design life of 75 years are also reported. The results of a careful set of full-scale wall tests show that EOC deformations are influenced by both compaction effort and global reinforcement stiffness when other factors remain unchanged. The paper is of value to design engineers by providing example deformation data to estimate the additional facing batter required to satisfy intended design alignment and to provide adequate clearance for adjacent structures.

Over the last 20–30 years the design methods used for geosynthetic reinforced soil structures have become well-proven and accepted. These methods generally are based on a "tied back wedge", or similar, approach to the design. These methods are considered to be very conservative in determining the required reinforcement strength and density for geogrid reinforced structures. This has been validated by monitoring results of instrumented structures which have demonstrated a significant discrepancy between predicted design tensile loads (strains) in geogrids and actual in-situ measured values. Recent research findings indicate that this is the result of the complex mechanisms of geogrid-soil interaction and the influence of planar layers of reinforcement vs. discrete strips which are not considered sufficiently in actual design. Most current design methods in use throughout the world give guidance on the determination of serviceability limit state (SLS) based on geosynthetic material properties. These proved to be impractical to calculate structural deformation at different design stages (e.g. a specific design life).

For the past 20 to 30 years, the design and construction of Mechanically Stabilized Earth (MSE) walls has transformed the available options for development. While this is an excellent technology, our design methods are relatively conservative, and tens of thousands of structures have been safely constructed, it appears that the lessons learned from this past experience are not sufficiently contributing to advancement of the MSE wall technology, specifically related to design. Design issues and resulted in various problems including serviceability issues (i.e., construction and post-construction occurrence of vertical and horizontal movement), construction delays, and other local performance issues including collapse continue to occur at a relatively high rate. One of the causes of these problems is that MSE walls are often not treated with the same diligence and care as other engineered structures (e.g., concrete cantilever retaining walls) in terms of the geotechnical investigation, the design review and/or the quality control/quality assurance requirements. Another principal cause is poor communication between all parties that should be involved in decisions from the concept stage and design phase through construction. Communication issues and other causes, which continue to complicate the success of MSE wall projects, are summarized and discussed in this paper. Recommendations to improve the ways MSE walls are designed and built are provided, which if implemented, will enhance the performance of MSE wall systems.

Over the past few decades, greater emphasis has been placed on environmentally responsible ("green") engineering and development of more sustainable products, structures and systems. Although Mechanically Stabilized Earth (MSE) walls are inherently more sustainable than many retaining wall systems, little has been done to evaluate or further increase the sustainability of the basic MSE wall concept. This paper describes a life cycle assessment (LCA) study that was performed to establish a baseline for the environmental impact of MSE walls from the "cradle" to dismantling and recycling the wall. This study consisted of identifying the goal and scope, performing an inventory analysis, assessing the impact and interpreting the results. The assessed impact categories are energy consumption, abiotic depletion, climate change, photo-oxidant formation and acidification. For comparison purposes, the environmental impact of a gravity wall is also evaluated using similar methods developed from the LCA study.

This paper examines the mobilization of reinforcement tension within geosynthetic-reinforced soil (GRS) structures at working stress and at large soil strains. Fully-mobilized reinforcement tension is assumed in most current design methods for the internal stability of GRS structures. In these methods the mobilized reinforcement tensile load is assumed to be equal to mobilized horizontal soil forces computed using active earth pressure theory. However, comparison with reinforcement tension loads measured in the field has shown that this approach is conservative (excessively safe) by as much as a factor of two. This observation has prompted the current study in which stress data obtained from a numerical study and two instrumented large-scale GRS retaining walls were used to examine the relationship between mobilized reinforcement tensile load and mobilized soil shear strength. The results show that the ratio of reinforcement tensile load and mobilized soil shear strength is not constant Only when the average mobilized soil shear strength exceeds 95%, is reinforcement tensile capacity mobilized significantly. Nevertheless, less than 30% of reinforcement strength is mobilized when the average mobilized soil shear strength reaches peak soil shear capacity. These results help explain why current design methods lead to computed reinforcement loads that are very high compared to measured loads under operational conditions.

The grade raising associated with the construction of a new runway at Seattle-Tacoma International Airport required construction of two near vertical tall reinforced earth walls that included the two tier 26 m North wall and the four tier 46 m tall West wall. Twenty reinforcement strips at critical wall cross-sections were instrumented with approximately 550 strain gages, and performance data were obtained during initial installation and backfill placement. The performance data facilitated the assessment of factors affecting the development of strains within the MSE walls. A simplified method to determine the amount of reinforcement strain due to installation of the reinforcement strip and subsequent compaction of fill is presented. Estimates of the initial reinforcement strain, fill thickness up to which compaction effects were noted, and the initial and geostatic reinforcement strain rate in terms of microstrain per meter of fill placed are tabulated and discussed. The local reinforcement stiffness contributes to the accumulation of initial reinforcement strain, but is found to play a significant role in controlling the rate of strain due to normal geostatic stresses. Implications for the design of very tall MSE walls are also given.

The Coherent Gravity Method has been used for more than three decades for design of Mechanically Stabilized Earth (MSE) structures. This paper discusses the history and essential characteristics of the Coherent Gravity Method and compares it to other available design methods such as the tie-back wedge, structure stiffness and simplified methods, and to monitored wall performance. These comparisons show that the Coherent Gravity Method accurately models the behavior of MSE structures reinforced with steel reinforcements, demonstrating why it is the design method recommended by the MSE industry and preferred by state departments of transportation.

This paper examines a number of second-order design factors and gives practical examples of the types of problems that may develop in the field if they are not properly accounted for in MSE wall design. Each example includes a brief review of standard practice, results from field observations and measurements and practical guidance for designers.

Mechanically Stabilized Earth Walls (MSEWs) have been routinely used over the past 20 years. Experience with MSE structures has identified several challenging issues facing public and private owners using MSEWs for transportation corridors and land development to maximize useable land area. This paper discusses the current `state-of-practice' and offers recommendations relative to procurement and design of MSE structures to minimize short-term problems and ensure a service life consistent with nationally recognized design methods (AASHTO & NCMA). Specifically, the benefits and challenges of the owner controlled aspects of the process are discussed, such as; various options to contract for MSEWs, integration/coordination with civil site design, compatibility with geotechnical site investigation, testing, analysis, and recommendations, as well as MSE design methods, procedures, and guidelines that influence both cost and performance. The division of professional design responsibility amongst the design team, a critical element to successful projects, is controlled by the owner. This paper provides guidance for future projects, primarily for owners, but also for design professionals, and contractors that balance prospective risks and rewards of the various procurement options available. Although focused specifically on MSEWs, Reinforced Soil Slopes (RSSs), and Segmental Retaining Walls (SRWs) encounter similar issues, such that the information presented is equally applicable to all types of MSE structures. Successful MSEWs will be significantly influenced by the manner in which existing and improved MSE structure design standards (AASHTO / NCMA) are implemented on future projects. This paper examines that implementation process and how lessons from the last 20 years can positively influence MSEW performance going forward.

Nevada Department of Transportation has over 150 mechanically stabilized earth (MSE) retaining walls at 39 locations. Recently, high levels of corrosion were observed due to accidental discovery at two of these locations. The resulting investigations of these walls produced direct measurements regarding the metal losses of the soil reinforcements and electrochemical properties of the MSE reinforced fill. One MSE wall (I-515/Flamingo) was replaced with a cast-in-place concrete tie-back wall at great expense because of the significant metal loss due to corrosion. There are two other walls at this intersection that were not mitigated. It is now known that aggressive reinforced fill has been used in a number of MSE walls in Nevada. In its characterization of MSE reinforced fill the Nevada Department of Transportation has used the Nevada T235B soil resistivity test method. The Nevada test method under-predicts the corrosive nature of reinforced fill soils when compared to the AASHTO T-288 test method. As the MSE wall investigations show, this under-prediction has proved detrimental to the service lives of MSE structures. The internal stability analyses (using AASHTO 2007 LRFD) of two remaining MSE walls at one intersection were performed using metal loss models developed from the statistical analysis of the direct measurements of metal loss. The results of an investigation that incorporates a statistical analysis in order to effectively undertake a prediction of the internal stability of the two remaining walls due to metal loss is presented in the paper. Other MSE walls in Nevada may also be experiencing similar high rates of corrosion which will result in a deteriorated internal stability.

Notable features of steel chain are its flexibility and ability to generate high pullout resistance than those expected from other inclusions such as round steel rods flat plain bars and flat ribbed bars. Neverthess little is known about the mechanism that generates pullout force of steel chain. In order to identify the mechanisms that generate the pullout force of chain, special types of chains were fabricated and laboratory pullout test conducted using Toyoura sand. Test results obtained from these chains showed that the frictional resistance between the chain and the soil, the passive bearing developed in front of the chain and shear resistance of the soil enclosed in the inner space of the chain are the mechanisms that generate pullout force of chain. Test results also showed that the passive bearing resistance is the most predominant mechanism accounting for over 50% of the total pullout force of chain.

The design of mechanically stabilized earth (MSE) walls is primarily based on the limit equilibrium approach. Wall deformations which are especially important for serviceability usually are not considered in the design when methods using limit equilibrium approach are utilized. Most agencies require minimum reinforcement length of 70 percent of wall height for the design of MSE walls. However, in some cases, due to existing site conditions and limited space behind a wall, it is not possible to accommodate these required reinforcement lengths. This study was performed to investigate the effect of reinforcement length on wall deformations for varying soil conditions. The effect of reinforced soil, retained/backfill soil, and foundation soil properties were considered. The modeling and analyses were performed using finite element method. The results showed that although wall deformations increase as the reinforcement length decreases, the use of soils with more favorable properties can help reduce the wall deformations and compensate for the increased deformations due to the use of shorter reinforcement lengths.

Some of the first Reinforced Earth MSE retaining walls built in the state of Maryland for public works projects were constructed at a rental car facility for BWI Airport in 1983. As part of later airport operations expansion, the site was considered for redevelopment and the rental car facilities were demolished in 2004. A comprehensive evaluation of the four existing MSE walls was performed to assess the suitability of the structures for continued use as part of new site development. In particular, several sinkholes had formed within the MSE wall envelope of one wall due to drainage alterations during demolition. In addition to evaluating the overall condition the four walls, the impact of these sinkholes was assessed. The evaluation included internal and external stability assessments. Field investigations included soil borings through the reinforced backfill, test pits to expose and inspect the integrity of the steel strip reinforcements, survey of the walls, and visual inspections. The walls were found to be suitable for continued use in the new development once remedial grouting, drainage, and maintenance measures were implemented. This case history presents a unique post-construction examination of the performance of MSE walls after more than 20 years of service and how such structures can be expected to perform over the remainder of the original 75 year design life.

A wild forest fire that swept through the Catalina Mountains near Tucson, Arizona, burnt much of the vegetation along the General Hitchcock Highway and in the watershed above. This event was followed by several precipitation events which caused large amounts of sediment to reach and plug drainage-ditches and culverts along the roadway. Several months later an exceptional storm produced over 9 inches of rain in the watershed and caused the roadway drainage features to become overwhelmed with flowing water and debris. As a result, an approximately 60-foot segment of an existing MSE wall was overtopped and undermined, and subsequently collapsed. Evaluation after the wall failure indicated that the increased flows, bulked by debris, were too large to pass through the corrugated metal pipe culvert installed through the wall, and a larger opening was needed to prevent similar occurrences in the future. Several repair recommendations were evaluated for permanently repairing the failed roadway segment. The following factors were considered: 1) the presence of loose, easily eroded superficial materials, 2) steep slopes below the wall, 3) widely variable bedrock depths, 4) environmental restrictions, 5) maintenance of traffic, 6) hydraulic requirements, and 7) construction costs. This paper summarizes the causes of the failure, the alternatives considered, and a rationale for the selection of the preferred alternative of a bridge to be constructed between the remaining MSE wall sections.

This paper summarizes investigations and studies that were undertaken to identify the cause of the collapse of several panels in an MSE wall. The collapse was caused by the presence of frost susceptible backfill near the top of the wall. This material, combined with the large freezing index at the site, probably resulted in frost penetration depths of up to 2m behind the wall panels. After the collapsed section was repaired, outward movement of panels in the same area continued, aggravated by accumulation of water from snow melt in a shallow depression behind the wall.

Three MSE walls were constructed along a two-lane rural highway where the use of non-standard on-site material as wall backfill initially appeared to be a great cost-savings measure with relatively low risk to the overall performance of the walls. The MSE walls ranged in maximum height from 8 to 11 meters and extended from 119 to 332 meters in length. Settlement of 76 to 152 millimeters was observed in the outbound roadway lane less than six months after construction. Longitudinal cracking appeared near the centerline of the roadway above the MSE walls, with voids up to 2 meters deep at spot locations. Roadway distress above the MSE walls was caused by settlement and piping of the wall backfill material. Several site-specific issues contributing to the MSE wall distress, including; (1) up to 45% fines in the wall backfill material, (2) rigid temporary shoring (soil nail walls) at the back of the more flexible MSE wall, and (3) an unlined ditch at the top of the wall, allowing surface water to flow through the clean gravel roadway base course to the back of the MSE wall. Three main lessons were learned from this case history. First, using non-standard on-site material as wall backfill ended up costing more than if imported fill was used, even though initial project costs were less and it appeared to be low risk to the wall performance. Secondly, temporary shoring systems need to be considered in conjunction with the design of MSE walls due to the differences in potential deformation between the two systems, which could lead to cracking at the interface of the two walls. Thirdly, prevention of surface water infiltration into the MSE wall system would have prevented piping of the wall backfill material and drainage improvements are relatively inexpensive costs compared to the overall cost of the entire MSE wall.

The nine-acre Olympic Sculpture Park (OSP) overlooks the Puget Sound and the Olympic Mountains, and required the import of approximately 164,000 cubic meters (215,000 cubic yards) of fill. The park is adjacent to Elliott Bay in downtown Seattle, Washington. Approximately 4,180 vertical square meters (45,000 vertical square feet) of wire basket face Mechanically Stabilized Earth (MSE) structures support the massive urban fill operation and provide the framework for the pedestrian park above. Pedestrian bridges over a busy street corridor and railroad tracks provide access to the urban waterfront. Design and construction of the MSE structures were completed by late 2006, and the park was opened for public access in early 2007. Unusual project features included: (1) True abutment MSE applications with geogrid reinforcement supporting vertical and horizontal bridge loads (2) MSE walls in pressure relief applications to eliminate lateral earth pressure on adjacent pedestrian bridge abutment walls (3) Two-stage MSE structures with multiple face inclinations (4) Back-to-back MSE walls with a narrow distance between parallel facings in a high seismic area

The first use of geosynthetic reinforced soil walls for integral bridge abutment construction in North America occurred on the Greenville Southern Connector (I-185) toll road in 1999. A second bridge with a longer span and higher loads was constructed in 2000 for the same project. Each of these four bridge abutment walls were constructed over 20 ft. (6 m) high using modular concrete block wall (MCBW) facing units and geosynthetic reinforcement with a silty fine to medium sand backfill around vertically driven steel "H-Pile" foundation elements. While the piles were designed to carry all the vertical live and dead bridge loads, the lateral loads due to momentum, braking, and thermal movement would be transferred through the integrally cast-in-place concrete abutment to the piles to the wall facing elements through the piles, located just 3 ft. (1 m) behind the MCBW facing, and resisted by the geosynthetic reinforcement within the abutment wall. This paper describes the engineering analysis, design procedures, and some of the installation details utilized for these geosynthetic reinforced Mechanically Stabilized Earth Walls (MSEWs) with MCBW facing for the traditional retaining wall loadings, plus the additional procedures to account for the pile induced lateral loads. The design procedures were based on the prevailing guidelines at the time, the 1996 AASHTO "Standard Specifications for Highway Bridges" as revised in 1998. Lateral loads were apportioned to the wall facing and geosynthetic reinforcement using "P-Y" curves for laterally loaded piles developed by Reese & Matlock. Seismic loadings were addressed by pseudo-static procedures. A review of the performance of these geosynthetic walls based on deformed shape measurement of the wall facing after ten years of service is also presented to begin to assess their performance.

In order to refine Geosynthetic Reinforced Soil Integrated Bridge System technology (developed by the FHWA), girders from a 42.7 m single span bridge were instrumented with strain gages, end pressures were measured using horizontal pressure cells, and the abutments were instrumented with vertical pressure cells and survey targets to measure girder footing and abutment wall movements. After a six month monitoring period, vertical deflections are within tolerable limits (ranging from 1.1 to 4.6 cm) and there are no visible cracks at the bridge-approach interface. The strain gages and earth pressure cells continue to collect meaningful data in terms of magnitude and trend. A change in ambient temperature causes a temperature induced strain in the steel, which affects the lateral pressure measured behind the steel girders as expected. Vertical pressures in the abutment are also affected by the thermal cycle. This paper will display the preliminary results from this project.

Underground structures have generally performed well during seismic events. Nevertheless the design of these structures has to consider seismic loading. Unlike above ground structures, the response of undergound structures to seismic loading is dominated by the seismic deformations of the surrounding soil. Seismic loading is characterized in terms of deformations imposed by the soil on the box structures and the interaction between them. This paper describes seismic analysis approaches for underground structures. The paper reviews available simplified solutions and describes procedures to conduct numerical analyses using pseudo-static and dynamic soil-structure interaction approaches. The paper describes some of the limitations associated with pseudo-static analyses for shallow structure and the need for dynamic soil-structure analyses under these conditions.

Field performance observations and experimental evidence indicate that well-built retaining structures that are composed of or surrounded by materials that do not lose strength as a result of earthquake shaking perform satisfactorily at moderate levels of ground shaking. Thus, seismic earth pressures need not be considered when the peak ground acceleration is less than or equal to 0.3 g. At higher levels of ground shaking, the seismic evaluation should include the effects of the retained earth. The use of a Mononobe-Okabe-type method requires the selection of the seismic coefficient, which largely determines the magnitude of the seismic load increment. The rational selection of the seismic coefficient requires proper consideration of the seismic hazard at the site and the amount of seismic displacement that defines the threshold between satisfactory and unsatisfactory seismic performance of the earth retaining structure. Thus, a robust seismic design procedure should include a calculation of the potential seismic displacement of the earth retaining structure. In this paper, seismic displacement design procedures for earth retaining structures are examined.

The current state of practice in the United States as implemented in the International Building Code now requires that structures with subterranean walls be designed for seismic earth pressures in addition to the normal static earth pressures. Since many design issues in the building code are introduced because of observed failure or deficiencies during earthquakes, the requirement to design for seismic earth pressures is peculiar because there is little or no evidence that any failures in engineered subterranean structures have occurred in past or even recent earthquakes. This paper examines how seismic earth pressures entered into the design practice and reviews some of the methodologies used to estimate seismic earth pressures in current engineering practice in the United States.

Which of the soil strengths should be used for seismic design of retaining walls with dry soil, peak or residual? The effort to answer this simple question leads to a generalized methodology that is based on the multiple-sliding plane concept using both the peak strength, for defining slip plane, and the residual strength, for evaluating earth pressure associated with the slip plane. Many of the assumptions made in the conventional seismic design, successfully applicable to retaining walls with dry soil, become questionable when the soil is submerged. Adequate characterization of the undrained behavior of sand under transient and cyclic loads is needed. Initial stress state of soil-retaining structure systems before the earthquake, often ignored in the simplified seismic design practice, can have a significant effect on the performance of the wall during earthquakes.

In this study, eight earthquake centrifuge tests were performed on a model of a cantilever retaining wall to study the seismic response of a retaining wall with anisotropic backfills. A special rigid container was designed and used to prepare models with different directions of soil deposition. The eight centrifuge models had 0, ±45 and 90 degrees of sand deposition angles and with dry or saturated backfills. It was shown that the fabric anisotropy had a strong effect on the settlement of backfill and the response of the retaining wall. It was also clear that the acceleration in the soil was sensitive to the fabric anisotropy of granular materials.

Retaining walls have failed either by sliding away from the backfill or due to combined action of sliding and rocking displacements, during earthquakes. Performance based design of the retaining walls in seismic areas must account for these displacements, in addition to the usual factors of safety against failure in bearing, sliding and overturning. A realistic model for estimating dynamic displacement, which accounts for the combined action of sliding and rocking and takes into consideration, non-linear stiffness of soil and geometric and material damping and coupling effects is now available, Wu and Prakash (2009). This model has been used to calculate the displacement for several combinations of backfill and foundation soil conditions. Based on this study, typical design charts for preliminary design have been proposed.

The continuing expansion of HOV lanes and limited right-of-way on urban freeways have led to increased use of back-to-back MSE walls. Because these lanes are often constructed in the existing median, many of the back-to-back structures have small aspect ratios, with the distance between the two walls often less than 0.7 times the wall height. There are no FHWA design guidelines for back-to-back walls closer than 1.2 times the wall height. In this paper, results from dynamic constitutive models are compared to those from the traditional Mononabe-Okabe pseudo-static approach. Compared to the results of conventional design methodology, data from the dynamic analyses show that walls with smaller aspect ratios exhibit greater deformation than walls having a higher aspect ratio, although this trend is obscured at lower accelerations. The study also shows that the current state-of-the-practice pseudo-static method is conservative.

Geotechnical explorations for a bridge crossing an unnamed creek near Vancouver, British Columbia, identified a sloping layer of soft clay approximately 10 m below the ground surface at one of the bridge abutment slopes. The clay was interpreted to be sensitive, and stability analyses suggested that ground improvement or other stabilization was necessary for the bridge abutments. Various design options were evaluated, and a stabilizing abutment-wall system consisting of pipe piles to support the bridge and its approaches and to stabilize the east slope was selected for the project. This paper discusses the design concept and the numerical modeling performed for the stabilizing abutment-wall system and summarizes field monitoring that was carried out during construction of the foundation system.

This paper re-analyzes the well-documented failure of a 30m deep braced excavation in underconsolidated marine clay using an advanced effective stress soil model (MIT-E3). The collapse of the Nicoll Highway during construction of cut-and cover tunnels for the new Circle Line in Singapore has been extensively investigated and documented. All prior analyses of the collapse have relied on simplified soil models with undrained strength parameters based on empirical correlations and piezocone penetration data. The current analysis use results from high quality consolidation and undrained triaxial shear tests that were only available after completion of the public inquiry. The current analyses achieve very reasonable estimates of measured wall deflections and strut loads using model parameters derived directly from the laboratory tests. The analyses confirm prior interpretations of the failure mechanism but provide a more rational basis for the modeling of soil-structure interaction.

The One-North MRT station is part of the Circle Line at Buona Vista Road in Singapore. The deep excavation involves a 30m deep cut into 20m of Jurong residual soils and 10m of Jurong sedimentary rocks to form a small stabilizing rock berm on the passive side of the wall, adjacent to the INSEAD building complex. The performance of the wall and excavation system is evaluated against field measurements and FEM analysis. Of significance is that the wall behavior showed that the Jurong residual soils behaved like a drained material despite its low permeability (k) values of less than 10^–8 m/s. This project illustrate how allowing for judicious lowering of GWT can lead to a safe and economical design of a very deep excavation in tropical residual soils conditions, with very high GWT. For correct c-phi reduction analysis, allowance must be made for possible wall and anchor failure.

This paper highlights a new earth retention strut-free scheme for deep excavation using counterfort diaphragm wall in a Singapore excavation project. This scheme is not restricted to any particular shape on plan like circular, peanut, donut, or other regular shapes which are required for the enhancement of the hoop stresses for structural resistance against soil and water pressures. These counterfort walls in thick soft soils deposit were founded on good base support for it to work well without any strut or tied-back system. It can be used for a large enclosed excavation provided there is stiff ground reaction at the base of each counterfort wall. Three dimensional (3D) finite element (FE) modelling with Plaxis 3D Foundation v2.2 is used to study and analyse the reaction and stresses of such wall system. A study on panel joint effect can only be made in a 3D model. The need for 3D versus 2D FE analysis is discussed as they may affect the estimation of bending moments for wall design.

This paper presents the numerical analysis completed for design of a 4,700-car underground parking garage located at the Microsoft Block C site in Redmond, Washington. The underground parking garage is about 335.3 m (1,100 feet) long in the east-west direction and 121.9 m (400 feet) wide in the north-south direction. The excavation and construction of the basement wall was completed using a permanent top down soil nail wall. Because of the length of the structure, the garage was designed as three separate sections with construction joints up to 152.4 mm (6 inches) wide in the east-west direction to accommodate the seismic movement of the soil as well as thermal expansion of the concrete. Permanent soil nails were utilized to resist the soil pressure and improve the seismic performance of the east and west walls. Finite element modeling was completed to estimate the seismic deformation of the east and west walls and to design gaps between floor slabs and the east and west walls. The purpose of the gaps was to prevent transfer of seismic soil pressures to the garage structure, hence, significantly reducing the number of the internal shear walls. Discussions on the effect of permanent soil nails on wall deformation and wall pressure under both the static and seismic conditions are also presented.

An elasto-plastic constitutive model for expanded polystyrene (EPS) geofoam was employed with a finite-element analysis addressing the lateral pressures on a rigid non-yielding retaining wall with geofoam inclusion. Experimental results from laboratory uniaxial and triaxial compression tests were used to derive the parameters of the elasto-plastic constitutive models for the geofoam inclusion and the backfill soil used in the finite-element investigation. The numerical study revealed that for the analyzed EPS density, specific combinations of geofoam panel thickness and retaining wall height may induce plastic strains in the geofoam. The results were compiled into design charts providing the distribution of lateral pressures on the wall and the geofoam isolation efficiency as a function of EPS panel thickness and wall height.

An existing reinforced concrete barrel of the Weber Branch Siphon and Weber Coulee Siphon along East Low Canal in the State of Washington, U.S.A. is used as an earth retaining structure during required excavation for construction of an additional cast-in-place reinforced concrete barrel in close proximity. A continuum-based numerical model is implemented to simulate significant construction stages in sequential order so that soil-structure interaction effects are considered. Results of the numerical simulation demonstrate that at the end of required excavation, lateral forces on the existing barrel from the back and front soils are close to the corresponding lateral force values obtained using conventional earth pressure calculations with at-rest and passive earth pressure coefficients, even though the development of horizontal and vertical stresses is more complex.

This paper will present the effects of the loading from an integral bridge abutment on a Mechanically Stabilized Earth (MSE) retaining wall structure. The analysis will mainly concentrate on the effects of thermal deformation phenomenon (contraction and expansion) of the bridge deck on an MSE wall structure and more particularly the induced tensile force in soil reinforcements and lateral displacement at the front face of the wall as a result of the bridge movement. A geotechnical numerical finite difference program, FLAC v5.0 2D, will be utilized to model a standard abutment (true bridge abutment on bearings) and an integral bridge abutment (bridge deck and beam seat rigidly connected). The paper will also discuss the concept, components and applications of numerical modeling results in design and construction of the MSE structures supporting standard and integral bridge abutments. The results of the two models will be compared to an empirical design methodology as developed based upon American Association of State Highway and Transportation Officials (AASHTO) design guidelines. Increasingly more complex design and construction techniques are being used for MSE walls, similar to the cases discussed in this paper, which makes the use of accurate design tools such as numerical modeling an effective verification of design assumptions.

The Donghu station of No.6 line of Guangzhou rail transit is divided two parts: the embedded excavation part and the open excavation part. The kilometer of the embedded excavation part is from k13+462.752 to k13+519.452. The tunnel crown is located 19 m below the ground surface. Soil stratums are mix-filled soil, alluvial deposit to flooded alluvial deposit, wholly-weathered rock of red strata, highly weathered rock of red strata, moderately weathered rock of red strata and slightly weathered rock of red strata from the top down. The largest span of tunnel is 20 m. Its largest height is 10.349 m. Because the tunnel has large span and section, the construction is very difficult. In order to control the surface settlement and ensure the security of tunnel, the Finite Element Analysis method is used to analyze the construction process and find the rationality of construction method and support system. A three-dimensional numerical simulation model analyzes the construction process of the tube pre-support and center drift method in Donghu station by FLAC-3D. The results indicate that the reinforced and construction method can ensure the safety of construction, and control the surface settlement. The final conclusions are two: (1) the surface settlement and displacement are small and the tunnel is steady in the construction process. (2) The main step of influencing the surface settlement and crown settlement are step I , step IV and removing of the temporary supporter. So the initial support must be installed in time and reinforced the supporter strength. The distance which is from working face to liner must be more than two times span of tunnel.

The paper discusses the implications of recent developments in codes of practice related to retaining structures, concentrating on European and American codes. In contrast to traditional working stress design, recent codes are generally based on a Limit State approach, with safety margins specified by partial factors. Both European and American developments have factors applied to loads, but while the American preference (LRFD) is to apply factors to resistances, such as bearing capacity, Eurocodes use a mixture of factors on resistances and on material strengths. Comparisons are provided for trial examples, a gravity wall and a propped embedded wall. It is shown that the precise way in which the factors are applied can have significant implications for the design result. The design to Eurocode of a new underground station for high speed trains in Florence is presented, illustrating some of the choices to be made in preparing calculations to satisfy code requirements.

Numerical analyses are performed on a routine basis in practical geotechnical engineering to assess the deformation behaviour of deep excavations under service load conditions, but it becomes increasingly common to use results from numerical analysis for ultimate limit state design (ULS). When doing so, compatibility of the design with relevant standards and codes of practice, valid in the respective country, has to be assured but there are no clear guidelines how this can be achieved. In this paper two aspects are addressed. First the influence of the constitutive model employed for modelling the mechanical behaviour of the soil on calculated structural forces of retaining walls is discussed and secondly the possibilities and limitations of introducing the partial factor concept as established in EC7 in combination with numerical analysis are highlighted.

This paper examines the advantages and limitations of employing ultimate limit state methods for the design of braced excavations. Braced excavation design requires both skill and careful evaluation of many factors that can affect performance. Traditionally in the US, braced excavations are designed with a serviceability approach where soil parameters are conservatively estimated and the performed analysis yields the service displacements, moments, and forces. Design forces are then calculated by applying a global safety factor on the service design results, while the wall embedment is determined by calculating limit equilibrium safety factors against wall rotation and passive resistance that range from 1.2 to 1.5. In Europe, in contrast to the US, an ultimate limit state design approach has been adopted in geotechnical design including the design of braced excavations. In this design philosophy both wall and supports are designed based on an ultimate limit condition. The ultimate design forces are typically determined by reducing the characteristic soil strength parameters or by multiplying the effects of actions and dividing the effects of resistances by various safety factors. At the end, a safety factor of one or greater is required for all structures and other types of safety factors. Back in the US, there is an increasing trend of promoting ultimate limit state design in geotechnical design, including braced excavations. In the author's experience the ultimate limit state method works reasonably well for most limit equilibrium methods but can produce very inconsistent results in many cases when numerical analyses are employed. Hence, the advantages and limitations of the ultimate limit state design should be carefully weighted by practitioners and academia in the US before, and if, the ultimate limit state philosophy is incorporated in a legally binding building code.

This paper describes the geotechnical aspects of earth retaining structures in the context of U.S. transportation practice based on Load and Resistance Factor Design (LRFD) methodology as adopted by the American Association of State Highway and Transportation Officials (AASHTO). The paper provides information on: (1) basic concepts of LRFD; (2) limit states, loads, load factors and resistance factors for earth retaining structures; (3) comparisons with past practice based on allowable stress design (ASD); and (4) considerations for limit state analyses for earth retaining structures in an LRFD framework. Fill and cut walls are addressed.

Previous calibrations of resistance factors for load and resistance factored design (LRFD) of mechanically stabilized earth (MSE) retaining walls with galvanized metallic reinforcements have not considered cross-sectional area as a variable in the reliability analysis. The remaining cross section at the end of a 75- or 100-year design life is considered, however the metal loss from corrosion is estimated using recommended rates of metal loss that render conservative estimates of remaining cross section. This paper describes reliability-based calibration of resistance factors for the rupture limit state considering the variability of observed corrosion rates. Results are compared with resistance factors cited in the current American Association of State Highway and Transportation Officials (AASHTO) design specifications. The comparison identifies conditions for which the current AASHTO resistance factors achieve the targeted probability of failure inherent to the LRFD strategy.

Rockeries, also known as dry stack walls, consist of earth retaining and/or protection structures typically comprised of rough onsite rocks stacked in an interlocking fashion with no mortar, concrete, or steel to retain cut or fill slopes. They are context sensitive solutions that in many cases are also relatively low cost. Several rockeries exist as historic or cultural features on many Forest Highway and National Park roads. Many were apparently built in the Civilian Conservation Corps (CCC) era of the late 1930's; some are still performing well and others have required extensive maintenance or have failed. Generally, rockeries have not been designed according to the AASHTO Standard Specifications or other accepted wall design procedures. The wide range of implementation suggests excellent performance can be expected when certain conditions are met. There is little guidance available to standard, acceptable design and construction guidelines. Therefore, a rationally based and tested design procedure is presented to provide designers and owners with the confidence that these structures can be used in modern highway engineering. Recommended guidelines for properly constructing rockeries are also presented.

Block-molded expanded polystyrene and related geofoam materials can reduce lateral pressures acting on a wide variety of earth-retaining structures to almost zero under both gravity and seismic loading. There are two distinctly different functional ways to achieve this: lightweight fill and compressible inclusion. This paper updates the state of knowledge in this area with an emphasis on seismic buffers. This particular application, which makes use of the compressible-inclusion function, has seen increasing research interest in recent years. However there appear to be significant misunderstandings about material behavior that require clarification.

Beginning in 2004, the FHWA Office of Federal Lands Highway (FLH) teamed with the National Park Service (NPS) to develop and implement a retaining wall inventory and condition assessment program supporting roadway asset management efforts underway throughout U.S. Parks. The vast majority of Park earth retaining structures were built prior to 1960, with many built circa 1935, making the assessment of this aging asset a high priority within the NPS. This paper briefly describes key development and implementation aspects of the WIP, overviews the condition assessment approach taken for the range of wall types inventoried, and notes preliminary program findings.

Many major landslides that have occurred throughout the U.S. in the last 15 years have been repaired using drilled and grouted elements. The projects include public and private facilities for many types of owners including State DOTs, railroads, casinos, shopping malls, and apartment complexes. The use of drilled and grouted elements such as tieback anchors and micropiles have enabled these owners to repair their problems for reasonable costs and within tight schedules. Details of seven different landslides are presented in this paper. They are located in a wide range of geological conditions and terrains. These areas include Southern California, the Rocky Mountains, the Appalachian Mountains, middle Tennessee and the southern Mississippi River region. Massive stabilization forces have been imparted to the ground by the drilled and grouted elements to provide adequate factors of safety to stabilize the slide masses.

Corfu Street is perched on an isolated, lenticular plateau midway up the south valley wall of the Ohio River in the West End neighborhood of the City of Pittsburgh. This plateau represents the head of a prehistoric rockslide that likely occurred following a period of glacial meltwater erosion during the Pleistocene Epoch. The plateau comprises the surface of a 40-foot thick layer of landslide debris (soil and rock) resting on a mildly sloping clay stone rock ledge. A steep, 150-foot high rock slope drops from this plateau to an active mainline railroad track below. Following a year-long period of above-average rainfall, the landslide, involving more than 120,000 cubic yards of accumulated soil and rock debris, was re-activated by an apparent significant rise in the groundwater table and resulting saturation of the slide mass. This paper describes the geotechnical evaluation and monitoring of the landslide and the design and construction of a system of ground anchors and surface reaction pads installed to stabilize the landslide.

This paper presents the innovative analysis, design and construction of permanent slope protection in highly seismic and landslide prone areas. The project scope of work consisted of creating a stable, large horizontal bench along a steep and long slope at the toe of an existing industrial complex, suitable for the construction site of a new industrial building complex 50 meters high. Several high slopes had to be developed, one after the other, and permanently stabilized to generate a safe and stable building site. Originally, the contract called for an Anchored Diaphragm Wall (ADW) to support permanently the vertical slope. However, during an early stage of construction, discrepancies on the Owner-supplied Project Site Soil and Geological Conditions were discovered that necessitated design amendment. Anchored Aligned Pile Wall (AAPW) with Pervious Piles (PP) in between to serve as vertical draining elements was finally adopted to retain the 18-meter high vertical slope, thus allowing the construction of the new building complex on the slope, free from problems of high earth and hydrostatic pressures and external surcharges. The ultimate purpose of the project was to provide local and global stability of the whole building complex during construction and ultimate service that required a design horizontal ground acceleration of 0.4g with performance objective of Immediate Occupancy after the earthquake. This paper describes the design concept of the retaining wall systems, the problems encountered due to "change of site conditions", and the solutions designed and developed to cope with the actual site conditions.

An ancient landslide was re-activated in 1998 in a hillside-graded area of San Juan Capistrano, California. The landslide occurred within a park site and threatened an adjacent residential development. The ensuing litigation led to the design and construction of a retention system with high-capacity tieback anchors in 2000. This system consisted of five levels of lateral support, with each level containing up to three rows of tiebacks and continuous walers. A total of 864 tieback anchors were installed, ranging in design load capacity from 300 to 365 kips. In order to account for the continuing slide movement that was expected to occur during the ongoing construction activities, the anchors were locked off at or above 50 percent of the design capacity. Post-construction monitoring has indicated that the measured loads in about 60 percent of the anchors continued to increase over a period of several years, up to 38 percent higher than the design capacity in some anchors. The anchor loads have now decreased to an equilibrium state at generally less than, or equal to, the design capacity (with few exceptions). This design approach has proven successful in resisting the active slide movement within the mitigated area.

This paper describes a numerical approach for the design of drilled shafts to reinforce slopes. The approach uses the program FLAC and the strength reduction method to predict the Factor of Safety of the improved slope. The method successfully predicted the maximum demands (shear force and bending moments) necessary for the structural design of stabilizing piles. The method also successfully predicted the critical mode of failure and slope stability Factors of Safety for piles with a plastic bending moment.

This paper will present the initial construction, subsequent repair, results of the instrumentation program and subsurface investigation, analyses and the stabilization measures implemented in 2004 to hold and restrain the side-hill fill. Since implementation of the two tiered, anchored sheet pile wall, monitoring has been ongoing and the results of the monitoring program since completion of construction are presented.