10NCEE

Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska

MITIGATION OF SOIL LIQUEFACTION BY DEEP SOIL MIXING FOR HOSPITALS M. Lew1, L. Shao2, M.B. Hudson3 and M.A. Murphy4 ABSTRACT Hospitals are considered essential critical facilities in California and many other jurisdictions. California law mandates that hospitals remain functional after a major earthquake. One of the seismic hazards that confront some hospital facilities is soil liquefaction, which can result in loss of shear strength, settlement, and sometimes lateral spreading of the ground. If there is a potential for significant soil liquefaction, the hospital will have to accommodate the liquefaction structurally, such as being supported on deep piles, or the liquefaction potential of the soil would need to be mitigated by soil improvement methods. Soil improvement by deep soil mixing is one of these technologies and involves the mixing of the soils in place with cement to form a soilcement mixture that would have sufficient strength to avoid the triggering of liquefaction and have sufficient stiffness to limit the amount of settlement of the soil. The soil improvement may allow for use of conventional foundations. Deep soil mixing may be appropriate for soil improvement applications when other potentially less costly methods cannot adequately improve the soil and/or limit the liquefaction-induced settlements or lateral displacements; this is especially true if there is a higher fines content that limits the effectiveness of other methods (such as stone columns, for example). Two California hospital case histories are presented and the design of the deep soil mixing configurations, the resulting structural systems used in the superstructure, and the quality control/quality assurance procedures will be presented.

1

Principal Engineer, AMEC Environment & Infrastructure, Inc., Los Angeles, CA 90040 Chief Engineer, Hayward Baker Inc., Santa Paula, CA 93060 3 Principal Engineer, AMEC Environment & Infrastructure, Inc., Los Angeles, CA 90040 4 Associate Engineer, AMEC Environment & Infrastructure, Inc., Los Angeles, CA 90040 2

Lew M., Shao L., Hudson M.B., Murphy M.A. Mitigation of soil liquefaction by deep soil mixing for hospitals. Proceedings of the 10th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.

Mitigation of Soil Liquefaction by Deep Soil Mixing for Hospitals M. Lew2, L. Shao2, M.B. Hudson3 and M.A. Murphy4 ABSTRACT Hospitals are considered essential critical facilities in California and many other jurisdictions. California law mandates that hospitals remain functional after a major earthquake. One of the seismic hazards that confront some hospital facilities is soil liquefaction, which can result in loss of shear strength, settlement, and sometimes lateral spreading of the ground. If there is a potential for significant soil liquefaction, the hospital will have to accommodate the liquefaction structurally, such as being supported on deep piles, or the liquefaction potential of the soil would need to be mitigated by soil improvement methods. Soil improvement by deep soil mixing is one of these technologies and involves the mixing of the soils in place with cement to form a soil-cement mixture that would have sufficient strength to avoid the triggering of liquefaction and have sufficient stiffness to limit the amount of settlement of the soil. The soil improvement may allow for use of conventional foundations. Deep soil mixing may be appropriate for soil improvement applications when other potentially less costly methods cannot adequately improve the soil and/or limit the liquefactioninduced settlements or lateral displacements; this is especially true if there is a higher fines content that limits the effectiveness of other methods (such as stone columns, for example). One California hospital case history is presented and the design of the deep soil mixing configurations, the resulting foundation system used to support the superstructure, and the quality control/quality assurance procedures will be presented.

Introduction Hospitals are subject to a higher standard of structural performance than ordinary buildings in California. In 1973, in response to the 1971 Sylmar earthquake (Mw 6.6) when four major hospitals were severely damaged and evacuated, the Alfred E. Alquist Hospital Seismic Safety Act [1] was enacted (California Seismic Safety Commission 2001). The Health and Safety Code Section 129680 states that “…hospitals that house patients who have less than the capacity of normally healthy persons to protect themselves, and that must be reasonably capable of 1

Principal Engineer, AMEC Environment & Infrastructure, Inc., Los Angeles, CA 90040 Chief Engineer, Hayward Baker Inc., Santa Paula, CA 93060 3 Principal Engineer, AMEC Environment & Infrastructure, Inc., Los Angeles, CA 90040 4 Associate Engineer, AMEC Environment & Infrastructure, Inc., Los Angeles, CA 90040 2

Lew M., Shao L., Hudson M.B., Murphy M.A. Mitigation of soil liquefaction by deep soil mixing for hospitals. Proceedings of the 10th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.

providing services to the public after a disaster, shall be designed and constructed to resist, insofar as practical, the forces generated by earthquakes, gravity and winds.” Liquefaction is one of the hazards that affect some hospitals in California. Liquefaction is a phenomenon in which the strength and stiffness of soil is reduced by earthquake shaking. When liquefaction occurs, the strength of soil decreases and the ability of a soil deposit to support foundations is reduced; in addition, liquefied soils will settle and possibly flow laterally. Hospitals sited in potentially liquefiable soils must be designed to mitigate the liquefaction hazard. Mitigation strategies can be structural, where the hospital building is designed to resist and accommodate the effects of liquefaction; an example may be supporting the building on deep pile foundations extending below the liquefiable soils. Other mitigation strategies would be to improve the ground so that the effects of liquefaction are reduced or eliminated to provide an acceptable level of performance; in the case of hospitals, this would allow for continuation of service after an earthquake. Common ground improvement technologies used to mitigate liquefaction hazard have included vibroflotation, vibro-replacement (stone columns), deep dynamic compaction, compaction grouting, earthquake drains, deep soil mixing (also known as wet soil mixing) and jet grouting. This paper addresses the deep soil mixing method and its recent application to new hospital projects in California. What is Deep Soil Mixing? Deep soil mixing (DSM), also known as the wet soil mixing method, is a ground improvement technique that improves the characteristics of weak soils by mechanically mixing them with cementitious binder slurry [2]. To construct columns, a powerful drill advances drill steel with radial mixing paddles located near the bottom of the drill string. The binder slurry is pumped through the drill steel to the tool as it advances and additional soil mixing is achieved as the tool is withdrawn (see Figure 1). To perform mass wet soil mixing, or mass stabilization, a horizontal axis rotary mixing tool is located at the end of a track hoe arm.

Figure 1. Deep Soil Mixing (Courtesy Hayward Baker)

The binder slurry is injected through a feed pipe attached to the arm. The process constructs individual soilcrete columns which can be in rows of overlapping columns providing essentially 100% mass stabilization or in grid patterns, all with a designed strength and stiffness. The technique has been used to increase foundation capacity, decrease static or seismic settlement, increase global stability, and mitigate liquefaction potential for planned structures, tanks, embankments and levees. DSM has also been used to construct in situ gravity retaining structures, and to facilitate tunnel construction or remediate the impact tunneling may have on nearby structures. Soil stabilization by wet soil mixing can provide structural support and/or it can greatly reduce lateral loads on bulkhead walls. DSM is best suited for soils with moisture contents up to 60 percent. Soft cohesive soils are usually targeted as other soil types can often be treated more economically with other techniques. If the moisture content is greater than 60 percent, dry soil mixing may be more economical. Soils vary widely in their ability to be mixed, depending on the soil type, strength, water content, plasticity, stratigraphy, and texture. Almost any soil type, including organics, can be treated with DSM although some soils may require significant binder and/or pretreatment, and soil containing cobbles or boulders can be more difficult to treat with DSM. With DSM, treatment is possible to depths up to 100 feet. Depending on the soil type, excess soilcrete generated may range from 10 to 40 percent of the treated volume. Stiff soils and obstructions are sometimes predrilled ahead of the soil mixing process. Although other ground improvement techniques, such as vibro-replacement or stone columns, might be more economical, DSM can provide much more assurance in the control of total and differential settlements, both under static and seismic loadings, especially in inter-bedded soft clays and loose sands. For hospitals, settlement is more of a concern than for common buildings as the expected performance goes beyond collapse prevention to maintaining service to the public in accordance with the law. Mixing shaft speed, penetration rate, batching, and pumping operations are typically adjusted after constructing one or more test columns in a convenient area on site. Pre-production laboratory testing is used to prescribe mix methodology, energy, and the grout slurry system. Specialty contractors have developed proprietary special equipment and software for the realtime monitoring of all mixing parameters during wet soil mixing. Wet sampling in freshly mixed columns and coring of cured columns can be used to verify strength and consistency of mixing. Test columns can be excavated for visual inspection of the soilcrete. Visual inspection is also possible with a down-hole camera lowered into a core hole. Also, the quality control measures for DSM are more straightforward and less variable than for other ground improvement techniques. Instead of performing standard penetration tests in borings or advancing cone penetration test soundings, usually quality control consists of verification of strengths for the soilcrete. Case History of Community Memorial Hospital, Ventura, California Community Memorial Hospital is constructing a new replacement hospital building. The main part of the new building will be six stories in height with one basement level as shown in Figure

2; there are also structurally separate one- and three-story wings over the common basement. The building will have steel-braced frames above the first floor and concrete shear walls on the ground floor and in the basement. A more detailed description of this project has been presented by Hudson et al. [3].

Figure 2. Proposed Community Memorial Hospital (under construction). Image provided by HBE Corporation. The site of the hospital is in the Ventura-Oxnard plain of Ventura County, California. The site is underlain by about 23 to 27 meters of Holocene age alluvial fan deposits that consist predominantly of loose to medium dense silty sand, interlayered with sandy silt, poorly graded sand, clay, and with some gravel and few cobbles. The Holocene deposits are underlain by late Pleistocene age alluvial deposits consisting of silt, clay, sand and some gravel. The basin is quite deep as Quaternary deposits extend to about 3,200 meters. Groundwater was encountered between 8.8 to 12.8 meters below the ground surface; however, local seepage was encountered as shallow as 5.2 meters. There are active faults that are within a few kilometers of the site. A liquefaction analysis following the procedures of Youd et al. [4] for the design earthquake indicated that there was potential for liquefaction of the Holocene age deposits (extending to about 27 meters below the ground surface at the site. The evaluation predicted that liquefactioninduced settlements of 16.5 to 27.3 cm could occur with differential settlements from liquefaction of up to 2.5 cm over a horizontal distance of about 9 meters. The analyses also indicated that below a depth of about 18 meters, the total liquefaction-induced settlement would be less than 3.8 cm and the differential settlement would be less than 1.3 cm over a distance of 9 meters. The potential for lateral spreading due to liquefaction was considered to be low. From a foundation perspective, the loose to medium dense nature of the Holocene age deposits would have dictated that deep foundations such as drilled or driven piles would be needed; the existing major hospital buildings at the site are supported on drilled piles. However, the use of piles for the new hospital given the liquefaction potential would require extremely long piles that would be up to 150 feet in length to account for the structural loads, loss of capacity in the liquefied soils, and added down-drag loads. Installing drilled piles to such depths was considered to be difficult and driving piles was not considered an acceptable option adjacent to the existing hospital, therefore reducing the feasibility of the use of piles. Thus ground improvement options were explored. In examining ground improvement options for the hospital, control of settlement

became the most important consideration. As the hospital needed to remain functional after the design earthquake ground motions, the amount of total settlement and differential settlement became the primary criteria for selection of the appropriate ground improvement method. Vibro-replacement was considered, however, it was not selected because it would not be adequate to improve the silty soils to limit the total and differential settlements to the tolerable limits to provide continued hospital operation. Dynamic compaction would not be able to improve the soils sufficiently because of the depth of the liquefiable soils; in addition, the noise and vibrations would be difficult to deal with in a hospital environment. Ground improvement by jet grouting was feasible; however, the cost would be very high. Ultimately it was determined that DSM performed to a depth of about 27 meters (below previous ground surface) would provide the needed ground improvement and achieve the desired performance with respect to allowable total and differential settlements. By improvement of the upper soils to the 27-meter depth, the settlement of the upper soils would be mitigated leaving only the settlement of the deeper untreated soils. It was determined that the settlements of the deeper soils would be acceptable in terms of total and differential settlement as the structure would be able to remain functional after the design earthquake event. In addition, tie-down anchors for seismic uplift resistance could be installed within the DSM columns; these tie-down anchors would otherwise have needed to penetrate through the liquefiable soils and into the deeper competent non-liquefiable soils and thus be much deeper. DSM would also have quality control procedures that would be easier to implement, and would result in less variability of resulting improvement than for other methods. Because DSM had not been used for a hospital in California before, the State of California Office of Statewide Health Planning and Development (OSHPD) required that an “Alternate Method of Compliance” (AMC) be submitted by the design team because DSM is not a method specifically addressed in the building code. The initial design prepared by the ground improvement design/builder consisted of a standard deep soil mixing configuration, consisting of deep soil mix columns in cellular patterns, with some intermediate independent columns for floor slab support. This cellular-type approach relies on the improvement provided to directly support footings, and also attract shear stresses to reduce shear strains in the unimproved soil within the cells and therefore prevent or reduce liquefaction (Nguyen et al. [5] and O'Rourke, and Goh [6]). The initial design and AMC was submitted to OSHPD for review. OSHPD agreed that the design could be reviewed as submitted, but the review process would take a significantly greater time since it would likely require extensive finite element modeling. OSHPD has expressed its interpretation that the deep soil mixing was essentially unreinforced concrete and could not be used in a cellular configuration for isolated foundations without extensive modeling to prove the adequacy of the configuration. Alternatively, OSHPD would allow a continuous mat foundation for the building and retain the cellular deep soil mixing approach; this combination could be thought of as allowing the foundation to “bridge” between soil columns. Another acceptable alternative would be to utilize spread footing foundations but the deep soil mixing would be required to be essentially continuous over the entire footprint of the building rather than in a cellular configuration. Of the three alternatives, the project team chose to perform deep soil mixing over the entire footprint in order to not change the design which was based on the use of spread footings. This approach was

selected to avoid a lengthy redesign and reduce the expected review time, and reduce the design and construction cost of the foundation elements of the structure. The layout of the essentially continuous soil mix columns is shown in Figure 3.

Figure 3. Continuous soil mix column layout under the footprint of the entire new building at Community Memorial Hospital. With this alternative, the deep soil mixing could be thought of as being equivalent to a conventional type of soil replacement; similar to excavation and recompaction, and the only additional AMC approval would be verification methodology that the complete replacement was effective. A nominal unconfined compressive strength of 0.25 MPa (36 psi) was specified as the minimum strength of the soilcrete mixture, which corresponds to the strength of a soft rock. The structure could then be supported on spread footings on top of the deep soil mix columns or supported footings established in engineered fill placed on the top of the soil mix columns. A net allowable dead-plus-live load bearing pressure of 168 KPa (3,500 pounds per square foot) was used for design. A test section was performed prior to DSM production. Continuous inspection was provided during the installation of the test section. Wet grab soil mix samples were retrieved and placed into molds from each test column. Samples were retrieved using an in situ wet sampler immediately after column construction and consisted of a set of at least 6 specimens (3 pairs) in 7.6 cm by 15.2 cm (3-inch by 6-inch) molded cylindrical specimens from each wet sample. The Geotechnical Engineer of Record cured all wet samples and performed unconfined compression testing at 7, 28, and 56 days. Continuous core samples were retrieved from borings and were 6.4 cm (2.5 inches) in diameter or greater. A full-length continuous core boring was performed in the soilcrete and selected length segments were tested for unconfined compressive strength. After the successful completion and approval of the test section by OSHPD, the installation of the 2,240 production soil mix columns began. The design/build contractor used a state-of-the-art computer-based data acquisition and control system to measure, control, and record all soil mixing parameters for every column installation once per second. The on-board computer

measured the mixing tool penetration rate, rotation RPM and torque, grout flow rate and density, and automatically calculated the binder content in place, blade rotation number, and drilling index. The grout pump was automatically adjusted by an active control loop in order to maintain the binder content prescribed by the design specifications. Multiple DSM rigs were used during installation as shown in Figure 4.

Figure 4. Multiple DSM rigs used at Community Memorial Hospital. Figure 5 shows the DSM mixing tool with an attached sample box to obtain wet grab samples of the soilcrete mixture in the DSM column. The design/build contractor also installed 268 tie-down anchors, within 3 mm of the design locations. The tie-down anchors were bonded into the soilcrete with allowable capacities ranging from 668 kN and 1114 kN. All of the anchors were proof tested to 133% of the design load. Five percent of anchors were performance tested up to 200% of the design load according to the Post Tensioning Institute anchor testing procedure [7]. Four anchors, i.e., two pre-production anchors and two production anchors, were creep tested to evaluate the anchor creeping deformation under long term loads. The DSM columns were installed successfully and the superstructure is being constructed on the DSM-treated soils at the time of this writing. Future Hospital Project A second hospital project in Los Angeles is also planned to use DSM to provide stable foundation support. The proposed hospital building will be four to six stories in height and will not have a basement, however, the site is located in a small valley and the site of the building is partly on the alluvial floor of the valley which has a gentle slope of about 7:1 (horizontal to vertical) and partly into a 1½:1 to 2:1 slope on the edge of the valley. The site has a difference in

ground surface elevation of about 9.1 meters (30 feet) from the high to the low side. The site is mantled by Holocene age alluvium consisting of silty clay and silty sand. Miocene age sedimentary bedrock consisting of interbedded siltstone and sandstone underlies the alluvium and the surface of bedrock slopes such that construction of the proposed hospital would cut into the bedrock on the higher portion of the site and the top of bedrock would dip below the alluvium on the lower portion of the site becoming deeper away from the slope. Thus the proposed hospital would extend partly into bedrock and partly in the alluvium; the deepest alluvium was encountered at a depth of about 21.3 meters (70 feet) below the ground surface. Groundwater seepage was encountered as shallow as 1.5 meters (5 feet) in the exploration borings.

Figure 5. DSM mixing tool with attached sample box to obtain wet DSM grab samples. The site is located within an area as having a potential for liquefaction by the California Division of Mines and Geology [8]. Analysis of the potential for liquefaction for the design ground motion using procedures outlined in Youd et al. [4] indicated that there was a potential for liquefaction within the alluvium, particularly in the silty sand deposits. The analysis also indicated that liquefaction-induced settlement would range from 2.5 to 10.2 centimeters (1 to 4 inches) in the portion of the building site underlain by alluvium and generally increasing with the thickness of the alluvium, as shown in Figure 6. Various alternative foundation treatments were considered. Driven piling was considered as well as several ground improvement techniques. Driven piles were not considered a satisfactory foundation system because of the noise and vibration installation of the piles would cause adjacent to the existing hospital. Other types of piling were also not considered to be desirable. Ground improvement techniques such as vibro-replacement (stone columns), jet grouting and deep soil mixing (DSM) were considered. Because of the variable depth of alluvium and the interlayering of the silty sand with silty clay, vibro-replacement was not considered desirable because acceptable differential settlement across the building footprint could not be achieved. Jet

grouting and DSM were judged to provide similar results with acceptable settlements, however, jet grouting would be more costly and DSM was selected.

Figure 6. Typical pre-improvement liquefaction analysis vs. the soil mixing design target. Fresh soil samples were obtained from the site and pre-production lab mixing tests were performed. As shown in Figure 7, the soilcrete strength developed as a function of curing time with difference binder dosage rates. Using the lab strength development curves as a benchmark, the engineers can predict the 28-day strength from the early soilcrete strength during the production stage, a good benefit for the timely construction QA/QC. Specifications for DSM were developed for the project and an application for an Alternate Means of Compliance was submitted to OSHPD. The project has not yet begun construction at the time of writing.

Unconfined Compressive Strength (kPa)

1600 1400 1200 1000 800 600 400 200 0

B150 B200 0

7

14 21 28 Specimen Age (days)

35

Figure 7. Soilcrete strength development in the pre-production lab mixing test. Conclusions To maintain the operation of the hospital after the design earthquake, structural performance necessitates settlement below specified criteria. Full replacement deep soil mixing treatment provides a high confidence of construction consistency and therefore confidence of meeting the design performance during the design earthquake, as well as flexibility for the configuration of spread footing foundation and tie-down anchor design. It also can reduce the overall building construction time compared to other ground improvement or foundation alternatives that are adequate to mitigate potential liquefaction. References 1.

California Seismic Safety Commission. Findings and recommendations on Hospital Seismic Safety. 2001.

2.

Federal Highway Administration. An Introduction to the Deep Soil Mixing Methods as Used in Geotechnical Applications. Publication No. FHWA-RD-99-138, March 2000.

3.

Hudson, M.B., Shao, L., Murphy, M.A., and Lew, M. “Sustainable Foundation Support of Community Memorial Hospital Against Liquefaction Hazards.” Proceedings, Geo-Congress 2014, Atlanta, 204.

4.

Youd, T.L., Idriss, I.M., Andrus, R.D., Arango, I., Castro, G., Christian, J.T., Dobry, R., Finn, W.D.L., Harder, L.F., Hynes, M.E., Ishihara, K., Koester, J.P., Liao, S.C., Marcuson, W.F., Martin, G.R., Mitchell, J.K., Moriwaki, Y., Power, M.S., Robertson, P.K., Seed, R.B., and Stokoe, K.H. “Liquefaction Resistance of Soils: Summary Report from the 1996 NCEER and 1998 NCEER/NSF Workshops on Evaluation of Liquefaction Resistance Soils.” J. Geotechnical & Geoenv. Engrg., Vol. 127 (10), 817-833, October 2001.

5.

Nguyen, T. V., Rayamajhi, D., Boulanger, R. W., Ashford, S. A., Lu, J., Elgamal, A., and Shao, L. “Design of DSM Grids for Liquefaction Remediation”, J. Geotechnical and Geoenv Engrg., Vol. 139 (11), November 2013.

6.

O'Rourke, T. D., and Goh, S. H.. "Reduction of liquefaction hazards by deep soil mixing." NCEER/INCEDE Workshop, March 10-11, 1997, MCEER, University at Buffalo, Buffalo, NY.

7.

Post Tensioning Institute. Recommendations for Prestressed Rock and Soil Anchors. 4th Edition, 2004. ISBN 1931085-29-3.

8. California Division of Mines and Geology. “Seismic Hazard Zones – Los Angeles Quadrangle Official Map,” March 25, 1999.