Between January and March, 1994, a drilling program was conducted at the Treasure Island Naval Station to evaluate the liquefaction resistance of soils that did and did not liquefy during the Loma Prieta, California, earthquake of October 17, 1989. A second goal of this drilling program was to test and calibrate a retrievable piezometer system that is designed to monitor dynamic pore-water pressure during liquefaction.

Retrievable Piezometer

A retrievable piezometer can be used to replace failed transducers without redrilling, and the external casing can be installed without the piezometer itself. Many external casings can be installed throughout a region and used only when necessary.

At two sites the USGS retrievable piezometer was placed at depths between 2.3 and 4.6m. The retrievable piezometer involves augering a hole to the testing depth and emplacing a 33-mm outside diameter pvc pipe with a porous stone. The hole is back filled and sealed with bentonite, the top of the boring is capped with a box flush to the ground. Later, a commercial transducer is connected to a 21-mm outside diameter pvc pipe and lowered down the 33 mm casing and screwed into the bottom porous-stone assembly. A calibrating transducer (the same type and model as in the USGS retrievable piezometer) was installed inside a penetrometer with a 60° conical tip and an external sleeve that protects the porous filter, located immediately behind the tip, during advancement through dry soil. After the instrument was advanced to the proper depth the tip with the porous filter was advanced past the protective sleeve. Pore pressure was elevated separately by dynamic impact and blasting.

The first calibration tests were conducted within the U.S. Geotechnical Test Site established at the Treasure Island fire station (building 157) (de Alba and others, 1994). A 590 kg weight (diameter 72 cm) was dropped 0.69 to 1.63 m onto a steel plate (91 cm square, 0.6 cm thick) to elevate pore pressure, each test involved dropping the weight one time. The surficial distance from the energy source to the piezometers ranged from 1 to 3.6 m. At an empty field (bounded by 11th and 13th streets and H and I ave) pore pressure was elevated using the 590 kg weight and no. 8 blasting caps (50 grains, 3 grams) and primer cord. The explosives were placed 2 m from the piezometers at depths of 2.4 to 2.7 m. A USGS explosives expert handled the explosives under the supervision of Navy personnel.

Liquefaction

Gray and brown, fine to medium grained sand was hydraulically dredged from San Francisco Bay to create Treasure Island (fig. 1). During the Loma Prieta earthquake Treasure Island experience a peak ground acceleration of 0.16 g (Shakal and other, 1989) and portions of the soil beneath Treasure Island liquefied and were vented to the ground surface as sand boils. During November, 1989, a survey of Treasure Island was made to document ground effects such as sand boils, settlement, and ground cracking. During this survey samples of more than 30 sand boils were taken for grain size analysis (Bennett, in press).

Although the soil beneath the fire station did not liquefy, surrounding areas did. The surrounding liquefaction may have affected the peak ground acceleration. Fifteen seconds into the acceleration record at the fire station there is a sudden drop in ground acceleration, and 16 seconds into the record there is practically no response (Idriss, 1991). De Alba and others (1994) ascribe the behavior of the acceleration record to the liquefaction of the underlying sand.

Besides the generation of sand boils, Treasure Island experienced significant ground settlement and lateral deformation that damaged lifelines for water and gas (Seed and others, 1990).

The primary objective of this report is to document the subsurface stratigraphy at the liquefaction and non-liquefaction sites (fig. 2), and to explore the relation between sand boils and subsurface sediment. This documentation adds to the geotechnical data base of liquefaction by clarifying which layers actually liquefy during earthquakes. Another objective is to briefly describe the piezometer calibration test in terms of what was done and where it was done, results of the calibration work will be reported later by the primary investigator, Behnam Hushmand of Hushmand Associates.