The first earth pressure balance (EPB) shield project in the United States, a 3,000 ft (915 m) long, 12.1 ft (3.7 m) diameter, tunnel is described. It is located in San Francisco and passes beneath a heavily trafficked commercial area known as Fisherman's Wharf. The soil profile consists of 20 ft (6.1 m) of rubble fill which overlies a 30 ft (9.1 m) layer of Bay Mud. A colluvium layer underlies the Bay Mud. The tunnel section is located within the Bay Mud, and has a cover of approximately 30 ft (9.1 m); water heads above the crown are typically 15 ft (4.5 m). The EPB shield was designed to deal with these difficult ground conditions without the use of compressed air or dewatering, and to cut through numerous abandoned wooden piles along the route. Surface and sub‐surface ground movements were monitored along the alignment during and after shield passage. The shield was found to initially heave the soil aside slightly as the face passed, but movements thereafter were towards the tunnel. Settlements were typically less than for conventional shields and, the ground water table was unaffected by tunneling.

A characteristic common to most tailings dam failures is that the mine tailings tend to liquefy and flow over substantial distances, with potential for extensive damage to property and life. In order to be able to assess the potential for damage in case of such failure, it is necessary to be able to predict the characteristics of the flow and the possible extent of flood movement. This paper presents analytical procedures for making such evaluations. The behavior of tailings materials during flow is represented by a Bingham plastic rheological model in these analysis procedures. It is apparent from the analyses that the flow of phosphate tailings would be expected to be turbulent but flows for other types of tailings would be expected to be laminar. The procedures described are applicable for flow of tailings on horizontal and sloping planes and in prismatic valleys. The analyses can be performed using dimensionless charts in the case of flow on planes, and a computer program in the case of flow in prismatic valleys.

A characteristic common to most mine tailings dam failures is that the tailings tend to liquefy and flow over substantial distances, with potential for extensive damage to property and life. The existence of these potentially hazardous earth structures is of great concern to the public and to the mining industries. In order to be able to assess the potential for damage in case of such a failure, it is necessary to be able to predict the characteristics of the flow and the possible extent of flood movement. A companion paper to this illustrated the various analytical procedures that have been developed for this purpose. The present paper describes a series of flume experiments that were conducted to check the validity of these analysis procedures. In light of the good agreement observed, the results of the analyses are then compared with flow movements observed in several case studies where failures have occurred.

Standards are available to estimate the vertical earth loads induced on pier or pile‐supported rigid pipes subjected to cohesionless settling backfill. There is no apparent direct field evidence to support existing design procedures. Therefore, a literature review was made to determine whether alternative methods existed. This review indicated that many other methods do exist for cases where the fill either settles relative to a fixed foundation (or pipe) or the foundation is subjected to uplift forces. It is concluded that existing standards consistently underestimate the vertical earth loads.

Standards are available to estimate the vertical earth loads induced on pier or pile‐supported rigid pipes subjected to settling backfill. There is no apparent direct field evidence to support existing design procedures. Therefore, an experimental program was undertaken to check these procedures. A small‐diameter steel pipe was placed in a prepared bed of dry sand and the sand was permitted to settle with respect to the stationary pipe. Both soil loads and relative movements were recorded and it was found that existing procedures significantly underestimate the induced soil loads, particularly at high depths of cover. Only small relative movements are required to mobilize peak loads. Alternative methods of predicting soil loads are also recommended.

Cemented soils are characterized by their ability to stand in very steep natural and man‐made slopes which are prone to sudden failures during earthquakes. Because of the hazard posed by such slope failures in many parts of the world, a program of field observations and finite element analyses was carried out to evaluate the seismic response to the steep slopes in cemented soils. The results indicate that failures of steep slopes in cemented soils will occur in moderate to severe earthquakes. The failures of vertical or near vertical slopes are apparently initiated by tensile failure in the upper parts of the slopes. The failures of moderate (less than 70°) slopes are likely to be slides parallel to the slope. Both types of failure are generally shallow, involving 2 m–5 m of material. Finite element analyses also show that the important parameters con trolling stresses and accelerations in steep slopes are the slope angle, the max imum input acceleration, and the ratio of the natural period of the deposit to the dominant period of the earthquake.

Analytical expressions for soil stiffness for a group of piles were obtained for both plane strain and three‐dimensional conditions. A comparison between these two expressions revealed that a concept of a Winkler soil model could be applicable to pile group problems for the frequency range higher than the fundamental natural frequency of the soil deposit. A simple solution of dynamic pile group response was developed using a Winkler soil model extended to pile group problems and a transfer matrix method. Natures of dynamic group effect were studied for various cases. Using a simple structure supported by a pile group, it was demonstrated how the dynamic conditions in a group effect affects dynamic response of a pile‐supported structure.

In the semi‐arid zones, seasonal variations in temperature and moisture content are extensive. This may induce severe cracking of the core of an earth dam, particularly in places where the soil is poorly compacted. The winter rains often result in flash floods leading to a rapid build up of reservoir level. Consequently, water flowing through the core has high gradients. Moderately dispersive soils, in such an environment, may become highly dispersive. Such a mechanism of failure has been postulated on the basis of a case study of an earth dam failure in Wadi Qattarah, Libya. Practical implications of this failure mechanism are pointed out in relation to earth dam construction practices in semi‐arid zones.

The probability of failure of California Delta levees was estimated using simple statistical procedures, based heavily on empirical data. Forty‐four islands were included in the study, and it was found that the probability of failure was strongly linked to the thickness of peat beneath the levees. When the islands were grouped according to peat thickness, it was found that sufficient historical levee failure data was available to estimate the average probability of failure for all the islands in a given peat thickness group. The relative probabilities of failure for the islands within a group were then estimated using slope stability analysis and simple statistical methods. For each island a Factor of Safety Index (FSI) was computed for ranges in water level outside the levee and a range of values of the strength parameter, $\varphi \prime \text{.}$ The water level was statistically characterized by the Gumbel Distribution and $\varphi \prime$ by a Normal Distribution. The probability that the FSI would fall below some critical value was used to rank the probabilities of failure for islands within a peat thickness group.