Recent changes in the API RP 2A Recommended Practice allow for selection of more favorable design parameters for piles in granular soils when detailed information, such as in situ cone test results, 1S available. This paper presents the results of recent in situ cone penetrometer tests at several sites in the Gulf of Mexico. It discusses the application of cone penetrometer results in view of the new API guidelines. Comparisons are made of computed pile capacity using both current (1984) and previous (1982) API guidelines. A brief discussion of the history of parameter selection for pile design in Gulf of Mexico sands is also presented.
Prediction of axial load capacity of a driven Gulf of Mexico pile in granular soils has, until recently, been based on strength parameters selected relying almost entirely on material grain size. This has been done primarily for convenience. Quantification of strength parameters for sand is difficult because sands cannot be easily sampled like most offshore clays and then tested in an "undisturbed" state.
A sand sample must usually be reconstituted for laboratory strength testing. Assumptions of relative density, drainage conditions, and confining pressure are also necessary. Other prediction problems include uncertain stress changes in the soil mass during pile driving and the non homogeneity of many deposits of granular soils (Dennis and Olson, 1983).
In October 1969, The American Petroleum Institute published the first edition of API RP 2A, Planning, Designing, and Constructing Fixed Offshore Platforms. Presumptive design parameters were based on four types of granular gradations ranging from clean sand to silt. These values remained basically unchanged until the fifteenth edition of API RP 2A was published in October 1984. The 1984 guideline bases design parameters not only on gradation, but also upon density. More importantly, it allows for deviation from these guidelines based on detailed in situ information about the site.
Use of the cone penetrometer in Gulf of Mexico sands is becoming more prevalent because of recent analyses of load test results (Coyle and Castello, 1981; Dennis and Olson, 1983; and Helfrich et al, 1985) that indicate a general trend of worldwide conservatism for parameter selection in siliceous to low calcareous content sands. Cone penetrometer soundings can also provide a continuous record of the profile which can result in a more accurate characterization of site stratigraphy. Results of a few cone penetrometer soundings and subsequent analyses are shown in this paper.
The static load that is carried by a deep foundation in sand can be expressed as:
(Mathematical equation available in full paper)
It should be noted that the ultimate values of unit skin friction and end bearing are commonly assumed, for simplicity, to occur simultaneously along the pile shaft and at the tip even though API RP 2A contains an admonition that this will not occur.
