The study summarized in this paper has two main goals:
verify the utility of static pushover analysis in accurately assessing a steel template-type offshore platform's ultimate limit state (ULS) behavior, and
provide results by which key assumptions for the program ULSLEA' can be verified. ULS evaluations were performed for two Gulf of Mexico platforms: South Pass 62A and Ship Shoal 274A. The results of these evaluations were compared to the historical performance of the platforms during hurricane Camille (SP 62A) and hurricane Hilda (SS 274A) in order to determine the accuracy of the solution approach. Biases and effects which skewed the analysis results away from "realistic" or historical performance were identified and studied. Also, the effects of horizontal framing in the jacket on first member failure and load redistribution were also examined.
Two major sources of bias which were examined were those affecting pile foundation modeling (sampling, testing and load-rate biases), and bracing members (calibration of buckling performance). The effects of material strength values (true yield strengths, strain-rate effects), proper load apportionment, and dynamic load reduction on overall platform performance were also briefly discussed.
Initial analyses for both structures were based on conventional design-basis foundation pile-soil interaction characteristics. In both cases, the use of these characteristics resulted in unrealistic performance of the platforms: the platforms were predicted to fail when they did not and to fail in a manner inconsistent with post-hurricane inspections. Realistic characterizations of the influences of soil sampling, testing, and load rates on pile foundation performance produced results results which were in agreement with the observed behavior of the platforms.
Biases associated with material strengths and bracing member performance were found to influence the estimated maximum lateral load capacities of the structures on the order of 20-30 %. Consideration of the dynamic responses of the platforms resulted in effective static loads 10–20 % less than those established without accounting for the time-varying nature of the loads.
The difficulties in predicting local failures such as localbuckling and cracking were also discussed. Local failures found in SS 274A following hurricane Hilda did not manifest themselves at the level of detail initially considered within the analyses; more-detailed evaluations of the stresses within members and at joints were needed in order to determine the failure modes. Consideration of biases in material properties and identification of stress concentrations must be made for assessment of local failures.
The effects of horizontal framing in the jacket of SP 62A on first member failure and post first member failure was studied. It was found that the presence of this framing had little effect on first member failure; however, horizontal framing was found to greatly influence load redistribution following the failure of the first member.
Results obtained for the two structures indicate the validity of using static pushover analysis to assess the lateral load capacity of steel template-type platforms subjected to hurricane wind and wave loads. With proper accounting for biases, dynamics and local stress effects, the analytical estimates were found to compare favorably with the historical performance of both structures.