Considerations for Selecting Approaches to Estimate Lateral Spread Displacements for Assessing Pipeline Performance

Liquefaction hazards, specifically lateral spread displacement resulting from liquefaction, are the most pervasive and often the most severe earthquake hazards to buried oil and natural gas transmission pipelines. While the potential threat of lateral spread displacements to buried pipelines has long been recognized, the availability of approaches and tools to assess the likelihood of liquefaction and estimate of the severity of the lateral spread hazard has improved significantly in the past 10 to 15 years. The most significant areas of improvements have focused on the development of probabilistic approaches, the use of alternatives to standard penetration tests such as cone penetrometer tests and shear wave velocity measurements, and methods to combine the results of dynamic site response analyses with what remains an empirical approach to liquefaction assessment. Approaches available for estimating lateral spread displacements can be generally divided into two categories, empirical methods and analytical methods. Empirical methods are generally limited to estimating the magnitude of lateral spread displacements at the ground surface and must rely upon judgment in defining ground displacements at depth, the size of the lateral spread, and the variation of lateral spread displacement both within and at the margins of the lateral spread zone. Current advanced analytical methods range from relatively simple mathematical analogs to finite element continuum models using either Eulerian or Lagrangian formulations. In addition, many researchers have developed specialty codes for assessing lateral spread displacement. As analytical methods have increased in complexity, there has been a commensurate increase in the cost to undertake these methods because of additional subsurface information requirements and the engineering effort to implement the analyses. Despite the significant advances in the ability to estimate lateral spread displacements, engineering assessment of pipeline performance for potential lateral spread hazards remains an approximate proposition requiring considerable judgment owing to practical limitations with respect to available information, modeling capabilities, and interpretation of modeling results. In many instances, these limitations raise questions regarding whether or not the added value provided by more sophisticated lateral spread modeling techniques justifies the substantial increase in costs. These questions are explored by comparing the advantages and constraints related to common approaches for estimating lateral spread hazard with respect to the ultimate goal of assessing the adequacy of a pipeline river crossing. The comparisons support a conclusion that the most advanced methods may not necessarily lead to a more reliable design solution.