Stochastic Incorporation of Uncertainty and Subjectivity in Deepwater Pipeline Route Optimization

摘要

Least-cost path pipeline route optimization based upon composite geocost surfaces has proven to be anrneffective and practical tool in challenging deepwater projects with complicated seafloor geology. Amongrnits strengths are the fact that least-cost path routing offers a reproducible logical framework that is wellrnsuited to sensitivity analysis and refinement as new and more detailed data become available. At leastrnthree major kinds of uncertainity enter into the calculations— bathymetric uncertainty of seafloor digitalrnelevation models and their derivative products, locational (geologic) uncertainty regarding the sizes andrnlocations of mapped seafloor features and hazards, and cost uncertainty regarding the weighted relativerncosts assigned to seafloor features and hazards. Because these different kinds of uncertainties can berndifficult to quantify and they likely interact in ways that are difficult to foretell, one pragmatic approachrnis to use resampling-based stochastic conditional simulation of composite geocost surfaces to assess thernsensitivity of calculated routes to input data uncertainties. Characteristic dimensions of seafloor featuresrnor geohazards can be used to estimate the number of random points necessary to characterize major orrnfirst-order features known with confidence; smaller features are then simulated using random distributionsrnwith the same statistical properties and spatial correlation structures of the sampled points. The final resultrnis not a single route, but rather a cloud of routes defining a most likely route corridor for furtherrnassessment and route refinement.