Evaluating the effects of input cost surface uncertainty on deep-water petroleum pipeline route optimization

A resampling-based stochastic simulation approach was used to evaluate the uncertainty that may be associated with geologically constrained least-cost path pipeline route optimization. A smoothed version of a composite geocost surface from a deep-water pipeline routing project was resampled and the results used to generate a series of equally probable cost surface realizations, which were in turn used as the basis for the same number of route optimizations. Eighty percent of the simulated routes followed a 500-2,000 m wide corridor nearly parallel to the baseline route (based upon complete information) between two hypothetical pipeline termini located about 25 km apart. Twenty percent followed an alternate corridor of approximately the same width. These results suggest that, while the general method of geologically constrained pipeline route optimization is a relatively robust one, uncertainties in geological input will at the least create a least-cost route corridor rather than a single least-cost route and may suggest realistic alternatives that must be critically evaluated in light of the available geological information.