Abstract

Pipeline capacity calculations are used to accept transport commitments, and on a daily basis depend on what is inside de pipe and what will be delivered into the pipe. In some cases, there is no time to prepare and stabilize product properties but only to assure maximum or minimum quality requirements so capacity cannot be precisely calculated. In this uncertainty scenario, this work presents a throughput prediction methodology based on a novel similarity measure of line fill and statistical analysis of observed quantity and quality values, in order to achieve proper throughput prediction. The proposed similarity measure represents the line fill of a pipe as an ordered set depicted as a vectorial path. This approach identifies days with similar initial line fill and by fetching and processing throughput of similar days, gives an estimated throughput for the analyzed day.

This content is only available via PDF.
You can access this article if you purchase or spend a download.