Analysis of the performance of a crude-oil desalting system based on historical data

Our study is keyed to the development of a methodological approach to assess the workflow and performance associated with the operation of a crude-oil desalting/demulsification system. Our analysis is data-driven and relies on the combined use of (a) Global Sensitivity Analysis (GSA), (b) machine learning, and (c) rigorous model discrimination/identification criteria. We leverage on an extensive and unique data-set comprising observations collected at a daily rate across a three-year period at an industrial plant where crude oil is treated through a combination of demulsification/desalting processes. Results from GSA enable us to quantify the system variables which are most influential to the overall performance of the industrial plant. Machine learning is then applied to formulate a set of candidate models whose relative skill to represent the system behavior is quantified upon relying on model identification criteria. The integrated approach we propose can then effectively assist to (a) modern and reliable interpretation of data associated with performances of the crude oil desalting process and (b) robust evaluation of future performance scenarios, as informed by historical data.