Classifying Crude Oil Emulsions Using Chemical Demulsifiers and Statistical Analyses

摘要

Crude oil emulsions are highly complex mixtures that can bestabilized by a number of naturally occurring species andconditions (e.g. asphaltenes, resins, acids, solids, solvency,viscosity, temperature, etc.). Emulsion resolution is oftenaccomplished using chemicals. Optimization of chemicaltreatment is generally accomplished in the field by bottletesting a large number of potential chemical intermediates andtheir combinations. Successful chemical formulations are ableto drop water rapidly, provide relatively clean interfaces, andproduce dry, saleable oil.The very nature of bottle testing produces a large amountof data. Some of the data provides information regardingwater drop, while part focuses on breaking the emulsion (i.e.producing dry oil and clean interfaces). To summarize resultsfrom multiple test sites, a highly structured bottle test wasdevised to make comparisons among different oilfieldemulsions. Thirty-eight chemical intermediates, tested at twodosages, were evaluated at nine different sites. From thetesting, ten bottle test performance parameters (four describingwater drop, three describing oil dryness, and three describingthe oil-water interface) were analyzed using several statisticalmethods: analysis of variance, multivariate correlations,cluster analysis, and principal component analysis. Theanalyses revealed a number of interesting trends. Forexample, the water drop and oil dryness parameters werefound to be more independent of one another than the waterdrop and interface parameters. These results suggest thatwater drop and oil dryness are likely governed by twodifferent mechanisms. This work has resulted in severalemulsion “maps” where crudes can now be classified withregard to their ability to drop water and break emulsion.