Rapid assessment of crude oil fouling propensity to prevent refinery fouling

A process for producing a liquid transportation fuel in a commercial petroleum refinery, comprising:

• a) analyzing a sample of a crude oil feed stock comprising unrefined petroleum by a spectroscopy method selected from at least one of near-infrared spectroscopy and nuclear magnetic resonance spectroscopy to produce spectral data comprising discrete digitized data points and transforming the spectral data to produce a sample wavelet coefficients data according to wavelet theory by applying a mother wavelet consisting of a member of the Symlet family of mother wavelets, at the third level of decomposition or greater;
• b) training a genetic algorithm to classify the sample into one of two groups selected from a first group and a second group to produce a trained genetic algorithm,
• wherein the first group comprises multiple aliquots comprising crude petroleum that are each of distinct geologic origin and that are each characterized by a fouling thermal resistance of less than 0.002 hr-sq.ft.-° F./per British Thermal Unit,
• wherein the second group comprises multiple aliquots comprising crude petroleum that are each of distinct geologic origin and that are each characterized by a fouling thermal resistance of at least 0.002 hr-sq.ft.-° F./per British Thermal Unit,
• wherein the training comprises performing the analyzing of part a) on each of the aliquots to produce training wavelets coefficients data and presenting the training wavelets coefficients data obtained from each aliquot in the first group and each aliquot in the second group to an untrained genetic algorithm that is instructed to recognize subtle collective differences within the training wavelet coefficients data obtained from aliquots of the first group compared to the training wavelet coefficients data obtained from aliquots of the second group to produce a trained genetic algorithm, wherein the subtle collective differences represent distinguishing spectral features between the first group and the second group that allow the trained genetic algorithm to classify an aliquot as a member of the first or the second group;
• c) classifying the sample of a) as a member of either the first group or the second group by presenting the sample wavelets coefficients data of a) to the trained genetic algorithm, wherein the trained genetic algorithm performs the classifying by examining the data features that collectively predict whether a particular aliquot is a member of the first group or the second group;
• d) refining the crude oil feed stock comprising unrefined petroleum in a commercial petroleum refinery when the sample of a) is classified as a member of the first group, and not refining the crude oil feed stock comprising unrefined petroleum in a commercial petroleum refinery when the sample of a) is classified as a member of the second group.