Flow rate measurement of individual oil well using multivariate thermal analysis

Flow measurement of individual oil wells mainly consisted of using test-separator units with many limitations including time and cost consuming, uncertainty of well isolation, and need to closing the co-line wells. The multivariate regression of thermogravimetric data were used to predict the accurate productivity of oil wells using a single sample-point at the blend-oil pipe-line. The results revealed that the Weighted Sum Model was appropriate for oils, which their thermal traces were significantly different from each other. The thermogravimetric data was arranged in a two-way array by taking the number of oil samples and the thermal signal intensities as columns and rows, respectively. The data matrix was decomposed using the commend of Matrix Left Division in MATLAB and the rank of individual oil wells was determined as a column vector. The error vector corrected the blend equation by considering the catalytic pyrolysis and eutectic point in higher and lower temperatures, respectively. The model predicted the accurate productivity of individual oil wells in an offshore oil field after 2010.

Figure optionsDownload as PowerPoint slideHighlights
► Multivariate thermogravimetric data predicted the oil wells productivity.
► Two-way multivariate data was decomposed by MATLAB’s Matrix Left Division.
► The model predicted the accurate productivity of oil wells in field tests.