Online estimation of wax deposition thickness in single-phase sub-sea pipelines based on acoustic chemometrics: A feasibility study

• Published in: Fuel (Guildford) Vol. 105; pp. 718 - 727
• Main Authors: Halstensen, Maths, Arvoh, Benjamin Kaku, Amundsen, Lene, Hoffmann, Rainer
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01-03-2013; Elsevier
• Subjects: Acoustic chemometrics; Applied sciences; Energy; Energy. Thermal use of fuels; Exact sciences and technology; Fuels; Partial least square regression (PLS-R); Principal component analysis (PCA); Single-phase flow; Wax deposition; Norway; Partial least square regression (PLS-R); Single-phase flow; Acoustic chemometrics; Principal component analysis (PCA); Wax deposition; Acoustics; Paraffin; Gas oil; Sea; Wax; Pipeline; PLS regression; Crude oil; Feasibility; Chemometrics; Principal component analysis
• ISSN: 0016-2361; 1873-7153
• DOI: 10.1016/j.fuel.2012.10.004

► Estimation of wax deposition thickness in pipelines based on acoustic chemometrics. ► Partial least squares regression was applied in estimating wax deposition thickness. ► Validation results for the wax deposition thickness model were satisfactory. ► Further investigations and subsequent implementation are recommended.
Wax deposition in sub-sea oil producing pipelines is a concern to the oil producing companies. The deposition of wax in pipelines can cause serious economic implications if not monitored and controlled. Several researchers have developed models and investigated the deposition of wax in crude oil pipelines. As of today, there is no off the shelf instrument available for reliable online estimation of the wax deposition thickness in sub-sea pipelines. Acoustic chemometrics was applied to investigate the potential for online estimation of wax deposition thickness in sub-sea pipelines. This feasibility study was carried out as a so called piggy back on experiments performed at Statoil research centre in Porsgrunn, Norway with real crude oil or waxy gas condensate. The first investigations focused on the repeatability of the acoustic chemometric technique followed by online prediction of the wax deposition thickness in a single-phase oil flow pipeline. A partial least squares regression model was calibrated and validated with a totally independent data set. The calibrated model had a root mean squared error of prediction of 0.28mm with a final wax deposition thickness of 3.36mm, a slope of 0.91 and R2 of 0.83 which were satisfactory results. The effect of varying oil flow rates on the wax deposition thickness was also investigated. The preliminary results showed the need for further investigations based on a robust experimental design and sample pre-processing. The general conclusion that can be drawn from this feasibility study was that the potential of adapting the acoustic chemometric technique for online estimation of the wax deposition thickness exist and must be further investigated.