Experimental Analysis for Flush Time Predictions in Pipeline Operations during Multi-Product Changeovers

Pipelines are the most common mode of transport used in the petroleum and petrochemical industries, owing to the advantages of high reliability and favorable environmental impacts [1]. These industries employ a batch mode of transport using large diameter interconnected pipelines, known as multi-product pipelines, which offer an economical way of packaging or transporting the varied refined products. A major challenge in multiproduct pipeline operations is maintaining product integrity between batches during product changeovers. The conventional method of product transport involves using the new product as a means of displacing the previous product. This operation is known as flushing. [2], [3],.

This flushing process typically gives rise to multi-layered zones: (i) the leading zone, consisting of the product from the previous batch; (ii) the mixed zone, a blend of previous product and new product on the line; and (iii) the new oil zone, representing the incoming product. In current industrial practice, flushing conditions are often determined through trial-and-error. This approach tends to require an excess volume of high-quality new products, which not only results in product downgrading but also contributes to economic losses and environmental concerns.

This study presents an experimental investigation of the flushing operation using a pilot plant designed to simulate the industrial operation. Two flushing operations were studied: (a) direct oil-to-oil flush, which involves using the new oil to push out the previous oil from the pipeline, and (b) air-blowing and flushing, which introduces compressed air into the pipeline first to displace as much old oil as possible before introducing the new oil. In these operations, viscosity changes were continuously monitored using an inline viscometer integrated with a real-time data acquisition system. This approach allowed precise identification of the time at which the product met the required specifications. The experiments were conducted across multiple oils categorized by viscosity ranges, enabling the development of a standardized framework for optimizing the flushing process. Viscosity readings and time scales were normalized using the pilot plant system parameters, such as flow rate and volume, to determine the optimal flushing time. The resulting correlations allow for the prediction of flushing requirements based on product specifications and system parameters. Due to established geometric similarity between the pilot plant and the industry unit, these correlations can be applied at the industrial scale to predict the time required for flushing, thereby minimizing mixed oil generation and improving operations.

References

[1] D. Ding, Y. Liang, Y. Li, J. Sun, and D. Han, “Numerical Study of Trapped Solid Particles Displacement From the Elbow of an Inclined Oil Pipeline,” Computer Modeling in Engineering &Amp; Sciences, vol. 121, no. 1, 2019, doi: 10.32604/cmes.2019.07228.

[2] B. Gao et al., “Improved Design of Flushing Process for Multi-Product Pipelines,” presented at the Foundations of Computer-Aided Process Design, Breckenridge, Colorado, USA, Jul. 2024, pp. 137–144. doi: 10.69997/sct.171679.

[3] S. S. Jerpoth, “Integrated Computer-Aided Design Experimentation, and Optimization Approach for Perovskites and Petroleum Packaging Processes,” PhD Thesis, Rowan University, 2023. Accessed: May 07, 2024. [Online]. Available: https://search.proquest.com/openview/214a6a6c745bbf735189de2363cd1875/1…