(122c) Experimental and Numerical Studies on the Phenomena of Interfacial Mixing in Pipelines Processing Lube-Oil

Pipelines are considered the most crucial mode of transportation in the chemical and petroleum industry. While a single-product pipeline system is comparatively easier to manage, The multi-product pipeline system poses many difficulties due to the high level of operational complexities (Kirschstein 2018). In the multi-product pipeline systems, different types and grades of products are successively transferred in batches. In these pipelines, a major challenge that has been long-standing and economically significant is the interfacial mixing of consecutive product batches when switching from one product to another. The changeover operation leads to the generation of contaminated/ mixed products at the interface of adjacent batches. The commingled product does not match the desired specifications of either of the two batches and is categorized as a downgraded product, which cannot be used for the desired application. To address this challenge, over the years, extensive research has been carried out, and systematic scheduling plans have been developed by industries (Liao et al. 2019; Dimas et al. 2018). Through these plans, the volume of contamination is reduced by preventing the products that have significant differences in physical properties from being processed adjacently. However, these scheduling plans are not completely efficient and therefore the industries undergo high economic losses due to the generation of large volumes of mixed products. Hence, another way to address this problem is to understand the interfacial mixing by studying the transport properties of the fluids (Patrachari and Johannes 2012). Over the years, several mathematical models have been reported by researchers for estimating the volume of the mixed fluid (Ramanujan 2008). The reported mathematical models are primarily empirical in nature and case study specific. To this end, our work focuses on the case study of a lube-oil processing pipeline that manufactures and packages over a thousand different products throughout the year.

Typically, to avoid contamination due to mixing, the following operations are adopted by a generic lube-oil industry, the pipelines are first gravity drained and then blown using compressed air. However, due to the high viscosities of the lube oil, gravity draining and compressed air blowing are not sufficient for efficiently cleaning the pipelines and therefore residues of the prior oil remain adhered to the inner pipeline walls. Hence, followed by air-blowing the pipelines are flushed by using the lube oil from the batch that is processed next. During the flushing operation, the residual oil from the pipeline walls diffuses into the fresh batch of oil and results in contamination. The flushing is continued until the new batch of oil meets the desired specifications when tested for various physical properties in the laboratory. The existing operation often results in excessive flushing, is based on trial and error, and leads to improper resource consumption and energy conservation. Hence, our work aims to help the lube-oil industries make informed decisions and minimize the generation of mixed oil.

For conducting an efficient flushing operation, the knowledge of mass transfer aspects plays a crucial role. It is important to accurately estimate the oil concentration distribution, flow rate, flow regimes, desired physical properties of the new batch (viscosity, density), pipeline parameters (length, diameter), pressure drop, and friction factor (Riazi 2005). In this work, we study these key predictive parameters to arrive at an optimum flushing operation using CFD (Computational Fluid Dynamics) simulations. Our results will give insights to help industries conduct an effective flushing operation. The knowledge of the key parameters will help to optimize the operation. Furthermore, our application tool will enable the lube oil plant operators to estimate the specific volumes of flush required to meet the desired specifications of a new batch. Thus, our work will improve the environmental, economic, and resource management footprint of the existing processing operations in the lube-oil industries.

References

Dimas, Diovanina, Valéria V. Murata, Sérgio M. S. Neiro, Susana Relvas, and Ana Paula Barbosa-Póvoa. 2018. “Multiproduct Pipeline Scheduling Integrating for Inbound and Outbound Inventory Management.” Computers & Chemical Engineering 115 (July): 377–96. https://doi.org/10.1016/j.compchemeng.2018.04.025.

Kirschstein, Thomas. 2018. “Planning of Multi-Product Pipelines by Economic Lot Scheduling Models.” European Journal of Operational Research 264 (1): 327–39. https://doi.org/10.1016/j.ejor.2017.06.014.

Liao, Qi, Pedro M. Castro, Yongtu Liang, and Haoran Zhang. 2019. “Computationally Efficient MILP Model for Scheduling a Branched Multiproduct Pipeline System.” Industrial & Engineering Chemistry Research 58 (13): 5236–51. https://doi.org/10.1021/acs.iecr.8b06490.

Patrachari, Anirudh R., and Arland H. Johannes. 2012. “A Conceptual Framework to Model Interfacial Contamination in Multiproduct Petroleum Pipelines.” International Journal of Heat and Mass Transfer 55 (17): 4613–20. https://doi.org/10.1016/j.ijheatmasstransfer.2012.04.017.

Ramanujan, Anirudh. 2008. “Deterministic Models to Explain the Phenomenon of Interfacial Mixing in Refined Products Pipelines,” 177.

Riazi, M. R. 2005. Characterization and Properties of Petroleum Fractions. W. Conshohocken, PA: ASTM International.