Full Content is available to subscribers

Subscribe/Learn More  >

Surrogate Models for Predicting the Performance of a Liquid-Liquid Cylindrical Cyclone Separator

[+] Author Affiliations
Jorge E. Pacheco, Miguel A. Reyes

Universidad Simón Bolívar

Paper No. IMECE2006-14759, pp. 51-58; 8 pages
  • ASME 2006 International Mechanical Engineering Congress and Exposition
  • Process Industries
  • Chicago, Illinois, USA, November 5 – 10, 2006
  • Conference Sponsors: Process Industries Division
  • ISBN: 0-7918-4777-2 | eISBN: 0-7918-3790-4
  • Copyright © 2006 by ASME


Liquid-Liquid Cylindrical Cyclone (LLCC) separators are devices used in the petroleum industry to extract a portion of the water from the oil-water mixture obtained at the well. The oil-water mixture entering the separator is divided due to centrifugal and buoyancy forces in an upper (oil rich) exit and a bottom (water rich) exit. The advantages in size and cost compared with traditional vessel type static separators are significant. The use of LLCC separators has not been widespread due to the lack of proven performance prediction tools. Mechanistic models have been developed over the years as tools for predicting the behavior of these separators. These mechanistic models are highly dependent on the inlet flow pattern prediction. Thus, for each specific inlet flow pattern a sub-model has to be developed. The use of surrogate models will result in prediction tools that are accurate over a wider range of operational conditions. We propose in this study to use surrogate models based on a minimum-mean-squared-error method of spatial prediction known as Kriging. Kriging models have been used in different applications ranging from structural optimization, conceptual design, multidisciplinary design optimization to mechanical and biomedical engineering. These models have been developed for deterministic data. They are targeted for applications where the available information is limited due to the cost of the experiments or the time consumed in numerical simulations. We propose to use these models with a different framework so that they can manage information from replications. For the LLCC separator a two-stage surrogate model is built based on the Bayesian surrogate multistage approach, which allows for data to be incorporated as the model is improved. Cross validation mean squared error measurements are analyzed and the model obtained shows good predicting capabilities. These surrogate models are efficient and versatile predicting tools that do not require information about the physical phenomena that drives the separation process.

Copyright © 2006 by ASME



Interactive Graphics


Country-Specific Mortality and Growth Failure in Infancy and Yound Children and Association With Material Stature

Use interactive graphics and maps to view and sort country-specific infant and early dhildhood mortality and growth failure data and their association with maternal

Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In