EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Solvent Extraction of Hydrocarbons from Petroleum Sludge: Experimental Optimisation Using Response Surface Methodology and Artificial Neural Networks

Raiane Hasseine, Abdelmalek Hasseine, Toufik Chaabane, Merzougui Abdelkrim, Djemoui Laiadi

Journal of Applied Engineering Science and Technology · 11.07.2026 · 6 (2)

DOI: 10.69717/jaest.v6.i2.177 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Applied Engineering Science and Technology, online erschienen am 11.07.2026. Verfasst von Raiane Hasseine und 4 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The sustainable management of petroleum sludge from refinery storage tanks has become a growing environmental concern. These semi-solid wastes, rich in hydrocarbons, present significant challenges due to their toxicity, persistence in natural ecosystems, and the absence of dedicated treatment pathways. Nevertheless, their high content of heavy hydrocarbons also offers considerable potential for energy recovery, which remains largely underexploited in the national context. This study aims to assess the efficiency of solvent extraction for hydrocarbon recovery from petroleum sludge, with a particular focus on n-hexane as the extracting solvent. Owing to its non-polar nature and strong affinity for light to intermediate petroleum fractions, n-hexane enables selective extraction at relatively low energy cost. A rigorous experimental methodology was applied, combining Design of Experiments (DOE), Response Surface Methodology (RSM), and Artificial Neural Network (ANN) to model the extraction yield as a function of temperature, agitation time, and solvent volume. The models were then coupled with a genetic algorithm to determine the optimal operating conditions. The ANN model outperformed RSM in terms of prediction accuracy (R2 = 0.9917 vs. 0.9891, RMSE = 0.71 % vs. 0.81 %). The genetic algorithm coupled with ANN predicted a maximum recovery yield of 65.8 % at a sludge-to-solvent mass ratio of 0.166, in good agreement with the experimental maximum of 65.0 % (ratio 0.172) and with literature data. These results confirm the potential of solvent extraction as a promising pathway for the valorisation of petroleum sludge in Algeria and demonstrate the effectiveness of ANN-based modelling for process optimisation. Highlights First study in Algeria combining solvent extraction with RSM and ANN for petroleum sludge valorisation. ANN outperformed RSM in predicting extraction yield (R2 = 0.9917; RMSE = 0.71%). ANN-GA predicted 65.8% recovery, matching the experimental yield of 65.0%. Solvent extraction recovers hydrocarbons while reducing the volume and toxicity of petroleum sludge.

Zitieren

APA

Hasseine, R., Hasseine, A., Chaabane, T., Abdelkrim, M., & Laiadi, D. (2026). Solvent Extraction of Hydrocarbons from Petroleum Sludge: Experimental Optimisation Using Response Surface Methodology and Artificial Neural Networks. Journal of Applied Engineering Science and Technology, 6(2). https://doi.org/10.69717/jaest.v6.i2.177

Vancouver

Hasseine R, Hasseine A, Chaabane T, Abdelkrim M, Laiadi D. Solvent Extraction of Hydrocarbons from Petroleum Sludge: Experimental Optimisation Using Response Surface Methodology and Artificial Neural Networks. Journal of Applied Engineering Science and Technology. 2026;6(2). doi:10.69717/jaest.v6.i2.177

BibTeX
@article{Hasseine2026,
  author = {Hasseine, Raiane and Hasseine, Abdelmalek and Chaabane, Toufik and Abdelkrim, Merzougui and Laiadi, Djemoui},
  title = {Solvent Extraction of Hydrocarbons from Petroleum Sludge: Experimental Optimisation Using Response Surface Methodology and Artificial Neural Networks},
  journal = {Journal of Applied Engineering Science and Technology},
  year = {2026},
  volume = {6},
  number = {2},
  doi = {10.69717/jaest.v6.i2.177}
}
Kostenlos starten →