@article{10.11648/j.ajaic.20200402.12,
  author = {Ikyerga Emmanuel and Alawode Adeolu and Falode Olugbenga Adebanjo},
  title = {Prediction of Asphaltene Precipitation During Gas Injection Using Hybrid Genetic Algorithm and Particle Swarm Optimisation},
  journal = {American Journal of Applied and Industrial Chemistry},
  volume = {4},
  number = {2},
  pages = {21-30},
  doi = {10.11648/j.ajaic.20200402.12},
  url = {https://doi.org/10.11648/j.ajaic.20200402.12},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ajaic.20200402.12},
  abstract = {Asphaltenes are precipitated and deposited during gas injection and this causes pore throat reduction, permeability reduction and wettability reversal. The result is reduced oil produced thereby leading to sizable revenue loss by field operators. To mitigate or completely prevent the occurrence of this phenomenon, this work has utilised Hybrid Genetic Algorithm Particle Swarm Optimisation-Artificial Neural Network (HGAPSO-ANN) for predicting the amount of asphaltenes deposited in the reservoir during gas injection. A number of methods are available for predicting the amount of asphaltenes deposited but some of them are either too expensive to execute or fraught with errors and deviations. This is due to the nature of asphaltene which is complicated and ambiguous. Some of the methods in existence include correlation with solvent properties, thermodynamic models and recently connectionist models (neural networks). There is however, no publication in the literature on using hybrid algorithms with neural networks to predict asphaltene precipitation during gas injection and this becomes an interesting area of research considering the enormous benefits that would be obtained from a robust hybrid asphaltene precipitation prediction model. The developed model performed well with an AARE of 0.09. This is lower than AARE values reported by Hue et al (2000), Rostami and Manshad (2010), Manshad et al (2015) which were 0.183, 0.153, and 0.121 respectively From the results of the model it can be seen that HGAPSO-ANN is more accurate in predicting asphaltene precipitation than other existing predictive models consulted. This method can therefore, be used as a decision making tool by field operators to set up procedures for the prevention or mitigation of asphaltene precipitation during gas injection. This will help prevent revenue losses and increase profitability of recovering hydrocarbons using gas injection.},
 year = {2020}
}

Copy | Download

TY  - JOUR
T1  - Prediction of Asphaltene Precipitation During Gas Injection Using Hybrid Genetic Algorithm and Particle Swarm Optimisation
AU  - Ikyerga Emmanuel
AU  - Alawode Adeolu
AU  - Falode Olugbenga Adebanjo
Y1  - 2020/10/26
PY  - 2020
N1  - https://doi.org/10.11648/j.ajaic.20200402.12
DO  - 10.11648/j.ajaic.20200402.12
T2  - American Journal of Applied and Industrial Chemistry
JF  - American Journal of Applied and Industrial Chemistry
JO  - American Journal of Applied and Industrial Chemistry
SP  - 21
EP  - 30
PB  - Science Publishing Group
SN  - 2994-7294
UR  - https://doi.org/10.11648/j.ajaic.20200402.12
AB  - Asphaltenes are precipitated and deposited during gas injection and this causes pore throat reduction, permeability reduction and wettability reversal. The result is reduced oil produced thereby leading to sizable revenue loss by field operators. To mitigate or completely prevent the occurrence of this phenomenon, this work has utilised Hybrid Genetic Algorithm Particle Swarm Optimisation-Artificial Neural Network (HGAPSO-ANN) for predicting the amount of asphaltenes deposited in the reservoir during gas injection. A number of methods are available for predicting the amount of asphaltenes deposited but some of them are either too expensive to execute or fraught with errors and deviations. This is due to the nature of asphaltene which is complicated and ambiguous. Some of the methods in existence include correlation with solvent properties, thermodynamic models and recently connectionist models (neural networks). There is however, no publication in the literature on using hybrid algorithms with neural networks to predict asphaltene precipitation during gas injection and this becomes an interesting area of research considering the enormous benefits that would be obtained from a robust hybrid asphaltene precipitation prediction model. The developed model performed well with an AARE of 0.09. This is lower than AARE values reported by Hue et al (2000), Rostami and Manshad (2010), Manshad et al (2015) which were 0.183, 0.153, and 0.121 respectively From the results of the model it can be seen that HGAPSO-ANN is more accurate in predicting asphaltene precipitation than other existing predictive models consulted. This method can therefore, be used as a decision making tool by field operators to set up procedures for the prevention or mitigation of asphaltene precipitation during gas injection. This will help prevent revenue losses and increase profitability of recovering hydrocarbons using gas injection.
VL  - 4
IS  - 2
ER  - 

Copy | Download