Empirical and Neural Network Modelling of Oil Recovery in Nano Assisted Enhanced Oil Recovery
In this chapter, an empirical model was developed to predict the oil recovery in enhanced oil recovery (EOR) application, based on rock permeability, rock wettability, particle size, and injection rate of nanofluids. The developed model efficiency is compared with several neural network models tha...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Published: |
Springer Science and Business Media Deutschland GmbH
2022
|
| Online Access: | https://www.scopus.com/inward/record.uri?eid=2-s2.0-85115439570&doi=10.1007%2f978-3-030-79606-8_23&partnerID=40&md5=1e6f703a19b5b5d67c2b829c0db0d9e7 http://eprints.utp.edu.my/28900/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In this chapter, an empirical model was developed to predict the oil recovery in enhanced oil recovery (EOR) application, based on rock permeability, rock wettability, particle size, and injection rate of nanofluids. The developed model efficiency is compared with several neural network models that are developed for the same purpose. A multi-layer feed-forward artificial neural network (ANN) model trained with an error back-propagation algorithm was employed for developing a predictive model. The model has considered input variables of particle size, rock permeability, rock wettability, nanofluid injection rate, and temperature, while percentage of oil recovery is the output. The scaled conjugate gradient (SCG) optimization algorithm was used to train the ANN model. An ANN with six hidden neurons was highly accurate in predicting the oil recovery. The developed ANN model accuracy was determined using statistical measures such as R- square and Mean-square error (MSE). The values obtained for the developed model are 1 and 0.0009 for training, 1 and 0.0009 for validation and, 1 and 0.0005 for testing data sets, respectively. The comparison between the developed neural network model and the polynomial fitting method concluded that the ANN is better in terms of accuracy for predicting oil recovery in EOR applications. © 2022, Institute of Technology PETRONAS Sdn Bhd. |
|---|