Modelling of Electrical Submersible Pumps for Petroleum Multiphase Fluids, an Intelligent Approach Supported by a Critical Review and Experimental Results

Abstract

This paper initially reviews existing empirical models which predict head or pressure increase of two-phase petroleum fluids in electrical submersible pumps (ESPs), then, proposes an alternative model, a fully connected cascade (FCC in short) artificial neural network to serve the same purpose. Empirical models of ESP are extensively in use; while analytical models are yet to be vastly employed in practice due to their complexity, reliance on over-simplified assumptions or lack of accuracy. The proposed FCC is trained and cross-validated with the same data used in developing a number of empirical models; however, the developed model presents higher accuracy than the aforementioned empirical models. The mean of absolute prediction error of the FCC for the experimental data not used in its training, is 68% less than the most accurate existing empirical model.

Comparison of empirical models using experimental results of electrical submersible pump under two-phase flow: numerical and empirical model validation

“…The flow rate values in each series have varied within a specified range, and the acquired head has been collected. In Group 1, Turpin's [26] and Zhou's [11], and Mohammadzaheri's [32,33] models were developed for head prediction, whereas Sachdeva's [29] model was developed to predict the rise in pressure.…”

“…Mohammadzaheri [32,33] proposed an empirical model for predicting head degradation using a shallow artificial neural network (ANN), which has proven to be more accurate than all the previous models. By following the other empirical models mentioned above, the ANN has a single output of H m (Equation).…”

“…Numerous researchers, including Turpin [26], Cirilo [27], Romero [28], Sachdeva [29,30], Duran [31], Zhou [11], and Mohammadzaheri [32,33], have developed empirical models for estimating the ability of pump for handling two-phase flow conditions, while Romero [28] established a practical model based on Cirilo's [27] experimental results. Later, Duran [31] developed an empirical model to extend the operating range of Turpin [26] and Romero's [28] models.…”

“…Later, Duran [31] developed an empirical model to extend the operating range of Turpin [26] and Romero's [28] models. Moreover, Mohammadzaheri [32,33] introduced a new empirical model using a shallow artificial neural network (ANN) as a substitution for all previous empirical models to calculate the multiphase head and pressure increment. Section 2 details all these empirical models, their associated components, and their implementation range.…”

Research Progress and Prospects of Multi-Stage Centrifugal Pump Capability for Handling Gas–Liquid Multiphase Flow: Comparison and Empirical Model Validation

“…Additionally, the values of coefficients (K, Kc, E1, E2, E3, E4, and E5), used in Sachdeva [38] and Zhou's correlations [12], are given in Table 3. Furthermore, the correlations proposed by Mohammadzaheri [41,42,45] were developed using different ANN, as a replacement to existing empirical models. Their ANN models have shown better prediction of the two-phase performance compared to existing models.…”