Prediction performance advantages of deep machine learning algorithms for two-phase flow rates through wellhead chokes

Hossein Shojaei Barjouei,Shadfar Davoodi,David A Wood,Nima Mohamadian,Hamzeh Ghorbani,Jamshid Moghadasi,Hossein Saberi

doi:10.1007/s13202-021-01087-4

Abstract

Two-phase flow rate estimation of liquid and gas flow through wellhead chokes is essential for determining and monitoring production performance from oil and gas reservoirs at specific well locations. Liquid flow rate (QL) tends to be nonlinearly related to these influencing variables, making empirical correlations unreliable for predictions applied to different reservoir conditions and favoring machine learning (ML) algorithms for that purpose. Recent advances in deep learning (DL) algorithms make them useful for predicting wellhead choke flow rates for large field datasets and suitable for wider application once trained. DL has not previously been applied to predict QL from a large oil field. In this study, 7245 multi-well data records from Sorush oil field are used to compare the QL prediction performance of traditional empirical, ML and DL algorithms based on four influencing variables: choke size (D64), wellhead pressure (Pwh), oil specific gravity (γo) and gas–liquid ratio (GLR). The prevailing flow regime for the wells evaluated is critical flow. The DL algorithm substantially outperforms the other algorithms considered in terms of QL prediction accuracy. The DL algorithm predicts QL for the testing subset with a root-mean-squared error (RMSE) of 196 STB/day and coefficient of determination (R2) of 0.9969 for Sorush dataset. The QL prediction accuracy of the models evaluated for this dataset can be arranged in the descending order: DL > DT > RF > ANN > SVR > Pilehvari > Baxendell > Ros > Glbert > Achong. Analysis reveals that input variable GLR has the greatest, whereas input variable D64 has the least relative influence on dependent variable QL.

Highlights

One of the key factors in determining the production performance of oil/gas reservoirs is to establish value of the variables influencing two-phase flow rate in oil and gas wells (Lak et al 2014; Choubineh et al 2017; Ghorbani et al 2017b, c, 2019; Mirzaei-Paiaman and Salavati 2013)
Two‐phase flow rate prediction accuracies achieved by machine learning (ML), deep learning (DL), and traditional mathematical models
In order to make comparisons between ML, DL, and traditional empirical methods for two-phase flow rate (QL) prediction, and to determine the best performance accuracy achieved for all 7245 data records in Sorush oil field, the following method is applied: Two-phase flow rate (QL) prediction accuracies achieved by the training subset (~ 80%), the testing subset (~ 20%), and the complete dataset (5796 data records) are presented in Tables 12, 14, respectively

Summary

Introduction

One of the key factors in determining the production performance of oil/gas reservoirs is to establish value of the variables influencing two-phase flow rate in oil and gas wells (Lak et al 2014; Choubineh et al 2017; Ghorbani et al 2017b, c, 2019; Mirzaei-Paiaman and Salavati 2013). The back pressure created by them helps to control pressure and flow rate (Mirzaei-Paiaman and Salavati 2013; Guo 2007; Al-Attar 2009; Chong et al 2009; Nasriani et al 2016; Mirzaei-Paiaman and Salavati 2012; Omana et al 1969; Poettmann and Beck 1963). The increase in well-bore pressure associated with well reducers is due to the presence of oil-soluble gas or gas condensate, which forms in the wellbore due to pressure drop in the production tubing. In variable chokes the bore or aperture can be increased or decreased to adjust fluid flow rate through it (Gorjaei et al 2015; Elhaj et al 2015)

Methods

Results

Discussion

Conclusion