Comprehensive investigation and evaluation for the different multiphase flow correlations in gas wells

Abstract

Gas production represents an important source of energy used in different industries. Accurate production forecast of gas wells helps in decision making and maximizing reserves.

Petroleum engineers model existing conditions of the fluid flow to help them in predicting wells’ future performance. The modeling of single-phase flow is considered the simplest case, but considering multiphase flow (oil, gas, and water) makes it more challenging.

We studied and evaluated the different multiphase flow correlations using two commercial software (A and B). A database of 537 bottom hole flowing pressure data points was used which covered 33 different flowing, well geometry, and tubing diameter conditions. 14 correlations were used from Software “A” and 26 correlations were used from Software “B” to calculate the pressure drop in the tubing for each case. The target was to determine the best correlation(s) for each case and establish guidelines to help petroleum engineers in selecting the best correlation when actual bottom-hole flowing pressure data is unavailable. We could determine the expected average percent error (APE) and average absolute percent error (AAPE) for each correlation for different flow conditions. Also, for different conditions, we calculated the different statistical parameters for the calculated pressure from each correlation including normalized root mean square error (NRMSE), coefficient of determination (R²), and the standard deviation (STD). We also created a map to show which software will give the lowest error for the different flow and well parameters. Then, a validation database was used to confirm the previously determined results using a new set of data. The results were validated with a percentage of 91% for Software “A” and 89% for Software “B”.

For both software programs, Gray’s correlation was found to be the best correlation for the entire database. It gives AAPE of 5.73% and 9.72% for software A and B, respectively.

This paper will help petroleum engineers select the best multiphase flow correlation(s) in different flow and well conditions. Also, it will give them guidance for the critical conditions which give a wide range of the calculated pressure drop using the different multiphase flow correlations.