Development of a computational method to estimate fluid flow rate in oil wells with an electrical submersible pump

Abstract

A significant challenge in the oil and gas industry is the simultaneous measurement of commingled gas, oil and water streams, using the three phase test separator or multiphase flow meter (MPFM). The major issue in the application of using either test separator or MPFM in oil field is the uncertainty of the measurements, due to different process and operations conditions. To date, there are no sets of rules proving the technique for comparison between the test separator and the multiphase flow meter. Hence, there is a need for more accurate and reliable methods to be adapted as alternatives to the current flow rate measurements techniques, which must be capable of working at any fluid composition and production flow environment conditions. At first, this thesis will show, through an experimental study, wide variations of the liquid rate measurements between the conventional test separator and MPFM at several periods of time. Then, it proposes a new computational method to estimate the electrical submersible pump (ESP) oil well flow rates. The research idea is to close the wellhead wing valve as the ESP is kept running normally, and the wellhead flowing pressure before well shut-in and the build-up of wellhead flowing pressure after the well shut-in is measured. The total shut-in time period is recorded, and it is dependent on the individual oil well production conditions. Explicit physics concepts for estimations of the multiphase fluid flow rate in a vertical pipe were employed. The formulas deal with changes of fluid flow parameters along the vertical pipe in the well, as a function of pressure and temperature variations with depth. A Microsoft visual basic program was also developed based on the oil mechanistic and empirical equations that can estimate oil rate for ESP oil wells. The new method was applied on 48 ESP oil wells in North African oil fields and lead to very reliable estimation results, which have about a +/-10% relative error. As a result, a regression correlation equation was developed based on the computational results. OLGA software has been used to make comparison with multiphase flow model available in the OLGA software against each nominated ESP oil well parameters obtained from measured field data. The objective was to verify the obtained shut-in wellhead pressure after closing the choke wing valve (WHPa) from the measured field data with the obtained shut-in wellhead pressure valve from the simulation model. The simulation results showed that the estimated WHPa are in agreement with the measured WHPa. The relative errors for individual oil field are within accuracy standard specification (typically +/- 5%). The overall relative errors are low and within acceptable uncertainty range, where the aggregate relative error for all wells was less than +/-4%. Therefore, the results have demonstrated that the new computational method can be applied to all fluid types and under any production conditions. Generally, the results show that the new computational approach is more accurate when compared with test separator measurement within the specified range of accuracy.