Production Rate Measurement Optimization using Test Separator and In-Well Sound Speed

Abstract

The litmus test for downhole multiphase flowmeters is to compare the measured phase flow rates with the test-separator rates. In most cases, the composition of the measurand is required for flowmeters and is typically obtained from bottomhole fluid samples. Extracting and analyzing fluid samples is an expensive process mostly done at the initial stages of field development. Because the composition is not always available and is usually old, flowmeters may have subpar performances when compared to test separators. In this work, it is shown that once the test-separator data are available, measuring mixture sound speed downhole will ensure the optimization of a multiphase flowmeter system without needing to obtain new fluid samples. This novel method is demonstrated in a North Sea case history.

This method of optimizing flow rates is independent of the measurement device because the required flow velocity and sound speed measurements could be obtained from separate and generic devices. For example, the fluid bulk velocity and mixture sound speed could be measured by a point measurement device and a distributed acoustic sensing (DAS) system, respectively. The main challenge in a flow-velocity/sound-speed measurement system is to determine individual phase sound speeds so that the mixture phase fraction could be correctly modeled using Wood's mixture sound speed model. The phase fraction from the separator tests could be used as the target value to optimize the performance of the system. This is done with a backward calculation of the sound speed of individual phases. Pressure and temperature variations at measurement locations as well as the pipe compliance effects are also accounted for in this approach. Following the adjustment of individual phase sound speeds, a forward calculation using the same model yields a phase fraction close to the actual value, which could be improved further by an iterative approach.

A downhole optical flowmeter in a North Sea field measured mixture velocity and sound speed. Well test results indicated that the water cut was underreported and phase flow rates did not match the test-separator rates. Instead of halting production and going through a fluid sample analysis cycle, the test-separator water cut was used as the target value to optimize oil phase sound speed using Wood's model with a backward calculation. The difference in oil sound speeds was extrapolated to other pressure and temperature conditions, and forward calculation showed that separator tests and flowmeter measurements closely matched. Subsequent flowmeter and test-separator data confirmed excellent agreement.

Using well test data and sound speed to optimize phase flow rates is a novel method that has not been previously demonstrated. This method is independent of device type, broadly applicable, and furthers the understanding of multiphase flow measurement.