In maturing gas wells, low amounts of liquids (water, condensate) can lead to production problems (liquid loading, salt precipitation, scaling, etc.). Improved monitoring of these low amounts of liquids will enable better diagnostics and mitigation of production problems. The liquid fractions are mostly low and fluctuate due to production instabilities. Currently gas and liquid rates of individual wells are measured periodically during a well test and the combined gas and liquid flows of multiple wells are measured continuously at the production separator.
State of the art
Direct metering of very low amounts of liquids in a gas flow is difficult and inaccurate when liquid flow rates are strongly fluctuating. Virtual metering is a proven technology (see 2.1) for gas wells and oil wells and is already utilized for back allocation purposes. These techniques are mostly data-driven and their performance depends on the flow stability and the availability and quality of production data. However, gas flows and liquid fractions in mature gas wells are typically low and unstable. At the same time availability and quality of production data is often limited. For these reasons, currently available techniques for virtual metering are not sufficiently accurate for North Sea gas well.
In recent years, investigations have shown that low amounts of liquids in a gas stream can lead to a number of flow, temperature and/or acoustic phenomena. We propose to investigate if and how we can make use of these phenomena to improve quantification / reduce uncertainty of the liquid fraction in a gas flow. This project explores, develops and evaluates new algorithms for virtual metering We will evaluate these for their ability to obtain (real time) gas rates and liquid rates from production data typically for Dutch or North Sea wells to a level of accuracy that is useful for well diagnostics and production optimization.
Outcome of the research
If successful, the outcome of this research could be implemented in workflows for real-time back allocation and event detection in North Sea gas wells. We envision that for a specific field case the selection of (a combination) of approaches to virtual metering will depend on the available data (data type, quality, history etc.).
First (WP1) we will define requirements for virtual metering algorithms (flow rates, liquid fractions, required accuracy etcetera) and evaluate the available data and their respective reliability. A synthetic case will be defined for use in the next steps. Then (WP2) the various approaches for virtual metering will be evaluated theoretically, using synthetic data which is generated analytically or numerically (for instance via OLGA / OLGA-ROCX). Finally (WP3) involves gathering and evaluation of field data and evaluation / validation of the algorithms for virtual metering on field data.
Wintershall, Total, EBN will support the project. Schlumberger has expressed their willingness to contribute in kind to the project by providing software licenses to Olga / Olga Rocx.