Parallel calibration of multiphase flow meters (MPFM) based on measurements of phase streams in separators

Abstract

The Alvheim field suffers from significant production deferrals of oil and gas, during calibration of multiphase flow meters used in ownership allocation. This thesis has developed an algorithm solving a new method, which effectively negates these deferrals. This is done through an object-oriented data science approach, in creating a framework for performing these calibrations in an elegant and efficient manner. The algorithm has been tested and compared to real world data and shows promising results. The tests during April 2019 showed an increase of 15000bbl of oil production during parallel calibration compared to a normal calibration. The Cognite Data Fusion repository helped in streamlining the development process with easy and swift access to process data. The algorithm was implemented and developed in the programming language Python. Additionally, this thesis covers the purpose and technical background of ownership allocation measurements and the systems and sensors involved in measurement and calibration. The details of the developed algorithm, and the calibration results are presented and discussed.