One central aim of multiphase flow metrology is to evaluate and reduce the uncertainties in multiphase flow metering in the oil and gas industries. While for single-phase flow metrology there exists a well-established reference network with norms and standards, such a network is lacking for multiphase flow metrology. This leads to a high level of uncertainty in multiphase flow measurement systems reaching up to 20 per cent [1]. The main objective of the European research projects “Multiphase flow metrology in oil and gas production”(MultiFlowMet I) and “Multiphase flow reference metrology”(MultiFlowMet II) is to reduce this level of uncertainty. Therefore, a comprehensive experimental intercomparison on multiphase flow is conducted on the one hand. On the other hand, the process of flow pattern formation as well as the quantitative influence of relevant flow condition parameters are studied by computational fluid dynamics (CFD).

The great advantage of CFD is that it gives insight into areas that are hardly accessible by experiments. Therefore, simulations can help to understand flow pattern formation as well as their influence on the measurement process. However, before CFD simulations can be used for predicting flows, they need to be validated first. In this contribution, the validation of the multiphase flow simulations is done by comparison with experimental video observations, which have been recorded during the MultiFlowMet I project at NEL. A tool for video analysis has been implemented [2], which extracts the liquid level over time from the experimental observations. Further analysis of the extracted liquid level allows a quantitative comparison with the corresponding CFD data and therefore a validation of the multiphase flow simulation.