Analysis of field data indicates that slugs generated in a near horizontal flowline can develop a long high void zone in front of a low void zone as the slugs move upwards in a high riser. The two void zones are separated by a relatively sharp void front. Reanalysis of Sintef Two-phase Flow Laboratory slug flow data supports the above hypothesis. However, both the field data and the lab data are gamma densitometer time traces that only give the average density in a cross section and the question is whether the low density (high void) zone is liquid continuous and a part of the slug or not.
The present work therefore investigates experimentally how pipeline-riser slugs develop when the slugs enters and travels up a vertical riser. The experiments were performed in the Well Flow Loop at IFE. The loop was configured with a 1080D long flowline entering a 150D long vertical riser. The flow has been monitored by three gamma densitometers and one capacitance tomography system at different positions. Close to the top of the riser bubble flow, slug flow and churn flow are detected, depending on the flow rates. For the slug flow regime it is seen that the high void zone increases in length with height and that the void front is maintained relatively sharp. From the tomography data it has also been concluded that the high void zone is liquid continuous and thus a part of the slug.
Prediction of void fronts with the OLGA code fails mainly because the mass gradients are smeared out due to numerical diffusion. In the present work a less diffusive numerical scheme has therefore been implemented in the OLGA code. Applied together with the slug tracking model, slug flow with the measured void zones and void front development in the vertical riser can qualitatively be reproduced. However, it proved necessary to tune gas entrainment at slug fronts and gas liquid slip in the slugs to obtain reasonable agreement with the data.
