Experiments and Modelling for the Control of Riser Instabilities: Forsøk og simulering for regulering av ustabil strøm i stigerør

Abstract

Multiphase flowline-riser systems sometimes encounter fluctuations in flow rate, pressure and liquid hold-up characterized as severe slugging. Slugging is undesirable because the piping and downstream receiving facilities are exposed to extra loads from the fluctuations that may damage the equipment and cause abruptions in the production.Two types of severe slug behaviour exist: Type 1 (SS I), which has a slug length larger than the riser, and Type 2 (SS II), which has a slug length smaller than the riser. In the literature a number of criteria are described that mark the boundaries of the flow regime where such slugs exist. An example is given in the figure below. At high liquid flowrates, severe slugging of both types is eliminated because of hydro-dynamic slugs that appear in the flowline upstream of the riser base, or by 'Stability of severe slugging', which means that the fast liquid production that causes gas blow-out and depressurization of the flowline in severe slugging cannot happen. At high gas flowrates, the cycle is eliminated because the flow in the riser becomes stable; that a positive flow perturbation is impeded by increased pressure drop in the riser. At low liquid flowrates, severe slugging will be of type 2 (SS II). The bound-ary between the two is marked by the Bøe-criterion, which states that severe slug-ging will be of type 2 if the inflow of gas is sufficient to prevent the slug of growing in both directions and into the flowline.The main aim of the current study was to reproduce the severe slugging cycle in flow loop experiments, and to investigate whether modifications to the system could minimize or even fully eliminate such slugs. The experiments have been conducted in the 2 inch water/air flowline-riser loop at Shell s technology centre in Amsterdam (STCA). The first part of the experimental activity in the project focused on finding the operating region where severe slugging can occur (the severe slugging envelope). The purpose of the second part was to achieve knowledge of how this region could be minimized by minor modifications to the system. A number of options for such modifications exist, and gas-lift, active topside choking using the Shell Smart Choke, or a combination of both was considered. However, due to a limited time span of the project and also technical issues that caused delays in the start-up of experiments, only gas-lift was finally performed.It was indeed possible to eliminate the severe slugging cycle in the experiments by the injection of lift gas. The mechanism is most likely that the lift gas stabilizes the flow in the riser. Experiments were carried out at various conditions and details are given in the report. The experimental results were compared with quasi-static criteria for the boundaries of the slugging regime known from the open literature. Since the theoretical criteria are functions of flowline pressure, hold-up and flow regime, and since no simple correlations exist for these properties where a horizontal flow line is concerned, an Excel model has been built as part of this study that links to Shell s proprietary package for multiphase flow the Shell Flow Correlations. In addition, the system has been modelled in the dynamic multiphase flow simulator OLGA.