Abstract Slugs are intermittent structures observed in horizontal co-current gas-liquid two-phase flow. Published data examining horizontal flow indicate the existence of intermittent sub-flow patterns. There is limited work performed on the study of this sub-categorisation. This paper presents a method, exploiting non-intrusive Electrical Capacitance Tomography (ECT) measurements, to assess and to characterise the spatial and temporal features of horizontal gas-oil flow. A logic-based algorithm is proposed to automatically identify and classify flow regimes. This algorithm uses the spatial distribution of the phases, based on measurements of the intrinsic dielectric properties of the fluids. The measurements were performed at the NEL UK national standard multiphase flow loop using a high-speed dual-plane ECT sensor array. The ECT measurements identified intermittent liquid structures with a sustained gas-core in the central region of the pipe cross-section with outer features analogous to those of liquid slugs. Assessment of the frequency and time span of the periodic structures showed that gas-core slugs are, typically, shorter than full slug structures but are more frequent for the same given time period. For both flow structures, the mixture velocity was a determinant factor, with gas-core slugs dominating the flow at higher velocities. The experimental work led to the modification of existing flow pattern transition models. The modified flow regime maps suggested in this study aim to widen the applicability of the results by enabling the prediction of gas-core slug flow for similar experimental and operating conditions. The present work comprises the initial steps towards automatic flow regime identification in multiphase flow metering. Automatic flow regime recognition has the potential to lead to better process control and improved accuracy in production allocation.

中文翻译：

---

关于气芯段塞的生活和习性：间歇性水平两相流的表征

摘要 段塞是在水平并流气液两相流中观察到的间断结构。检查水平流的已发表数据表明存在间歇性子流模式。对该子分类的研究开展的工作有限。本文提出了一种方法，利用非侵入式电容断层扫描 (ECT) 测量来评估和表征水平油气流动的空间和时间特征。提出了一种基于逻辑的算法来自动识别和分类流态。该算法基于对流体固有介电特性的测量，使用相的空间分布。测量是在 NEL UK 国家标准多相流回路中使用高速双平面 ECT 传感器阵列进行的。ECT 测量确定了在管道横截面的中心区域具有持续气芯的间歇性液体结构，其外部特征类似于液体段塞。对周期性结构的频率和时间跨度的评估表明，气芯段塞通常比完整段塞结构短，但在同一给定时间段内更频繁。对于这两种流动结构，混合速度是一个决定因素，气芯段塞在较高速度下主导流动。实验工作导致对现有流型转换模型的修改。本研究中提出的修改后的流态图旨在通过预测类似实验和操作条件下的气芯段塞流来扩大结果的适用性。目前的工作包括在多相流量计量中实现自动流态识别的初始步骤。自动流态识别有可能导致更好的过程控制和提高生产分配的准确性。