In this work we investigated the fluid dynamics in a pilot-scale bubble column equipped with an asymmetric “tree” type sparger at superficial velocities between 0.01 and 0.37 m s–1. We present an extensive experimental data set consisting of overall holdup measurements, liquid velocity, gas velocity, and local gas volume fraction profiles as well as bubble size distributions. The second part of this work examines the modeling of the column using computational fluid dynamics (CFD). It is found that the computationally efficient single bubble size model used in this work offered satisfactory predictions of the complex flow patterns found inside bubble columns operated in the industrially relevant heterogeneous flow regime.

No abstract

Two-phase pipelines offer the potential for substantial cost savings in the offshore transport of hydrocarbons. Their design has been hampered by uncertainties in two-phase pressure drop relations, in flow regime determination, and in liquid slug length prediction. This uncertainty makes difficult the choice of pipe size and the design of downstream separation facilities. In order to reduce the uncertainty in two-phase flow pipeline design, Esso, Statoil, Mobil, Texaco, and Getty have sponsored the construction of the Norwegian Two-Phase Flow Laboratory now being operated by SINTEF. A large pipeline-riser expe imental facility has bee constructed to obtain laboratory quality data in a field scale apparatus. In addition, analytical modeling о two-phase flow is proceeding to provide a framework for interpretation of the experimental measurements. The experimental facility, results of experimental measuremen s, and analytica modeling techniques are described. The two-phase flow data and analysis resulting from this project promise to reduce substantially the uncertainty in the design of two-phase pipelines and downstream separation facilities.

In order to predict the consequences of a leak or rupture of a pipeline or vessel, or to design a controlled pressure release system, a method of predicting the mass efflux is required. Although a large body of literature exists on blowdown simulation for vessels or short pipes, relatively little analysis and experimentation has been performed for the depressurization behavior of pipelines. In addition, the impact of the thermodynamics of multicomponent hydrocarbon fluids has been relatively unexplored. A mechanistic homogeneous equilibrium model applicable to both vessels and pipelines has been written. Since fluid thermodynamics are evaluated external to the model, single or multicomponent fluids can be simulated as well as single or multiphase flows. In addition, a series of small scale blowdown experiments has been made using both gas bottles and coiled tUbing. The coiled tubing has a length to diameter ratio of SO,OOO, comparable to that encountered in field piping. Air, carbon dioxide, and carbonated water have been used in the experiments.Excellent agreement between the model and experiments has been obtained, along with excellent agreement with at least one field data set.

In SPE 26565 presented at the 68th Annual Technical Conference in October, 1993, analysis and experimental results were presented for the blowdown of pipelines and vessels using air, carbon dioxide, and carbonated water. Since then further experiments have been carried out using several hydrocarbon gases including both methane and heavier mixes. The pronounced difference in the blowdown behavior between pipelines and vessels noted in the earlier non-hydrocarbon experiments was confirmed for the hydrocarbon gases tested. In addition, for vessel blowdowns, the presence of condensed hydrocarbon liquid was shown to increase the mass efflux rate over an otherwise equivalent single-phase methane blowdown. By contrast, pipeline efflux rates were found to be virtually insensitive to gas composition over the range of hydrocarbon gases tested. These experimental results were compared to the analytical methods described in the earlier paper, and initial efflux rates were found to agree with measured values within about 20%. In addition, the measured efflux rates can serve as test cases for comparison to other analytical methods.