Behavior of Pressure Gradient and Transient Pressure Signals during Liquid-Liquid Two-Phase Flow

2006 ◽  
Vol 29 (10) ◽  
pp. 1183-1195 ◽  
Author(s):  
D. P. Chakrabarti ◽  
P. Ghoshal ◽  
G. Das
Author(s):  
Maral Taghva ◽  
Lars Damkilde

To protect a pressurized system from overpressure, one of the most established strategies is to install a Pressure Safety Valve (PSV). Therefore, the excess pressure of the system is relieved through a vent pipe when PSV opens. The vent pipe is also called “PSV Outlet Header”. After the process starts, a transient two-phase flow is formed inside the outlet header consisting of high speed pressurized gas interacting with existing static air. The high-speed jet compresses the static air towards the end tail of the pipe until it is discharged to the ambiance and eventually, the steady state is achieved. Here, this transient process is investigated both analytically and numerically using the method of characteristics. Riemann’s solvers and Godunov’s method are utilized to establish the solution. Propagation of shock waves and flow property alterations are clearly demonstrated throughout the simulations. The results show strong shock waves as well as high transient pressure take place inside the outlet header. This is particularly important since it indicates the significance of accounting for shock waves and transient pressure, in contrast to commonly accepted steady state calculations. More precisely, shock waves and transient pressure could lead to failure, if the pipe thickness is chosen only based on conventional steady state calculations.


1966 ◽  
Vol 8 (1) ◽  
pp. 107-109 ◽  
Author(s):  
D. Chisholm

Relationships between the friction pressure gradient and the cross-section of the tube occupied by the liquid are developed for the flow of air-water mixtures in rough-walled horizontal tubes. The data indicate a pronounced change in the form of the relationships when the pressure gradient reaches a value of about 60 lb/ft2ft.


Author(s):  
E T Pak ◽  
J C Lee

Pump performance characteristics change drastically under two-phase flow conditions from those of single-phase flow. This is due to a change in flow characteristics in the impeller. Owing to a positive pressure gradient the air bubble moves more slowly than the water in the impeller channel, but in the suction surface region of the impeller inlet, where a negative pressure gradient prevails, the bubbles move more quickly than the water. Thus, in the space just after this region the distributions of the void fraction obtained are considerably higher and uneven. The change in the pressure distribution owing to air admission is also particularly evident in the inlet region of the impeller. These changes bring about an alteration of the whole flow pattern in the impeller and also cause a drop in pump performance. The Reynolds-averaged Navier-Stokes equations for two-phase flow in a centrifugal pump impeller are solved using a finite volume method to obtain the pressure, velocities and void fraction respectively. Good agreement is achieved when the predicted results are compared with those measured experimentally within the range of bubbly flow conditions.


2018 ◽  
Vol 31 (1) ◽  
pp. 102-116 ◽  
Author(s):  
Sudarja ◽  
Aqli Haq ◽  
Deendarlianto ◽  
Indarto ◽  
Adhika Widyaparaga

2009 ◽  
Vol 283-286 ◽  
pp. 599-603
Author(s):  
Hikmet Ş. Aybar ◽  
Mohsen Sharifpur ◽  
Roozbeh Vaziri

Many researches have performed some studies about pressure gradient in the one phase flow, particle-liquid two-phase flow and porous media. However, there is no any report about interaction among them. In this study, this interaction idea is developed by using the same particle size for particle-liquid two-phase flow and porous media. For the experimental study, an apparatus is designed, and at the first step one phase water pressure gradient is investigated, next in the further steps, little by little spherical particles are added to the cycle till accumulation of the particles did not allow any movement to the particles (i.e. porous media occur), and after that well pack porous media is investigated. The results confirm the relation between pressure gradient over mass flow rate in the one phase flow, particle-liquid two-phase flow and porous media obeys as a parabolic curve.


1973 ◽  
Vol 15 (3) ◽  
pp. 235-236 ◽  
Author(s):  
D. Chisholm

From consideration of the homogeneous equation for frictional pressure gradient, and the relation between friction and void fraction, an equation for the ratio of the phase velocities during two-phase flow, is developed which is interesting in form, easy to use, and agrees with experiment.


Author(s):  
Catalina Posada ◽  
Paulo Waltrich

The present investigation presents a comparative study between two-phase flow models and experimental data. Experimental data was obtained using a 42 m long, 0.05 m ID tube system. The experimental data include conditions for pressures ranging from 1.2 to 2.8 bara, superficial liquid velocities 0.02–0.3 m/s, and superficial gas velocity ranges 0.17–26 m/s. The experimental data was used to evaluate the performance of steady-state empirical and mechanistic models while estimating liquid holdup and pressure gradient under steady-state and oscillatory conditions. The purpose of this analysis is first to evaluate the accuracy of the models predicting the liquid holdup and pressure gradient under steady-state conditions. Then, after evaluating the models under state-steady conditions, the same models are used to predict the same parameters for oscillatory and periodic conditions for similar gas and liquid velocities. The transient multiphase flow simulator OLGA, which has been widely used in the oil and gas industry, was implemented to model one oscillatory case to evaluate the prediction improvement while using a transient instead of a steady-state model to predict oscillatory flows. For the model with best performance for steady-state pressure gradient prediction, the absolute percentage error is 12% for Uls = 0.02 m/s and 5% for Uls = 0.3. For oscillatory conditions, the absolute percentage error is 30% for Uls = 0.02 m/s and 4% for Uls = 0.3. OLGA results underpredict the experimental pressure gradient under oscillatory conditions with errors up to 30%. Therefore, it was possible to conclude that the models can predict the average of the oscillatory data almost as well as for steady-state conditions.


Author(s):  
Rinaldo Antonio de Melo Vieira ◽  
Artur Posenato Garcia

One-dimensional single-phase flow has only one characteristic velocity, which is the area-averaged velocity. On the other hand, one-dimensional two-phase flow has several characteristics velocities, such as center of volume mixture velocity and center of mass mixture velocity. Under slip condition, usually they are quite different. In a simple way, one may think that the petroleum correlations and the drift-flux model are an attempt to “adapt” the single-phase momentum equation for a mixture of more than one phase, where the several parameters in the single-phase equation are replaced by average-mixture ones. These two models use different considerations for this “adaptation”. For instance, for friction loss calculation, petroleum correlations use the mixture volume velocity while drift-flux models use the mixture mass velocity. Normally, the volume velocity is higher than the mass velocity, and petroleum correlations may calculate friction gradients higher than the ones obtained by drift-flux models. This is very important, especially for horizontal and slightly inclined upward flows, where the friction pressure gradient is dominant. This work compares the pressure gradient evaluated by these two models for horizontal and slightly inclined upward flowlines using available data found in literature. The comparison shows that, depending on the situation, one model gives better results than the other. Based on the results, a new approach for two-phase flow friction calculation is proposed. The new model represents a combination of the approach used by the Petroleum Correlations and the Drift-Flux Model, using different characteristic velocities (volume, mass and a new one defined by the authors). The new model is very simple to implement and shows good agreement with the tested data.


Sign in / Sign up

Export Citation Format

Share Document