Gas-liquid flow: a prototype

Story of a successfully built prototype project describing the characteristics of two-phase flow under different conditons

20150917_103357 [427862]
The prototype!
Two-phase flow or gas-liquid flow is an important measure to account when we study the behavior of fluids especially in gas/oil fields and wells where lifting is usually done through deep-well jet pump or where retrograde condensation happens or the most common example that we can easily observe is inside of a coffee maker and countless others.

Let us first examine the basic phenomenon and then we shall discuss the prototype we had made. The simplified Bernoulli’s equation suggests that at constant water flow-rate the pressure difference is mainly due to the energy losses but when air (precisely gas) is introduced into the tube the pressure difference decreases due to decrements in the average density of the fluid but when we increase the gas flow-rate a point (may) will reach where the decrements in the average density will be countered by the increase in the friction losses and this will cause the pressure difference to rise again. Through this whole process, we may encounter different flow patterns each corresponds to a different observation.

While making this prototype we had several difficulties, one of which was the non-consistency of the ASTM fittings’ standard among acrylic tube and UPVC and for the remedy, we had to make our own standard even we had made some of the components just because it weren’t available to fit it accordingly, like the inlet portion of the compressor just after the push-in fitting of moisture trap etc. and we did with the help of Machine Shop and Fluid Mechanics Lab in Mechanical Engineering Department. The earlier design included a NRV, which was installed just after the moisture filter but due to severe leakage problems and halting of gauges; we did multiple free flow tests by dis-engaging the NRV and finally decided to remove it permanently. Due to the fluctuations in the gauge reading and lack of finance we weren’t able to get the desired sensitive pressure gauge though we’d installed a globe valve to regulate the flow but in the end agreed on the installation of two valves i.e. one for the drain purpose and the second one for the generation of some back pressure. As this whole working is based on the recycle and somehow a by-pass cum drain; we’d used un-wound hose pipe to recycle the water stream but it dwindled and caused hindrance in the out-flow and in order to rectify this issue we’d removed the hose pipe and replaced it with the UPVC pipe.

2015-09-04 21.34.38 (2)


The demonstrative apparatus consists of two inlets as mentioned above i.e. for the water and air in-flow & two outlets i.e. one serves as a recycle stream and the other one acts for drainage as well as by-pass for the smoothness of flow. Besides some standard fittings it also has some custom pneumatic. We had intended to use the auto-drain moisture trap for its standard purpose but due to some unavoidable circumstances we couldn’t and it now plays the role of an intermediate medium between the water and air which can be better demonstrated and understand rather than explaining here. Precisely, fathom and a foot long acrylic tube (30mm O.D. 26mm I.D.) has been used for the flow visualization rest includes UPVC and couple of pneumatic pipes. The pressure gauge ranges to 10 bar and placed 3.5 foot apart from each other. The 2 HP air-compressor is a single-stage oil-cooled capable of producing 198 L/min at a maximum pressure of 115 psig and has a 24 L tank. It also had come with a built-in regulator apart from the opening valve so we didn’t need to put any further regulation after that. The ½ HP centrifugal pump has the capability to pump about 40 foot head at its standard operating point.

2015-09-04 21.36
The results obtained below are pretty much verifying this graph.


Fortunately, we have been remained successful to get the desired flow-patterns as well as the bulls and bears in pressure drop and somehow able to obtain the hold-up and ratio of gas-to-liquid flow from the graph:


Flow-pattern Pressure (1) (kPa) Pressure (2) (kPa) ∆P


Bubbles 6.89 14.47 7.58 0.724
Plug 9.65 15.16 5.51 0.526
Slug 13.78 18.95 5.17 0.493
Churn 18.61 23.43 4.82 0.461
Annular 26.88 33.08 6.2 0.592
Mist 30.32 38.60 8.28 0.791


All the above readings have been taken at (maximum possible) constant flow-rate but it might vary for the desired flow-pattern. Due to air-water flow the reading at the gauge fluctuates so we have picked the best possible value for explanatory purpose. As we haven’t placed any rotameter so the values mentioned here might not justify with the graph.

* ɣ = 9810 N/m3 & h = 1.067 m

There are a lot of improvements that can be made in this project. For the precise value of the gas-to-liquid ratio in the feed, rota meter should be installed. Also in order to keep the valve handling constant we have to consider controlling the flow-rate by using a variable frequency drive for the pump. These are just recommendations which may be implemented in the near future.


Actual picture of the demonstration showing slug along with bubbles



 p. 344; Smith/Van Ness/Abbott; Introduction to Chemical Engineering Thermodynamics

 p. 60/418; De Nevers; Fluid Mechanics for Chemical Engineers

p. 419; De Nevers; Fluid Mechanics for Chemical Engineers

p. 420; De Nevers; Fluid Mechanics for Chemical Engineers

p. 420; De Nevers; Fluid Mechanics for Chemical Engineers

 During construction:

 20150525_130054 20150525_13130720150518_131407 20150525_130225