Two Phase Flow

J

Thread Starter

James Thomas

We have an application where we pressurize a tanker with nitrogen to offload solvent into a storage tank. We want to measure the quantity of solvent being offloaded but are experiencing problems with 2 phase flow i.e solvent gas mixture when the tanker is almost empty. We are using a coriolis flowmeter with local indication only. It is not used for control and is not inputed into a DCS. Has anyone got any suggestions on how we can over come this problem in on order to achieve accurate flow measurements. Thanking you.
 
The flow meter should be insensible to the weight density of the fluid. Check with YOKOGAWA and their Vortex. This device is less sensible.
 
A

Alan Rimmington

Could you use a break tank or similar before the flow meter, eg tank is higher than final process and is vented giving chance for gas to leave liquid before gravitating (or pumped) through flow meter
 
H

Hakan Ozevin

Coriolis or mass flowmeters will not give accurate results if the material is a liquid/gas mixture. I dont know how, but you must find a mechanical way to get rid of that nitrogen before the mass flowmeter. Once I have read about such vessels, but I dont remember the details.
I am sure someone from the list will remember it.

Good luck

Hakan Ozevin
 
C
We have run into similar problems with two-phase flow in mass flow meters. You can add a gas separator prior to the flow meter. Essentially this would be a vessel with a tangential side nozzle for the inlet, a bottom liquid outlet and a top vapor outlet. Be aware that it will not be pure nitrogen coming off the top (i.e. there will be solvent with it). The vapor valve can be a simple float operated "air eliminator" valve or the valve can be controlled via level. This is very difficult to control though because the vessel is usually small, and the gas flow is often erratic. The best thing may be to look for a different flow meter or transfer method.
 
H

Hakan Ozevin

I checked and saw that those mechanical equipments I mentioned are called "gas separators" (simple isn't it:)). You can make a search in the net to find a supplier for those.

Another idea, although I do not know if it fits to your design, is to use vacuum on the secondary side. Then you do not need nitrogen and no bubble problem at all. But of course all the system including the secondary tank should be pressurized (or should I say vacuumized?:)). We are using this method in pharmaceutical industry, i.e. it is not practical to transfer tonnes of liquid.

Hakan Ozevin
 
James,
I can see several possible approaches to this problem. I will work under the assumption that this measurement is being sought for accounting (custody transfer or verification) purposes. I will try to rank the solutions in increasing costs.

1. Have the transport provide certified weights before and after
unloading, calculate the volume and ignore the meter.
2. Do not draw any solvent out of the tank while unloading the
transport, then measure the change in level and forget you have the
inline meter.
3. Install an "air eliminator", which is a small (less than 50 Gallon)
vessel that is in the line before the meter. This provides a place
for the nitrogen to separate from the solvent. A ball type float
will cause a valve to open and vent the vapor space to your vapor
recovery/thermal oxidizer system.
4. In addition to the above, you might want to consider requiring your
solvent supplier to provide a transport with a tractor mounted
pump.
5. If that is not feasible and the number of loads of solvent is
sufficient, you may consider installing an unloading pump. Given
your cost of Nitrogen and the number of loads you receive daily, it
might be more economical to change the way you unload.

Good luck in your adventure.

John Beck
 
M

Miguel Borges

Ok, but the problem with multiphase technology is the high price of these meters( about $140,000). Fluenta as well as others manufacturers offer that
technology.
Best.
Miguel Borges

 
J
Some mass meters do not respond well to entrained gases of any amount in the fluid. I am assuming this is the case here and that the duration of this two phase flow is prolonged.
You may perhaps consider using the concept of multiphase metering but it doesn't have to be as expensive as suggested, we are only talking two phases of clean fluid and gas. You need a volumetric meter and an entrained gas density meter. This is a device which will continue to report the true density even though there may be anything upto 100% entrained gas. Since the gas density is very low compared to the solvent density, and you describe this phase as primarily solvent with some nitrogen carry over, you may simply multiply the volumetric flow by the density.
If, however, there is a progressive increase of the nitrogen content so that this simpel assumption is not reliable then you will also need a pressure sensor so that the flow computer can correct the density signal to find the true solvent concentration in the gas.
The trick is now to chose a volumetric flowmeter which will prevent phase slip and which will not be damaged by or will resist overspeeding. Possibly a triple helix meter (the flow component of at least one multiphase meter) but this system will be far less expensive than a multiphase meter for oilfield service, though more expensive than a simple massmeter. By the way, you don't say what the mass meter people say about the problem.
Entrained gas density meters are not that common. I only know of the Solartron tube density meters at www.solartron.com. But i would recomend a web search for vibrational density transmitters and ask some questions of the different manufacturers you find even if they don't advertise entrained gas capability. Tripple helix meters can be obtained from Kral, to name but one manufacturer.
 
M

M. Ramchandran

We manufacture Centrifugal gas Separators which we have supplied for many Mass Flow applications on LPG, Ammonia & Hydrocarbon fluids. The centrifugal Gas Separator is a vessel which has tangential inlet & outlet and is sized to retain the liquid inside in a centrifugal spin for approx. 5 - 10 secs so that the lighter parts viz. Air, N2 etc can separate out to the centre of the vessel and move upwards. We also manufacture the automatic air-venting system which is a ball float operated air vent which can sense the level of the liquid so that only gas is vented and liquid remains inside the vessel.
Please let me know whether you require any details on these vessels.

M. Ramchandran
[email protected]
 
Top