Offshore Sonic Flare

[EmmanuelTop]

We are about to commission offshore facilities and LNG trains within the next couple of months. At present, I am focused on offshore platforms and verifying critical design parameters and operating manuals, control philosophy etc. There are couple of issues related to the offshore flare system that I'm not much familiar with.

The flare is with a sonic tip, designed to release full blowdown flow @ backpressure of 8 barg. The system design pressure (flare collection header, KO drum, and line to the flare tip) is 15 barg. Maximum relief flow is 49 MMSCFD of gas, and the line size from KO drum to the flare tip is 6". The choked flow pressure for design relief flow is about 1.8 bara (6" line) if I remember well.

The system has been designed with the following philosophy:

- In case of total platform blowdown, there will be a shock wave progressing through the 6" line toward the flare tip (due to choked flow), causing accumulation of gas and pressure buildup. Design relief flow can be achieved only when the flare tip backpressure reaches 8 barg.

- This dynamic event is envisaged to last very short, perhaps a few seconds - when the system will be exposed to sonic velocities. After reaching 8 barg upstream of the flare tip, the flare system will achieve its design relief capacity (49 MMSCFD), and the flow regime will be well below sonic conditions (Mach = 0.3 to 0.5, depending on location in the system and according to API rules).

My concern here is - and perhaps I'm just guessing because I am not familiar with sonic flares - whether the choked flow conditions and sonic velocities through the system (particularly in the 6" line) will last for a sustained/extended period of time, as opposite to what has been calculated and written in the Flare Design Basis. What happens when backpressure reaches 8 barg? The flare starts to transmit full relief flow, but at what conditions and what Mach number upstream of the flare tip? Also, as the blowdown continues and flow rate drops, the backpressure should drop as well. Does this cause flow regime to approach sonic conditions again? I would appreciate if someone experienced with sonic flare systems can explain the sequence of events, from the beginning to the end of blowdown - accompanied by (qualitative of course) description of flow regimes during blowdown.

Thanks in advance,

 

[EmmanuelTop]

No hints on the subject?

http://antwrp.gsfc.nasa.gov/apod/astropix.html

 

[rneill]

Sorry I can't guess at what might be happening in the system nor do I suspect can others since it would all depend on the size of the flare stack, headers, relief rates, and fluid composition.

What you need is a dynamic simulation of your flare system. You might be able to set up a simple calculation that iterates in an excel spreadsheet. Calculate the initial release rate, pressure drop through pipe, relief through the tip - assume this is constant for some small period (perhaps 0.25 sec or less) - then reduce the system pressure by the amount of fluid relieved and recalculate the relief rate, pressure drop, etc. and keep doing this until the system is stabilized.

In general, I think that if you have a large header designed for little pressure drop, you could be sonic in the tip while much less than sonic in the header but if you have a "small" header you might in some circumstances be sonic in both.

There is a little information on "High Pressure Smokeless Flares" in API 537.

 

[Flareman]

Your conditions are simply flow and orifice issues.
The sonic flare is just a final orifice on the pressure relief. The smallest hole in the system will control what happens to the flows and pressures. I have to suppose that the total orifice area through the flare tip is that hole. If so it creates a back pressure in the system which allows all the upstream velocities to be less than that through the final orifice, regardless of the flow rate.

How quickly the flow decays will depend on the entire system volume which you "pump up" to 8 barg. Just for the sake of an estimate, work out the contained mass of the system at 8 barg. Divide by the mass flow rate at the same pressure. That becomes the time constant k for a decay equation. w = Wo * (1 - exp(-t/k)).
w = mass flow rate at time t
Wo = contained mass.
Final time to equilibrium will be 5 to 10 time constants.
When the internal pressure eventually drops below the critical presssure the characteristics will change and lengthen the decay a little.

David