Gas Blending in a Pipe

[KernOily]

Hi guys. Hope you are all well in this new year. Question. Client wants to blend three gas streams for use as fuel in a steam generator. The streams are: 1000 Btu/scf pipeline utility gas, 600 Btu/scf casing gas, and 1300 Btu/scf raw gas from a well. I maintain that I can insert all three of these streams into a pipe and as long as I provide sufficient turbulence, I will obtain complete mixing of these gases in a few diameters, due to turbulence and Brownian motion. I might throw in a static mixer for fun. Client says no, you need a big vessel to do the blending.

Have you guys ever seen a correlation or rule of thumb for blending gases to achieve complete mixing in X pipe diameters downstream of the blend point?

Thanks guys! Pete

 

[georgeverghese]

At flow turndown, without some mixing device, mixing of these streams will be much poorer. A static mixer will help to ensure mixing over a wider turndown, as will a series of cut baffles ( like we see on the shellside of shell and tube heat exchangers). Obviously, the static mixer can only be used in clean serive, while the cut baffles would tolerate some fouling. A more expensive option would be to use a throttle control valve ( any of ball or butterfly or globe style trims will do ) that operates on constant diff pressure on this combined stream.

 

[zdas04]

Pete,
It all depends on Reynolds Numbers (both velocity and density have control). The 600 BTU casing gas is probably about 20% CO2 and has a SG around 1.3? The raw gas is probably around 0.8 SG? The commodity gas is 0.6 SG. So you have a pretty good range of densities. If the velocities are very different then you have a serious mixing problem and probably need a static mixer. On the other hand if your Reynolds Numbers are below 3,000 or above 6,000 and within about 5% of each other you may not need the mixer. It all depends on relative velocities and densities. If all three gases were put into a vessel at separator velocity then they would surely blend quickly, but just bringing three pipes together can easily get velocity stratification that can take a while to work itself out.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[BigInch]

Or nitrogen.

Brownian motion isn't going to mix anything in a couple of diameters. Maybe it would work in a thermos bottle over night. Turbulence is what you want. A vessel is proably the way to go, so that you can limit high velocities to short distances within maintainable equipment. It would avoid higher pressure drops and well stream particle erosion from affecting a long pipeline.


Reaction to change doesn't stop it

 

[LittleInch]

Like zdas04 says - keep it in the turbulent zone, add a mixer and also depends on how you add the gases.

If you look at things like injection of chemicals, using a centre quill or injection point would be best rather than a tee or branch connection.

Then maybe 40 - 80 D to get complete mixing.

If you stuff it through a static mixer or a control valve then maybe 10D downstream of that.

Wholly dependant on velocity, pressures, quantities, line size, SG.

Yu would be waiting an eternity for Brownian motion though, especially with that level of difference of SG

I have seen some fancy transient modeling for different liquids (condensate into Crude Oil. CFD modeling could be used.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bimr]

Specialised equipment is seldom needed for mixing gases, which because of their low visocsities mix easily. The mixing given by turbulent flow in a length of pipe is usually sufficient for most purposes. Turbulence promoters, such as orifices or baffes, can be used to increase the rate of mixing.

Tank mixing is usually more expensive, but has the significant advantage of maintaining an accurate proportion down to zero flow.

One issue is that an expensive control setup will be required:

http://www.controlglobal.com/articles/2010/controlvalves1003/

Another issue is that the gas to the boiler should have a consistent Btu.

 

[KernOily]

Guys - Thanks for the great replies. Much appreciated. We've decided to use both the static mixer and the big vessel. The vessel will go 80D downstream of the mixer. In addition to blending, the vessel will also provide surge capacity to account for the lag in response time of the control system (the three streams are throttled to maintain a Btu setpoint of the blended stream). The control strategy is identical to the one in the link from bimr.

Thanks again guys. Pete

 

[BigInch]

Great. I think you made a good choice.

Reaction to change doesn't stop it

 

[KernOily]

Thanks brother Bigs.