Sure, the CFD can solve this mixture/mixing problem. Just make sure you provide appropriate initial/boundary conditions.

Thanks all for valuable comments. Here, I would like to add some more details to better clarify the problem.

The distant source is a set of centrifugal compressors of limited capacity. The system requires intermittent consumption of at much higher flow rate. 14” piping from distant source is connected to an equal “Tee” in such a way that its one branch joins to buffer vessels and other branch to the system. Buffer vessels supply the excess flow during short time higher consumption by the system reducing the pressure up to some extent. During non-consumption time pressure again gets built up in upstream of the system.

To me, it looks that flow rate from buffer vessels (only 20 m away from system) may be higher than the piping from distant source because of two reasons:

1) Flow from a large quantity of gas stored nearby, and

2) Direction of flow is straight to the system at “Tee” joint. (Flow from distant source has to turn 90o, where both flow combines).

But how to quantify is the problem. The information available are trends of upstream pressure and flow rate to the system.

From the information provided, I understand that the system needs gas from short duration and two sources are used one continuous supply from the compressor and other from a buffer pressurised tank source. Normally for a steady state flow systems, a 1D flow models are used and they a pretty good. This is transient flow situation, I suggest that you can connect compressor line to the buffer tank and from there you can take it to your system. The analysis of this is simpler and easier and probably you can analyse using 1D flow models. The analysis using 1D models is very fast.

Thanks Mr. Rao for suggestion, but I can not change the configuration. Can any method other than CFD be used for this problem?

From what I can see, if you have two sources with different pressures at the T junction, the flow from higher pressure source flows to the system first till the pressure is equalised at T junction, and later the flow takes place from both the sources almost equally.

You only have to check that the flow requirement of your system for specific duration ( in terms of mass flow is available to you from both the sources. There may not be any need for detailed analysis if you are only looking at you can complete the tests with the configuration and sources you have.

N Nanjunda Rao

I think that 1-D models are not appropriate for this problem. You must use 3-D at least for simulate flow in T-connection.

I agree with Tishkin that 3D flow simulation analysis should result in more accurate prediction and is the best if you have time to do it.

To: Dr. Tishkin - Could you highlight why 1D models are not appropriate for this problem?

Becouse in T-connection 3-D effects determine what will be the fraction from each source.

Thanks Mr. Rao and Mr. Tishkin for your valuable comments.

Mr. Rao, I never stated that there is “two sources with different pressures at the T junction”. In fact when system is not consuming gas, the pressure in piping from distant source as well as buffer vessels are almost in equilibrium because the continuous supply of gas from compressors will slowly gets filled up in buffer vessels. So, during non consumption time pressure in upstream piping and buffer vessels increase slowly. During short time consumption of gas by the system pressure falls rapidly. Approximate constant supply rate from distant source is 4 Nm3/sec (0.13 m3/sec) and short time consumption by the system is 30 Nm3/sec (1 m3/sec). Hence during consumption, system draws gas from long distance piping coming from compressors as well as nearby buffer vessels. Total volume of piping is approximately 200 m3 (about half of the buffer vessels). I feel that during short consumption period long piping will experience more pressure drop compared to nearby buffer vessels. This may result instantaneous lower pressure at ‘T’ joint in comparison to buffer vessels and allow more flow from buffer vessels. But how to quantify it, how much more?

Pritam. Since you had not indicated the pressures of 2 sources, I had to start from T junction which is valid even the source pressures are equal. As you have correctly pointed out the pressure drops in pipelines are different if the flows are equal from two sources. In reality, what happens is that the pressure drops gets adjusted so that the pressure at the T junction becomes same. You should therefore find out the flow from each source to get same pressure drop for the actual piping configurations by iterations. Since one source is continuous and other source is depleting in pressure(buffer vessel) it may be combersome to estimate the transident situation.

Thanks Mr. Rao for your valuable input, pertcularly for the suggestion to "find out the flow from each source to get same pressure drop for the actual piping configurations by iterations."