Liquid gravity flow - problems with pulsations?

[AndersE]

Hi,

I am tring to determine if there will be problems with air entrainment in the pipe and/or pulsations for the following problem:

I'm working with a project on an offshore installation (Norwegian sector). The storage tank for wax inhibitor is filled by connecting a flexible hose from the transportation pods in which the wax inhibitor is shipped. The company wishes to replace the hose with a permanent pipe.

The elevation difference between the pods and the storage tank is about 23 m and the line will be approx. 70 m in length. The line will have sections that are either sloped (minimum 1:100) or fully vertical. Of course, since elevation difference < pipe length most of the pipe will be sloped rather than vertical.

The storage tank is a vented atmospheric tank with four 2 inch vents. The inlet to the tank is not submerged, so air can enter the pipe via the tank.

The pipe is supposed to be 2 inch SCH 10 and has a maximum allowed design pressure of 18 barg (according to company piping spec).

The density of the wax inhibitor is 884 kg/m3 and the viscosity is around 10 cP.

I have calculated the maximum possible flow from the static height and a pressure drop spreadsheet (Darcy formula). The total length incl all bends etc have been taken from a preliminary piping isomtric. This gives me a velocity of 2.79 m/s and a flow of 23.6 m3/h.

However, this assumes a full pipe. There have been some discussions earlier on this forum regarding problems like this, for example thread378-81608. The thread descibes how vertical gravity lines can "gurgle" when air passes through the pipe.

My questions are: Will the pipe be liquid filled? If not, will air entrainment in the pipe cause serious problems with vibrations/pulsations? Is the "gurgling" a problem or will it just mean that it takes a while to establish full flow? Could one install a throttling valve at the end of the pipe to decrease the flow and the pulsation problems?


Some safety considerations (in case someone asks...):

The volumetric flow is to small to overpressurize the tank even if one vent is blocked and the tank is liquid filled.

The pods are equipped with a vacuum relief valve.

The wax inhibitor is not flammable according to the MSDS.

The storage tank is protected from overfilling with a actuated ball valve (closing time is 8 s to prevent hammer effects - enough?). The storage tank is also fitted with an overflow line.

The static pressure from the liquid cannot overpressurize the pipe.

For a picture of one of these pods, see

http://www.conserveplc.com/content/patent01.htm

 

[hyposmurf]

I would fit a tee at the pipe top part at the start of it just after the pod valve , and a branch to be higher than the liquid level of the pod , so the air will not enter from the down tank, it will take from the atmosphere, and mean while the pod is empty the vacuum release valves will allow air enter to the pod .

 

[BigInch]

Why not just put a check valve on the line just before it enters into the storage tank?

Its not clear what the elevation is entering the storage tank, so I can't be sure if a check would also help prevent possible contamination of the pod by liquids backflowing from the tank (moved by back/up flowing air from the storage tank,.. as they might if there was no check valve), but a check would tend to prevent that possible contamination source.

http://virtualpipeline.spaces.msn.com

 

[katmar]

Anders,

I agree with you "full pipe" flow calc rates. At the other end of the scale if you go for "self venting" flow your flow rate will be about 0,3 litre/second. The time difference for unloading 680 gallons (assumed imperial i.e. 3.1 m3) is huge. A full pipe scenario will discharge in 8 minutes. If you go for self venting you are looking at almost 3 hours. Somewhere between these two extremes you will get the slugging and gurgling that you mentioned.

I think that it was in the thread that you referenced that I stated that the important factor to consider is the resistance to flow from your pod to the fixed pipe. If that can deliver 24 m3/h then you are OK. I would go for the full pipe option in preference to self venting flow. It seems to me that your safety measures have been well thought out.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[MortenA]

The Froude number should be sufficiently low - this will ensure that "gurgling" should not occur.

You can check out this NORSOK standard:

http://www.standard.no/pronorm-3/data/f/0/13/26/4_10704_0/P-001e5.pdf

The Fround no is calculated this way:

F=V/(sqrt(D*g))

V: Vel (m/s)
D: ID (m)
g: gravity constant (9.81 m/sec^2)

According to the standard the Froude no should be less than 0.3

Best regards

Morten

 

[AndersE]

Thank you all for your help, it is always valuable to get some extra input from other engineers.

MortenA:
Your consideration is very interesting, but I have interpreted the reference from NORSOK P-001 as being only valid for vertical pipes that are NOT full.

Chapter 6.2.5 says:
"Pipes that are running full, and do not require a minimum downward slope to avoid particle deposition, shall
be sized according to the total available static pressure head, and the maximum allowable velocities for
liquid lines."

And later in the same chapter (under vertical lines):
"Vertical gravity lines with or without submerged outlets (e.g. so-called “dump caissons”) shall be designed
such that the Froude number is less than 0,3. This is to avoid air entrainment and ensure undisturbed flow
without pulsations."

The way I have interpreted this is:

If you know that the pipe will be running full it shall be sized on available head and maximum allowed velocities.

If you know that the pipe will NOT be running full it shall be sized so that the Froude number becomes lower than 0.3.

In the case I am working with the Froude number is about 3.8. If it has to be lower than 0.3 the flow will have to be throttled to 1.9 m3/h, which would mean that it would take several hours instead of 20 min to refill the storage tank.

But do I know that the pipe will be running full? That is the question...

 

[katmar]

Anders, I agree with your original formulation of the full pipe scenario. If the static head exactly matches the friction drop then the pipe will run full. If the flow is less than this rate the upper interface will be at some level below the inlet of the pipe and there will be at least a portion of the pipe that is not full.

You could check the situation in the sloped sections using Mannings equation to see if the pipe is full for a given slope and flowrate. This should give very much the same answer as the static height vs friction drop analysis. At 24m3/h I think you are virtually certain to have a full pipe.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[MortenA]

Sorry i didnt read far enough

However maybe you could check that you calc. velocity gives a Froude no high enough to ensure full pipe? Im not sure but maybe you could discover it yourself.

Best regards

Morten

 

[AndersE]

MortenA and katmar:

It is probably a valid assumption that the pipe will be full, so I will base the design on that. The Manning equation gives me full flow in the horizontal sections.

 

[BigInch]

I think all this thinking cost more than the check valve.

http://virtualpipeline.spaces.msn.com

 

[sailoday28]

I believe for vertical pipes, full flow and pulsations are related to a critical Grashoff number. That number may also be tied in to the flow in the original question. I do not have time at present to search it down.