How to evacuate vapor from a pipline which is not full initially. AFT impulse 3

[bsteel]

I'm using AFT impulse, and i have pipline with no flow initially (valve closed). I have notification that vapor exists in this pipeline in steady state and so i cannot move to the transient analysis.what can i do?
I tried to use vacuum breaker valve baut it doesn't work!
Sincerly,

 

[LittleInch]

Is this a simplified profile?

what is the end point?

You need to draw this out in terms of distance vs height and then add on the (water?) flow and pressures.

Is the fluid just flowing out faster than it is coming in at your vapour point?

Too many questions, too few answers



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

 

[pennpiper]

What is the line size?
Do you have the ability to pig this line?

Sometimes its possible to do all the right things and still get bad results

 

[georgeverghese]

The steady state may be rightly predicting vapor breakout at this high point due to the high static head presented by the final discharge point.

This vapor is not an air trap in the system - at that point, the liquid pressure is below saturation vapor pressure at the local operating temperature in the pipeline.

If a throttle valve were installed at the discharge end of this line and you gradually increased its differential press (by decreasing throttle valve Cv), this vapor breakout may disappear ?

 

[bsteel]

Hi,
Thank you so much for your ansewrs.
LittleInch: Actually i'm trying to resolve the problem at steady state "NO FLOW IN THE LINE: valve closed". Unfortunatly , there is vapor at this point. The end point is a Tank.
pennpiper: the line is 36 inch , P1 is 2321 meters, the difference between the high point and the low one is 11 meters.
georgeverghese: actually it is an air trap (i guess..) besause there is no flow in the line! i'm trying to resolve the problem at steady state (no flow) because i guess it has an effect on water hammer..(after openning the valve) what do you think?

 

[LittleInch]

We still don't have much data here, but it sounds like you are pulling a vacuum due to static head being greater than the vapour pressure of the liquid, hence the fluid just flows out of the line into the tank until it equalises.

A more comprehensive elevation sketch of the system including the tank would help a lot.

Is 2321 metres the elevation above sea level?

Yes, if you have an air or vacuum bubble and you turn the flow on, the collapse of that bubble can lead to very high surge pressures

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

 

[pennpiper]

bsteel,
Good, you answered one of my questions, line size 36".

Now, what about my second question? Can this line be pigged?

If it can be pigged, then you can use a Foam batching Pig and water or product to push all the air out of the high points.

Sometimes its possible to do all the right things and still get bad results

 

[LittleInch]

pennpiper.

pushing the air is only one of his or her problems. If there is nothing creating sufficient back pressure either flowing or stationary, the fluid will simply pull a partial vacuum and continue to drain out of the line.

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

 

[bsteel]

LittleInch: 2321 meters is the lengh of the pipe P1 only.(between valves).. Yeah that's why the software refuses to model the transient event with cavitation model before resolving this problem in the steady state.

pennpiper : i'm sorry..this line is not pigged because it's just a flush line. The main pipeline is washed with water then a valve is opened to lead this water (with this line) to special tanks. The highest point in this flush line is on the pipe P1 (where the vapor exists before having water flow).

 

[bsteel]

If this bubbles exist now, i guess they will exist during water flow too. I want to evaluate the loads generated by water hammer (and amplified by these bubbles) and see their effects on the pipeline by Caesar Software..

 

[bsteel]

I want to clarify that i have a real problem of water hammer in this line. This is just a little part of the real problem..

 

[georgeverghese]

You may have created this vapor bubble during the steady state operation (before you stopped the flow).

The collapse of this vapor bubble (on resuming flow) as the bubble goes down this pipe (and internal pressure builds up ) is one reason commonly cited as one of the root causes for liquid hammer

 

[georgeverghese]

Even after you get rid of this vapor bubble by throttling this valve at the end of the line, you still may have a liquid hammer problem if you close the valve at the base of this pipeline at elevation 11metres too quickly.

Liquid hammer only becomes an issue if the hammer pressure exceeds mechanical design pressure or if piping components (anchors) cannot handle the resulting generated forces.

 

[georgeverghese]

Is this some kind of temporary line ?

If this line is a permanent line, and if flows vary all the time, you may need to check if one valve setting keeps the pressure at el 2321 at above sat vap pressure for all flow cases.

If one manual valve Cv setting doesnt, then a PIC - PCV loop may be required. PT at el 2321 and PCV at el 11m; stroke speed high limit should kick in at below 30% of design case Cv or so.

 

[gerhardl]

bsteel : 'I tried to use vacuum breaker valve baut it doesn't work!'

The obvious solution seems to be a vacuum breaker (in combination with an air inlet valve), as you have already tested. Why does it not work and in wich way does it fail?

Both air inlet and outlet valve has to have an area sufficient to equal outlet and inlet of air in volume equal to the respective liquid flow in and out of the pipeline roughly calculated at 'normal' air (gas) speed, say for calculation purposes 20-40 m/s (giving a good measure with spare volume).

 

[bsteel]

Thank you georgeverghese for you're ansewers!! Actually it didn't exceed the design pressure but the supports are not able to handle the generated forces as you said!.. umm i don't know how to modelize the PIC - PCV loop on the software..

gerhardl: because on AFT, vacuum breaker valve acts as a dead end during steady state! so it doesn't work...can you explain more how could i calculate the inlet and outlet area please?Thank you!

 

[georgeverghese]

Pls ask a process engineer to help you with the study on flow variation to find out if you need a PIC - PCV.
As far as your modelling is concerned, this PCV can remain as a block valve with different Cv for different flows.
Also ask the process controls engineer to set up the additional logic for the PIC as suggested - one more detail - this high limit on PCV stroke is only applicable to closing stroke ( not applicable to opening stroke)

This is a very high pressure line - static pressure at el 11m is >227barg (3400psig) at no flow ( not including pump dead head pressure) !! And line size is 36inch !

What is the design pressure you are using for this line, and which piping code ?

 

[gerhardl]

Bsteel:

By AFT impulse I believe you are refering to

http://www.aft.com/products/impulse.

With the 11 m hight difference, can I assume your piping class to be EN PN16 (16 bar) or somthing equal in PSI?

Back to the original question: Your problem seems to be both practical and theoretical.

Your practical problem is that you have a long, large pipeline filled with fluid with gas (air) on the top point(s?), which (as expected) will give flow problems and waterhammer problems when you start the flow.

Your theoretical problem is that you can not get the application of software (composed of emphirical data and theoretical data supposed to be able to give correct overview of dynamic situations?) to suit your case.

Advice:

1. Forget for the moment the theoretical part.

2. As others already have pointed out: You must in some way control the flow of fluid in the pipeline, especially by the outlet point if you let it out downhill to a lower (atmospheric?) pressure in a (open?) tank. The tank inlet is drawn without a valve. In my opinion the best solution would be to have a control valve here.

Only special valves can do this without cavitating. In some cases (for some valves) this will require special solutions with additional air indraft for the valve, especially if there is no liquid after the valve. You will also have to protect the pipeline for underpresure with a suitable size vaccuum breakers at highpoints, combined with air outlet valves. For this length of pipeline you will need this equipment at several points, disregarding solutions else.

3. Your system seems to be an open-end, unpressurized system, partly filled with fluid, with gas at highpoints. Each time you start your system, you have the same problem as filling a long municipal water main, which must be filled with care, giving time letting the air escape from all highpoints.

4. A preliminary suggestion (if feasible) is to go for the solution pointed out in #2: let the whole pipeline stand pressurized (filled with liquid) with suitable air inlet and outlet valves, between flushings, with a control valve at tank inlet.

5. If it is not possible to have the line permanently pressurized between use, then there is necessary with a careful, very slow start of inlet to let air escape (eg. controlled outlet) and large enough air inlets for larger flow after filling, to avoid vacuum and waterhammer for the downhill pipeline. It might also be (have to be checked against flow data and total end layout) a possible solution with an on/off valve at top-point, and pressurized pipeline before this (with air outlets), combined with a controlled fluid inlet amount, and a large air inlet for the P1down to tank, and only an on/off valve at the tank.

6. As others have pointed out (again): to give further advice will require a full set of details and a considerable amount of consulting engineering work.(Choice of control valves at outlet is interesting. Needle valves (German: ringstempelventile, and probably in lesser dimension than the pipe) double-excentric ball valves or cone valves could possibly be considered etc. etc.)

7. Very rough calculations of air inlet and outlet is very simple, as stated before. Just compare inlet/outlet of water (cubic), to same amount of cubic air at an assumed normal gas speed, forgetting all fancy calculations taking gas weight and comprimation into consideration.

 

[georgeverghese]

Okay, if elevation difference is 11metres, and line length is 2321metres.
For a 36inch line, use a butterfly valve at the base for throttling(for either fixed setting or automated PIC - PCV loop arrangement)

 

[gerhardl]

georgeverghese:

Be careful!!

Your butterfly valve will easily cavitate at regulating openings, even at 11 bar or less pressure difference, if the required opening is small under the whole or part of the regulating sequence, especially if the regulating is out to atmospheric pressure.

At openings less than say 5 to 15% roughly, differing from exact inner constructio and flow amount, a BFL will anyway give poor and unpredictible regulation, disregarding pressure difference.

Conditions must be checked and calculated for the exact valve and conditions!

(BFL is not impossible, but will with high probabillity give problems.)