10 replies to this topic

[ChemEng_Novice]

 

I'm trying to figure out, on a depressurisation line on a subsea pipeline. The question I'm trying to answer is, if there is a hydrate formed closed to the base of the riser, would a 2" manual globe valve be sufficient to allow sufficient throttling control to depressure the section in say 1-2 bar increments. 

 

I've taken about 250 meters of 14" pipe. giving a volume of about 19.4 m3.

 

The pressure between the hydrate and this throttling valve is 3350 psi, (231bar), downstream is open to flare, so atmospheric

 

To simplify the question: Would a 2" manual globe valve provide sufficient control to vent pressure in a 19.4m3 vessel at 3350psi to the flare in 1-2 bar increments. or, even at the virtually closed position), the globe valve would be too large to allow any throttling control and the trapped pressurewould deplete too quickly.

 

Used a simple stream of 99.6% methane and the rest c2-c5

 

I'm not sure what is the equivalent flow area available from a globe valve. I've simply punted 35 mm to be the equivalent orifice size of a fully open globe valve, and further assume minimum opening throttling would lead to a 5% opening of the available area - not sure how valid this is. I'm then entering short durations of 10-30 seconds in the Depressure utility to see what the final pressure turns out to be.

 

I've not run the depressurise utility for the longest time and not even sure if I've set it up properly- I'm also uploading the Aspen Hysys 7.3 HSC and XML file if anyone wants to take a look.

 

Would appreciate if

 

1) Someone could guide me on equivalent orifice diameter for a 2" globe valve - is assuming 35 mm a good enough assumption

2) globe valves are used extensively on throttling service - not sure if te 5% assumption is too optimistic, can someone challenge that even a turn or couple of turns would open the valve much more?.

3) Any comments on the attached hysys model would be very much appreciated.

4) Any alternate, simpler method to answer the main question.

 

Would appreciate any help.

 

Regards.

RS

 

 

Attached Files

•  compar.XML   753.11KB   236 downloads

[Bobby Strain]

I believe the depressuring utility was abandoned many versions ago. You must use dynamic simulation now. Good luck.

 

Bobby

[ankur2061]

Bobby,

 

HYSYS depressuring-dynamics utility is still the part of the standard Aspen HYSYS module in v7.3 which I use.

 

Regards,

Ankur.

[ChemEng_Novice]

Indeed its called the Dynamic Depressure utility now.

[thorium90]

The Depressuring utility still exists in V8.0. To answer the OP, is it necessary to create a model in Hysys just to find out if a globe valve can control the pressure in a pipeline? If the pressure drops too fast, just radio the operator to be more careful. I dont think this is a modelling problem. All you need is a pressure gauge near the valve (that is in clear view of the valve) and a technican with strong arms...

 

The fact is, without knowing stuff like valve hysteresis, stiction, deadband, etc etc, this model is not going to be able to predict what will actually happen when the operator turns the handle on the globe valve by some arbitrary "5%" or "1/4 turn" etc...

This would all depend on how good your operators are, modelling wont help here. ..

 

If you really need such fine control, a good control valve and sizing will be needed. You cant expect an operator to operate a valve with 231bar and be able to drop the pressure at 1bar increments..... Big arms wont help here...

 

Or perhaps what you need is a tubing connected to the main line, probably 1/4" with a needle valve. That will be much more controllable... 2" globe valve is impossible to finely control.

 

For those who cant open the file, I've attached some screens for your view

Attached Files

•  example depressure results.jpg   114.74KB   174 downloads
•  connections.jpg   99.3KB   84 downloads
•  heat flux.jpg   86.59KB   85 downloads
•  valve.jpg   91.07KB   86 downloads
•  option.jpg   102.94KB   67 downloads
•  op con.jpg   92.55KB   64 downloads

Edited by thorium90, 18 February 2013 - 08:53 AM.

[gegio1960]

Be very carefull with Depressuring utility.

In my experience a "real" dynamic simulation shall be made to "approach" the reality.

Various papers have been published on the non-realistic behavior of the Depressuring utility.

Good luck!

[ChemEng_Novice]

Thanks everyone for your valuable feedback.

 

The thrust of the question is - a 2" globe valve, when "just cracked open", whats the Cv avaialble for flow at this position. I'm not having much luck finding any representative globe valve characteristics.

 

This is part of a review where the review team has raised the point around controlled pipeline depressurisation. The bigger picture is a 10.5 km pipeline which has had a fully formed hydrate plug form right at the base of the riser, so, on the downstream side of the hydrate, you have circa 200m of pipe holding high pressure gas, and the upstream side has 10.3 kilometres. In this situation, company practice is to go for double sided depressurisation i.e, drop one bar  on the downstream side via this 2" proposed globe valve,  close the globe valve, then drop 1 bar pressure on the upstream (the upstream side is depressurised through the wellhead via a umbilical core back to the platform)., repeat the process until the pressure on both sides reaches hydrate dissociation point and it begins to melt.

 

Given the upstream side has orders or magnitude more inventory, it could be depressuirised in 1 bar increments quite easily, and infact this process is expected to take days to bring the pressure down to out of hydrate region. However, on the downstream side, with only 200m or so of inventory, fine depressurisation via the 2" globe valve is a concern in that depending on valve characteristic, it could be that even if the operator "just cracks" the valve open, the resultant flow area would deplete the pressure by many bar in a matter of seconds, leading to a big dP across the hydrate, turning it into a projectile.

 

My assumption of a 35mm equivalent orifice in a 2" globe valve, and minimum opening of 5% of that cross sectional area taking about 15 seconds to drop pressure by 1 bar wasn't entertained very much by the review team. They want to see some real valve characteristic showing indeed the assumption is valid.

Edited by ChemEng_Novice, 18 February 2013 - 11:51 AM.

[thorium90]

Ok, after reading your description. I can safely conclude,

 

You cant have a human operator control 231bar with a 2" globe valve. It will almost certainly become a projectile.

 

The amount of force exerted on the valve at that pressure is very large. It will take quite abit of force just to turn it. And when it does releases abit of gas, the operator will have to go against that force to close it back. All that opening and closing will likely have to be done in a fraction of a second. With the typical human response, you would likely have dropped quite alot of pressure.

 

I once had a 1/4 inch line fitting blow off a 150bar vessel containing roughly 2000m3 of gas. It dropped 10bar in just 2 mins (granted this was fully open). You have 19.4m3 at 231bar, to be controlled by a 2" globe valve, thats gonna be quite a feat. Even if cracked open, it will likely drop very fast, nowhere near 1 bar increments

 

You cant find Cv for crack open because I doubt anyone publishes that. Normally its like 50% or fully open Cv.

 

If you have to do this procedure often enough, manually, there is no guarantee for repeatability. Some weaker operator might not be up to the task. This as you describes poses some safety issues.

 

Have you tried turning an actual 2" globe valve before? I can tell you they dont turn like the water taps at home.

 

You would find difficulty in modelling it with real valve characteristics. Firstly, because its a manual valve, you cant precisely define how many % its open, that why those calculations for hysteresis is for control valves as you can actually type a % output to the valve. Second, the idea is just proposed right? You dont actually have the valve on site to try out right?

 

Perhaps you can try smaller sizes?

Edited by thorium90, 18 February 2013 - 12:33 PM.

[ChemEng_Novice]

Thank you very much everyone for your input, specially Thorium. Taking your advice, I'm convinced as that the 2" globe valve, taking 4-5 turns to fully open would not do the trick. So, looking at most likely a needle valve type arrangement. The problem is, for this extremely onerous case of 20 m3 at 3350 to be depressured in 1 bar increments in say, 1.5 minute (to give operators time to react etc), a tiny flowrate of just about 440 lb/hr is needed (an orifice of circa 1/16th of an inch!!) . However, for normal depressurisation (i.e we've confirmed no hydrate and the entire pipeline can be taken down without the 1 bar increment requirement, the needle valve would need to allow flowarates of the order of 26,478 lb/hr to give reasonable depressurisation times. This is managed by a 0.5 inch RO downstream the manual valve. The needle valve would need a turn down of about 60 (=26478/440) in order to not become a bottle neck. Is it feasible to get high pressure gas service needle valves that can handle such a major swing in flowrates?.

[Erwin APRIANDI]

Hi ChemEng_Novice,

 

I got the same problem here, the solution for me is to use a manual choke valve and an RO downstream, rather than having a needle valve, yes needle valve can be use but I'm not sure about restricting a flowrate using a needle valve at such high differential pressure accross the valve.

[haivsp]

Dears,

I am trying to do depressurizing for fire case. My question is:

- For fire case, I defined that the liquid inventory for equipment is taken up to LAH level of the equipment (22.2 m3 in my example). Why Hysys often assumes that the initial liquid volume is 50% of inventory volume? How can I change this parameter?

 

Please see image bellow for more detail.

 test_Page_01-up.jpg   161.54KB   32 downloads

 

 

Thank you & best regards.