Depressuring systems 3

[supermax]

Hello, I have a gas depressuring system in wich a pressure vessel has to be depressurized from about 29 bar to environmental pressure, using a pressure relief valve.
This is the question and I wonder if you could help me: do you know a mathematical model that is able to tell me what is the minimum temperature value of the pressure vessel wall? I mean: if the gas reaches a minimum temperature value of -50°C at the end of the depressuring process, may I know what is the minimum wall temperature value?
This point is very important for me, because the wall material choice in the vessel depends on this minimum temperature value.

Thank you

 

[gkandy]

I think you can estimate the temperature by Joule Thomson co relation. The gas temperature and the wall temperature for all practical purposes can be considered to be the same

 

[SooCS]

As far as I know, there is not a very good way to estimate the wall temp of the vessel.

What I did I assumed the vessel temp is maintained with the fluid temp as it depressurizes. Depending on the vessel sizes and the wall thickness of steel, heat content in the vessel wall is quite significantly more as compare with the fluid. As such the final temp may not be as low as you assumed only the fluid depressurizes.

Just courious, how can you depressurize the vessel with a relief valve to atm pressure?

 

[supermax]

Thank you for your answers; the final pressure wasn't the atmospheric pressure, but this is a value that I posted for istance. I think the problems depends on:

1) The vessel mass

2) The vessel wall thickness

3) The thermal conductivity and the specific heat of the wall material

3) The depressuring time from the beginning to the final pressure value (and so, the time needed to reach the final temperature value!)

If I have to consider the process as it is, this is a not stationary heat transfer process and it is necessary to write the equations concerning this case. I only tried to find a semplified way.

 

[svv]

hi supermax,
I am very sure you can depressurise a system in Hysys provided you know the time you want to depressurise it in.
If you go into Hysys/utlities, you can obtain the depressurisation utility.
Provided you know your metal mass, lqd inventory, timeyou want to depresursie it, inventory mass and inventory vessel heat capacity , initial pressre and temperature and what pressure you want to depressurise it, you can do it in hysys.
hope this helps
svv

 

[supermax]

Thank you svv, but I have to tell you that I actually use Hysys! I know the feature you were talking about and I used it. The problem is that I can't understand the applied method: the heat transfer problem that Hysys solves refers to an external fire case (the mathematical model used by Hysys seems concern to a fire case)? I read something like that in the API RP 521 statement...

P.S.: Please, forgive me for my terrible english!

 

[joerd]

In HYSYS you can pick (I think) one of 3 methods: adiabatic (no heat input), with specified heat input (e.g. by an external heat source such as a steam coil), or a fire case. The fire case method uses the formulas from API RP521 to calculate the heat input based on wetted vessel wall area.

You guessed right about the important parameters in this problem. All of these can (and should) be input into the HYSYS utility somewhere.
HYSYS will perform a sequential flash in a series of time steps to achieve a "pseudo-dynamic" result. Heat transfer from wall to gas is calculated based on the temperature difference (flux = heat transfer coefficient * temp. difference), the wall temperature drops according to heat loss = mass * Cp * (Tinitial - Tfinal), the gas temperature drops due to depressuring/flashing.

Take care if you use HYSYS 2.4 or lower, I found a bug in unit conversion, when you input the heat transfer coefficient to the wall (or the heat flux, I don't remember) it doesn't convert to SI properly, so you should check your results.

HYSYS 3.0 has a better depressuring utility, based on a real dynamic approach.

You can check support.aspentech.com for more details.

Regards,

Joerd

 

[Guidoo]

We use PRO/II for these depressurizing simulations. These simulations assume that there is no heat input from the surroundings into the vessel (conservative approach).

Emergency depressurization is usually applied in case of an external fire, to prevent the metal temperature to reach a level at which stress rupture could occur. Note that a safety relief valve will not prevent this from happening! This is indeed described in API RP 521, section 3.19. Probably this is also what Hysys is referring to.

With respect to the material selection, note that the low temperatures (e.g. -50 °C) only occur in combination with low pressures. Assumption here is that the vessel will not be repressurized before it has heated up again (operating manual!). Because of this fact, it may still be possible to use carbon steel instead of stainless, without having the risk of brittle fractures. Please ask a material specialist for details.

 

[SooCS]

Heat input to from the supporting is not critical unless a the blowdown system is design with the time delay in activation. In which case the pressure can built and effect the blowtime duration.

When designing a blowdown system, the key issue is to blowdown duration. API has indicated for a 1" wall thickness vessel, a 15 min blowdown duration. Many designers has taken this as the design blowdown time duration. If you look at Fig 1 in API you will see that for a 1" wall thickness plate it take about 15 min to reach 1200 deg F which is about critical temp when the material start to deform. For a 1/2" wall thickness vessel, looking at Fig 1 again, it only take 7 minute to heat up to 1200 deg F. So if you design the blowdown system to depressurize within 15 min for all cases, you will probably have BLEEVE before the system complete it blowing down.

So, set the blowdown duration depending on the wall thickness of the vessel. For a system blowdown watch out for the piping wall thickness. Piping wall thickness is usually much less than that of vessel wall. I have not seem many piping system more than 1" wall thickness unless for very high pressure system.

Also the blowdown system should include the delay in activation of the blowdown, if any. Some companies set a time delay of about 5 min to allow the operators to confirm a fire situation before starting to blowdown.

 

[supermax]

I use HYSYS 2.2 for my symulations, but I'm not licenced to operate in the dynamic mode; however, Hysys is a very expensive sym software! Do you know another sym software less expensive than Hysys? Thank you.

 

[JKEngineer]

I skimmed this thread and am a little confused by the fact that little mention, except in the case of external fire, is made of heat transfer from the environment to the vessel walls. The internal temperature of -50C may be the worst case for the vessel wall, but it does not seem to me to be a realistic approximation. As the internal gas cools by expansion and takes heat from the internal wall surface, the external wall surface should be getting heat from the surroundings. If they are "ambient" then that is something of the order of 20C. Conservatively, I would think you need to treat the outside heat transfer coefficient as though it were still air, while the internal one may be higher, since the contents are in motion as the vessel depressurizes. The metal temperature will then be some value between the ambient outside and the internal gas T. This would not necessarily be the final internal gas T, since the internal gas T is dropping through the process. If the vessel is insulated (not mentioned above) then that would need to be taken into account and would lower the wall temperature.

Am I missing something? Jack M. Kleinfeld, P.E. Kleinfeld Technical Services, Inc.
Infrared Thermography, Finite Element Analysis, Process Engineering

http://www.KleinfeldTechnical.com

 

[MortenA]

supermax:

A inexpensive simulation tool is design II/winsim:

http://www.winsim.com

Best Regards

Morten