Wall Vessel Temperature after BDV Gas Depressurization 1

[Max1976]

Dear All,

I need to determine the Wall Vessel Temperature after Gas depressurization trough an orifice ( API 521 ) and to verify the necessity of applying Low Temperature Carbon Steel. I know the possibility of using Hysys ( ASPEN Tech. )Depressurization Utility, however I have the following questions:

Assuming adiabatic process, for given initial conditions ( Gas Temperature, Pressure and composition), Vessel Volume and Final Pressure, will be the Final Temperature a function of the depressurization time ( or orifice size )? Is there an analytical relation to estimate the depressurization temperature? Any rule of thumb?

Many thanks and best regards,

Max

 

[davefitz]

I assume the adiabatic surface is at the OD of the vessel.

The vessel temp at the end of the process will be the same regardless of the rate of depressurization, but the vessl wall thermal stress may be lower if the rate of depressurization is slowed down.

The rough time constant for the vessel metal wall is about s^2/alpha, s= wall thickness, alpha= thermal diffusivity. If the rate of change of fluid temp F/hr is less than 100 F/3/t, where t is the above vessel time constant, then the vessel thermal stress should remain below .75* yield, but you can also prove that with ansys or other finite element program. Other seconday effects can gratly change this estimate depending on inside film heat transfer coeffcient or use of internal insulation- If you need to depressurize it instantaneously and the vessel has a thick wall, then an internal liner may be needed.

 

[Max1976]

In order to apply the suggested criterion, how can I quickly estimate the rate of change of the temperature?

 

[dcasto]

This a no win study. With gases, the JT coeficient for HC's is 7 degree/100 psi. An instantaneous blowdown from 1000 psig is a 70 Degree F drop, so, if the vessel starts at 50, thats -20. But even at 2000 psi start, the final temperature will be -90 F and nothing will happen to the vessel, by the time the temperature gets to -20, the stress is 50% and mass of the steel is so high that the -20F of the gas will not draw enough heat from the vessel to get the whole vessel below -20F.

If you really want a worse case scenerio, try liquid ethane at 2000 psig and having a blowdown valve (or a rupture). The ethane will boil off chilling everything to -120F, all carbon steel. The stress is so low on the vessel and lines by the time it gets to -20 F at 160 psi, the system is ok. Now, I would not beat on the vessel, but its fine.

On blowdowns, we did limit per a company policy, blowdowns to under 25,000 pounds mass per hour.

 

[davefitz]

well, there are worse cases available. On power boilers, the steam drum is typically 7" thick, operating at 66 F and 2800 psig. Depressurizing quickly , with the concomittnet drop insaturation pressures and boiling heat transfer coeficient, can overstress drum components.

 

[sailoday28]

If there is no inflow (to the upstream of the orifice volume)to the vessel the change in internal energy , U,in the upstream volume can be approximated as equal to the inflow of heat from the pipe walls. U will decrease as a function of the product of mass leaving and the specific enthalpy of the upstream volume.
dU = Q - mh
How can the Joule-Thompson effect apply?
Typically, with instantaneous blow down, the walls tend to give up little heat.
And with the adiabatic process that follows, pressure and temperature of the gas can be obtained by following isentropes on a TS, PH or HS diagram.

 

[davefitz]

typo error- not 66 F, but 665 F initial temp, can be depressurized to ambient press, and resulting sat. temp at ambient press (212F)

 

[MortenA]

dcasto

While the gas temp downstream a valve or orifice expands following approx. en isenthalpic process inside the vessel its more like an isentropic process - the temperature is thus even lower inside the vessel at the end of blowdown.

Best regards

Morten

 

[FredExergy]

Thanks MortenA for correctly distinguishing the thermodynamic changes undergone by the fluid left in the vessel and the fluid expanding through the nozzle.
The fluid in the core of a vessel can - with a large pressure drop - eventually fall to a very low value (Isentropic or Joule expansion). The fluid passing through the nozzle undergoes isenthalpic or Joule-Thomson expansion. The inlet temperature to the nozzle will also fall with time due to the first effect. The fluid in the vessel nearer to the walls will fall much less than the core due to heat transfer from the walls. The extent to which the wall and central fluids mix or exchange heat depends on the vessel internals and how much turbulence is generated. British Gas did experimental work on this in connection with Rough Field used for gas storage, Haque and Richardson and Saville wrote some interesting papers on this in the early 1990s.
Peter Kauders wrote a good practical article on the subject in The Chemical Engineer around 1980.