22 replies to this topic

[shan]

Hi Everybody,

 

If a natural gas stream (90% C1) is letdown from 1300 psig, 120 F to 20 psig, the letdown temperature is 33 F per Hysys.  If the same stream is blowdown from 1300 psig, 120 F to 20 psig, the final blowdown temperature is -130 F per Hysys blowdown utilities.  Please help me to explain why the letdown temperature and blowdown temperature are so much different although the initial conditions and final pressure are identical.

[PaoloPemi]

the first value (33 F) seems reasonable for a pressure reducer solved as H-P flash operation with Prode Properties,

the second value (-130 F) seems the result of a S-P flash operation (Prode Properties gives similar values for a C1 0.9 plus C3) , is that the case ?

Edited by PaoloPemi, 07 May 2013 - 11:35 AM.

[shan]

The blowdown mode is defined as "adiabatic".  I think it is the same expansion as letdown valve dH=0. 

Letdown is a stable simulation and blowdwon is a dynamic simulation (process conditions are varied with time).  Should stable/dynamic simulations make that much temperature difference to the same final pressure?  If yes, why?

[PaoloPemi]

there is something unclear in your problem,

are you comparing the final condition in a pseudo-dynamic (integration over time) simulation against the result of a single (steady state) operation ?

In the first case you should have defined additional specifications as vessel data etc. and probably final temperature will be different (different specifications).

Differently it may be that blowdown temperature is calculated with a single operation (see above post)

[shan]

I have varied the vessel sizes, blowdown time, and PV Work Term Contribution.  There is no difference on the final temperature. 

[latexman]

Sounds like a dynamic, batch-wise blowdown to me.

[Dacs]

Yeah looks like it.

On the letdown case, you have a constant condition upstream but in blowdown, your upstream conditions vary as you proceed with blowdown.

[Propacket]

Shan,

 

This is common. As Dacs said, there is a constant condition at the inlet (1300 psig/120 degF) in let down case. Howver, blow down being dynamic in nature has two types of expansion processes in series.

• In vessel/equipment which is being depressured and is losing its contents.
• Across BDV which is further cooling the vessel/equipment contents.

So two expansion processes in series should yield a colder temperature.

Edited by Propacket, 08 May 2013 - 06:03 AM.

[shan]

Thank you Propacket.  I accept your two expansion processes explanation as my question's answer.

[PaoloPemi]

>I have varied the vessel sizes, blowdown time, and PV Work Term Contribution. 

>There is no difference on the final temperature

 

that could be true for a insulated system (i.e. for fluid dH = 0) 

however in real world the fluid exchanges heat with equipments (vessels, piping etc.)  and  (within certain limits) the procedure should calculate a different temperature profile if you define a different blowdown time.

[shan]

I have defined no heat transfer at all to simplify the problem.

[PaoloPemi]

then you could model the blowdown as a  isentropic expansion (see above),

just for curiosity sake, have you compared the result  against a S-P flash operation ?

[shan]

There is no S-P depressuring option in Hysys.  Per Mollier Chart of C1, if 100% C1 is isentropic expanded from 1300 psig, 120 F to 20 psig, the temperature is about -240F and 100% liquid.

[PaoloPemi]

I do not know your test composition so I consider a mixture of C1 0.9 and nC4 0.1 with PR (std.) model ,

Pin  1200 psig, Tin 120 F Pout 20 psig

H-P operation returns 35.2 F

S-P operation returns -146.4 F

 

not too far from the values given in your first post.

Edited by PaoloPemi, 08 May 2013 - 07:52 AM.

[shan]

The final temperature is -164F, 0,9129 vapor mole frication for isentropic letdown from 1300 psig, 120 F to 20 psig per Hysys.

[PaoloPemi]

so with your test case there is a 13% error in calculated dT  (-130-(-164))/(120-(-130)) for blowdown and isentropic expansion,

perhaps the cause could be some correction factor introduced in blowdown as Work Term Contribution etc.

ideally if you set dH = 0 the two values should be close.

[Dacs]

* In the vessel the temperature starts to decrease (due to the combined effect of JT across the depressuring valve and also due to the expansions of the contectns inside the vessel),

I just want to make a comment on this.

 

I think the temperature decrease inside the vessel has nothing to do with JT effect on the valve. The vessel undergoes adiabatic expansion and by that alone where the temperature decrease depends on.

 

JT effect only matters on the temperature at the depressurizing valve outlet.

[marchem]

>there are two kinds of tempertaure values HYSYS reports under the depressuring dynamics.
>1 - The minimum temperature value and
>2 - The final temeprature value.
>The two are different and the minimum temperature value will be lower than the final >temeprature value for depressuring with no heat input

 

I agree with Paolo, this point is not clear to me,

could elaborate the concept ?
according my knowledge  (I studied these matters many years ago and may be I do not remember correctly) modeling the depressurization as adiabatic process (dH = 0 meaning no heat exchanged with vessel, no additional work due to friction etc.) the minimum temperature should be the final temperature (minimum pressure inside the vessel),
for this particular case (dH = 0, see above) the final (and minimum) temperature simulating the depressurization cycle should not be very diffent from the temperature calculated as isentropic expansion from P1-T1 -> P2,
consider the last mole of gas in vessel, for an adiabatic process the final temperature (isentropic expansion going from P1-T1 to P2) will be the same no matter of time or number of (ideal) pressure reduction stages etc.
things may be a bit different if there are changes of phase but the concept remains the same.
Of course things can be VERY different when you consider the real world (i.e. dH > 0 , heat exchanged with vessel,  work due to friction etc. etc.) and that is the reason why we need to simulate the whole cycle.
I do this with a different software but I presume results should be quite comparable.

Edited by marchem, 09 May 2013 - 04:13 AM.

[Dacs]

depressurization as adiabatic process (dH = 0) the minimum temperature should be the final temperature (minimum pressure inside the vessel),

Adiabatic process only tells us that dQ=0 (heat exchange between control volume and surroundings is zero). dH in fact is not zero especially if we're talking about isentropic expansion (dS=0).

 

Not too long ago I posted a thread about depressurizing basics (since I'm developing a utility similar with Hysys in MS Excel) and I was about to match the result of what Hysys generates (for adiabatic without heat loss or metal heat effects) when I set the change in entropy at 0 and in fact the enthalpy of the vessel contents before and after each time increments don't match.

[marchem]

Dacs,

for temperature calculated as isentropic expansion from P1-T1 -> P2, I mean S1=S2 or if you prefer dS=0 ,

I think it is clear reading the post,

are you suggesting that it is not correct ?

If in the above considered conditions (no heat exchanged with vessel, no additional work due to friction etc.) you think that dS is different from 0 could elaborate the concept  ?

Edited by marchem, 09 May 2013 - 04:05 AM.

[Dacs]

My understanding with isentropic process (where reversible adiabatc expansion falls into) is that your change in entropy is zero. Plus the fact that you don't have heat exchange (being adiabatic), so something has to give in your system. And that's where enthalpy changes.

[marchem]

the admin has closed this thread so I stop here,

my understanding is that for the above discussed isentropic expansion (dQ =0, dW=0 etc. etc.) the final temperature should be the same as calculated by a depressuring utility or with a single flash operation solving for constant S (S2=S1) from Pin.Tin to Pfinal,

no matter of time or number of ideal (dW=0) pressure reducing, etc. etc. elements.

perhaps some confusion was introduced by notation dH=0 (delta Heat=0) instead of dQ=0,

however the description of the process is clear as the conclusion.

Edited by marchem, 10 May 2013 - 02:07 AM.