×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

Modelling very large pressure pressure drops in liquids
2

Modelling very large pressure pressure drops in liquids

Modelling very large pressure pressure drops in liquids

(OP)
I have a system where there is a near 100Barg pressure drop in a liquid stream over a valve. According to HYSYS this pressure drop results in an almost 1% increase in the absolute temperature of the liquid.

Is this correct? If so, what is the fundamental reason why this happens?

Read the Eng-Tips Site Policies at FAQ731-376  

RE: Modelling very large pressure pressure drops in liquids

This is enough dP on a liquid to change the volume by about 1 %.  I know 3000 psi will reduce water's volume by 1%.  Throttling through a valve is usually isenthalpic.

h = u + pv = 0
du = Cv dT

pv decreased, so T had to increase.

Good luck,
Latexman

RE: Modelling very large pressure pressure drops in liquids

Latexman (Chemical)

du=CvdT-  d/tau [p*tau]dV      where the bracket term derivative is at const volume AND
tau=1/T     T=absolute temp
p= pressure
V= specific vol.
Cv= spec ht at const specific vol
u= spec internal energy.

or
du= CvdT  +  Tdp/dT[@const v * dV   -  p dV

Sorry, I don't know how to type in the partial derivatives.

Regards

RE: Modelling very large pressure pressure drops in liquids

sailoday,

The equations I pulled from memory were strictly for ideal gases and not correct for this case.  Thanks for the correction.

Good luck,
Latexman

RE: Modelling very large pressure pressure drops in liquids

If you express the fundamental equation for enthalpies,

H = U + P*v

Then, dH = dU + p.dv + v.dP

At constant enthalpy, dH = 0.  Also, the p.dv term in this case is very small compared to the v.dP term, since the specific volume change is negligible.  Therefore, we can say that dU approximates -v.dp.  The change in temperature, of course, is proportional to the dU term.

In Hysys, they solve the appropriate equation of state iteratively at the final pressure to obtain an enthalpy equal to the value at the inlet.  Note that it is possible to end up with a two-phase mixture at the outlet condition, depending on how close the incoming liquid enthalpy is to the outlet liquid's enthalpy at the bubble point temperature.  This check is made rigorously in Hysys.

RE: Modelling very large pressure pressure drops in liquids

Then, dH = dU + p.dv + v.dP

Refer to my previous posting on this thread for dU.
While specific vol change is negiligible, what about the temp change?

Yes an appropriate equation of state should be used.

For example

dH=CpdT   d [ (v*Tau)/dtau  at const press ] *dP

Where Cp  is spec ht at const pressure,
v  spec vol
Tau= 1/T
T= abs temp.
Assuming a const enthalpy process, the change in temp can
obtained from an equation of state.
If v=v(P, T} use the above eq.
If p=P(v,T) use my previous posted equation to get dU.

Regards.

Having the change in press

RE: Modelling very large pressure pressure drops in liquids

sailoday28:

Solving the equation of state (EOS) numerically at constant enthalpy is the proper way to get the new temperature after the valve.  The solution method for the Soave or BWR equations of state is found in, e.g., T. F. Daubert: "Chemical Engineering Thermodynamics" (McGraw-Hill, 1985).  The input data required are:  critical temperature and pressure, Pc, acentric factor, and the ideal gas enthalpy coefficients, h0=f(T) for all components in the mixture.

ColourfulFigsnDiags's original question is best answered, I think, by reminding him that maintaining fluid enthalpy after pressure is reduced drastically cannot be done without a temperature increase for most fluids, unless you have the rare case of a negative Joule-Thompson coefficient:

(?T/?P)|H = 1/Cp*[-T*(?V/?T)|p + V] , where the ? operators denote partial derivatives.

The Joule-Thompson coefficient is negative only for real gases above reduced temperatures of about 6, and also for hydrogen and helium gases.  For liquids, it is always negative.

RE: Modelling very large pressure pressure drops in liquids

Correction to my previous post.
dH=CpdT  + d [ (v*Tau)/dtau  at const press ] *dP

or
dH=CpdT   +vdP  - T?v/?T|p dP

or for JT coefficient  that of posting
by UmeshMathur (Chemical)

"In Hysys, they solve the appropriate equation of state iteratively at the final pressure to obtain an enthalpy equal to the value at the inlet.  Note that it is possible to end up with a two-phase mixture at the outlet condition, depending on how close the incoming liquid enthalpy is to the outlet liquid's enthalpy at the bubble point temperature.  This check is made rigorously in Hysys"

Does Hysys check to see if flashing occurs within valve?

Regards





RE: Modelling very large pressure pressure drops in liquids

If the outlet stream from the valve will be a two-phase mixture, Hysys gives you the molar vapor fraction (i.e., vapor moles/feed moles) at outlet T and P.

RE: Modelling very large pressure pressure drops in liquids

UmeshMathur-My real question is== how does Hysys compensate for flashing within a valve?
Regards

RE: Modelling very large pressure pressure drops in liquids

The exact geometry of a valve is extremely irregular, and so an increasing amount of flashing occurs from the point where the local bubble point pressure is reached until the valve exit.  If the fluid is already in a two-phase condition at the entrance to the valve, the percent vaporization simply increases as pressure drops.

I rather doubt that anyone is interested in the progression of flashing within the valve body.  The pertinent engineering question is: "What is the condition at the end of the valve for a specified total pressure drop?".  This calculation requires an isenthalpic flash to be performed at valve outlet T and P.  This provides the final value of the V/F ratio.  To make this calculation, Hysys uses the user-specified thermodynamic option, e.g., an equation of state such as PR, SRK, BWR, etc.

As always, the user remains responsible for the proper choice of thermodynamic specifications from the many options available in Hysys (or any other process simulator).

RE: Modelling very large pressure pressure drops in liquids

(OP)
Fascinating discussion guys, and thank you for the answers!

Read the Eng-Tips Site Policies at FAQ731-376  

RE: Modelling very large pressure pressure drops in liquids

(OP)
Ok, I guess (after reading this thread) it is clear that if the pressure drop across the valve occurs isenthalpicly, then there will (nee. must) be a corresponding temperature rise in the fluid.

The next question is, how valid is the assumption that the pressure drop across the valve is isenthalpic?

I mean, if you were out in the field, with a valve, would the temperature increase be a real thing that you would measure?

Read the Eng-Tips Site Policies at FAQ731-376  

RE: Modelling very large pressure pressure drops in liquids

If there isn't any vaporization of the liquid you will often see an increase in temperature, albeit small.  This is especially the case for water and is sometimes referred to as a "reverse JT effect".  JT systems are typically used to lower temperature by vaporizing a portion of the liquid.  Check your college thermo boook, it should have a discussions of JT systems and times when the JT coefficient is negative (this is your case).

By definition a JT system is isenthalpic and is an accurate (and necessary) assumption here.  Remember your first principles.  Without heat flow, enthalpy must remain constant.

RE: Modelling very large pressure pressure drops in liquids

??By definition a JT system is isenthalpic and is an accurate (and necessary) assumption here.  Remember your first principles.  Without heat flow, enthalpy must remain constant.??
IF AND ONLY IF SPECIFIC VOL CHANGES ARE NEGLIGIBLE--ie change in KE is negligible. Which for liquids is a reasonable approx.   Oh--I forgot to include change in elevation----

regards

RE: Modelling very large pressure pressure drops in liquids

(OP)
Sorry, perhaps I was not clear enough. I understand that any system without heat flow will be isenthalpic, however, I guess what I was asking is if there is any significant transfer of heat via friction within the valve, or via heat transfer / losses to the outside world.

It is now clear to me that most liquids will experience an inverse JT effect (albeit very small), and that HYSYS takes account of this in its calcs (which was my original question). But if we walk away from the cleanliness of theory, and look at the messiness of reality, the pressure inside the valve itself will clearly rapidly change and vary dependent on the valve geometry. In general, will this effect be observed in the plant, or will external losses / effects swamp it?

Read the Eng-Tips Site Policies at FAQ731-376  

RE: Modelling very large pressure pressure drops in liquids

I'm afraid we have some misconceptions regarding the fundamentals.  The following comments are offered in an attempt to clarify the issues.

A negative Joule-Thompson coefficient is observed only rarely - hydrogen and helium gases, and very hot gases (above a reduced temperature of 6 or so).  For those who have forgotten, reduced temperature is the ratio of the system temperature to the critical temperature of the fluid (absolute units).  Thus, most fluids cool down when pressure is reduced isenthalpically.

The degree of flow turbulence and sudden changes in direction of flow across the valve are irrelevant to the issue here.  The enthalpy is preserved, except for heat losses.  These losses are negligible, considering that the total surface available for heat transfer to the outside world is barely a few square feet in most valves and, therefore, quite insufficient to allow significant heat transfer by natural convection to the outside air.  Recall that convective heat transfer coefficients are also very small.

Finally, huge changes in velocity from inlet to outlet can have effects on outlet temperature arising from the kinetic energy term.  Again, these are generally negligible for most real-world valves.

Therefore, the theoretical prediction of outlet temperatures is pretty accurate, as long as you choose the right thermodynamic options in Hysys.  In fact, these changes are quite small for liquids that do not show a phase change.

If, on the other hand, the incoming fluid flashes to any significant extent, the valve outlet gets much cooler, as the latent heat of vaporization is being drawn from the enthalpy of the fluid itself.  That is why, in the refrigeration business, such valves are called "JT valves".

RE: Modelling very large pressure pressure drops in liquids

If the internal characteristics of the valve are not considered, can't flashing within the valve affect the flowrate?
Regards

RE: Modelling very large pressure pressure drops in liquids

Colourfulfigs,

The answer is yes.  But remember that steady state operation is something that is rarely seen in a plant, no matter what people say (depending on your definition of steady state that is).  So you're probably not going to get the model to match to 1%, which is the change that you noted in the simulator.

Without any vaporization the volume isn't going to change appreciably, so there will be a small temperature increase.  Almost every other effect for this system is typically negligible.

But whether the operator will actually see such an effect depends more on the accuracy of their instrumentation than anything.  Such a change is typically not measurable in a plant (in a lab, yes).  It's not really going to be seen in operation unless you start vaporizing something and get a large delta T.

UmeshMathur, you're right on for gases (H2 and He).  We're talking a liquid here, and maybe in reality the JT coefficient term doesn't necessarily apply (I don't really know).  But I do know (dt/dp)h is negative when the temperature increases during a throttling application.

RE: Modelling very large pressure pressure drops in liquids

(?T/?P)|H = 1/Cp*[-T*(?V/?T)|p + V] , where the ? operators denote partial derivatives.

The Joule-Thompson coefficient.........   For liquids, it is always negative.  MAYBE

The maximum density of water is at 4C at atm pressure. Therefore ?V/?T|p  can be negative.

Regards

RE: Modelling very large pressure pressure drops in liquids

The Joule Thompson coefficient, (dt/dp)|h, is all that matters when discussing isenthalpic processes.  For water, and almost all normal fluids, this is always positive so there is cooling as the pressure goes down at constant enthalpy.

If there is flashing across the valve, the valve pressure drop will go up.  That simply increases the total DP to be overcome to maintain the measured flow, assuming there is a pump.  (Of course, the flow is not measurable in the two phase region after the valve - it must be measured upstream while you are in a single phase condition.)  However, we are not discussing the effect of DP on flow here, I believe.

RE: Modelling very large pressure pressure drops in liquids

Umesh, in your reply of April 28 you said "The Joule Thompson coefficient, (dt/dp)|h, is all that matters when discussing isenthalpic processes.  For water, and almost all normal fluids, this is always positive so there is cooling as the pressure goes down at constant enthalpy."

But back in your reply of April 22 you said: "The Joule-Thompson coefficient is negative only for real gases above reduced temperatures of about 6, and also for hydrogen and helium gases.  For liquids, it is always negative."

Which is it for water and other liquids, positive or negative?

RE: Modelling very large pressure pressure drops in liquids

UmeshMathur (Chemical)/DickRussell (Chemical
Please refer to my posting of property of water at atmos press and 4C where water has a maximum density/

At the condition of atmos press
J-T coef (?T/?P)|H = 1/Cp*[-T*(?V/?T)|p + V] becomes
@ 4C   (?T/?P)|H=v/Cp  which is positive.
At temp below 4C  
(?V/?T)|p  will be negative and J-T positive.

Regards

RE: Modelling very large pressure pressure drops in liquids

UmeshMathur:

Your post on 22 Apr 06 19:09 appears to be misleading in stating that "maintaining fluid enthalpy after pressure is reduced drastically cannot be done without a temperature increase for most fluids".   You may be referring to only liquids but "fluid" generally means gases and/or liquids.

As noted in earlier posts most real gases, except for some few exceptions like H2, will cool on expansion through a throttling valve.  

RE: Modelling very large pressure pressure drops in liquids

DickRussell and rbcoulter:

I apologize profusely for the confusion created by the contradiction re. the J-T coefficient in my earlier posts.

Unfortunately, there is a “Joule-Thompson inversion curve” for every substance:  On a P-T diagram, lines of constant enthalpy above this curve have a negative slope, and a positive slope below this curve.  Further, this phenomenon can be exhibited both below and above the critical temperature.  At very low temperatures and above the saturation pressure, the slope becomes consistently negative.

However, for sub-critical temperatures and at pressures below the saturation curve, the slope is positive (“normal” condition).  Thus, T goes down as P is reduced at constant enthalpy for such “normal” vapors.  For compressed liquids above the saturation line, at very low temperatures, the J-T slope is consistently negative.  However, closer to the critical point, it can become positive for liquids and then reverse sign at higher pressures.

A good description of these phenomena can be found in K.E. Bett, J.S. Rowlinson, and G. Saville: “Thermodynamics for Chemical Engineers” (MIT Press, 1975).


To check the points made above, using the SRK EOS, I ran a simple series of isenthalpic flashes for compressed liquid methane, starting each case at 500 kg/cm2 and -125 C before the valve:

END P [kg/cm2] =   450.     400.     350.     300.     250.     200.     150.     100.      50.       40.       30.        20.        10.       5.         2.         1.
END T [C]          = -122.9 –120.8 –118.9 –117.2 –115.6  -114.4  -113.4  –113.0  –113.3  –113.5  –113.8  –114.1  –124.6  –138.0  –152.5  –164.4

It is clear that temperature increases until the pressure drops to about 100 kg/cm2.  Hence, the J-T coefficient is negative up to this pressure.  After this point, however, the temperature begins to decrease, at first slowly and then far more rapidly; the J-T coefficient has clearly switched sign to become positive.  This behavior is entirely consistent with Bett et al’s description.

This pattern of behavior has been exhibited by all hydrocarbons and organic liquids that I have tested.  Hydrogen and helium are the main exceptions (based on the special equations of state on the NIST website).

If anyone has found a case where the description provided above has NOT been observed, it would be extremely interesting to study it.

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources