15 replies to this topic

[Steve90]

Hi !

 

While simulating a refrigeration loop on HYSYS to decrease the temperature of natural gas from -5°C to -20°C using propane as refrigerator at -20°C , the propane leaving the chiller become colder and have a -21°C ! is it normal ?

[Bobby Strain]

You tell us nothing about the exchanger model.

 

Bobby

[curious_cat]

Hi !

 

While simulating a refrigeration loop on HYSYS to decrease the temperature of natural gas from -5°C to -20°C using propane as refrigerator at -20°C , the propane leaving the chiller become colder and have a -21°C ! is it normal ?

 

Flow rates? Pressures? 

[Steve90]

Ok ! Sorry it's a shell and tube heat exchanger :

 

Tube side :(Natural gas)

Temperature inlet -5°C / outlet -19.28°C

Pressure 30 barg

molar flow : 801 kgmol/hr

 

Shell side :(Propane)

Temperature inlet -20°C / outlet -21°C

Pressure 1,428 barg

molar flow : 80,6 kgmol/hr

 

Thanks

[gegio1960]

pressure drop? property package?

[Steve90]

0.1 bar for each side ! peng robinson

[gegio1960]

everything seems reasonable... it's difficult to answer without the simulation and the simulator ;-)

it could be an effect of the heat balance with the changing state and the pressure drop....

my only suggestion is to vary a little bit the parameters and to observe the variations

sorry

 

ps: you can also analyse the heating/cooling curves of the HE to localize strange points, if any

Edited by gegio1960, 29 April 2015 - 11:35 AM.

[Zauberberg]

It also depends which parameters are specified and which ones are left to be calculated by the simulator. If Propane is specified as saturated vapor at the preset pressure at the exchanger outlet (or compressor inlet), temperature is calculated according to the P-T diagram for Propane.

[serra]

Steve90,

in real plants the evaporator should superheat the refrigerant after all the liquid has evaporated,

I use a different software (Prode Properties) which allows to solve these operations in Excel and that allows to verify heat balance for each component, in this case evaporator (Hgas=Hpropane),

possibly you can follow a similar procedure, for natural gas Peng Robinson should do the work while for  C3 a Mollier chart (or electronic tables) may help.

[curious_cat]

The only way this is thermodynamically possible seems to me if propane phase changes. Fundamentally, there doesn't seem anything obviously paradoxical in what you are seeing. Or at least I don't see the paradox. 

 

What gives? Why do you suspect your answer? 

Edited by curious_cat, 30 April 2015 - 02:23 AM.

[serra]

since for C3 saturation pressure at  -20C is about 2.439 Bar.a calculated with Prode but values are similar to those reported by Steve90 and given the low dP assuming it is a flooded evaporator,

to me it seems strange that for C3 (cold side) outlet temperature is lower than inlet temperature but it may depend from some wrong settings as suggested by others...

Edited by serra, 30 April 2015 - 02:43 AM.

[shan]

Lower C3 temperature is because the pressure drop.  C3 bubble point (-20C, 1.442 barg), and (-21.16 C, 1.342 bag) per Hysys Peng-Robinson.  Please try your simulation if you specify no pressure drop, there will be no temperature drop and if you specify pressure increase, there will be temperature increase under the condition of specified outlet stream vapor fraction <= 1.00

Edited by shan, 30 April 2015 - 07:14 AM.

[serra]

of course (my values a bit different -20 C 2.4388 Bar.a -21.15C 2.3388 Bar.a see above, but similar dTsat vs. dPsat)

but, as previously noted, there is no superheat, may be changing some settings will improve solution.

[shan]

There is only C3 latent heat involved in the heat exchanger. 

[serra]

latent heat is predominant and probably sufficient to solve heat balance in many cases

but I suppose you can instruct the software to heat the vapor above saturation temperature to simulate a refrigeration process,

that could have a noticeable impact on resulting cycle and remove the negative dT noticed by Steve90...

Also the way how latent heat for C3 is calculated can introduce additional errors.

[varun_harikumar]

@steve - It'l be true if you are talking about liquid propane flashing.

 

Instead of cooling to a very low temp, propane is just going a degree lower. Rest of the energy(latent) is used to cool the hot fluid - NG.

 

Similar phenomenon would have been difficult had we not considered flashing.