9 replies to this topic

[avm58]

Hello, I have tried to simulate AARP (ammonia absorption refrigiration plant) using hysys or other systems .
I tried Aspen One, but it seems very complex to simulate the whole plant. So I used Hysys.


I fit VLE data using Wilson model, but there are low accuracy in some cases, that leads to increasing COP value(0.6) in comparison with real values ( 0.45- 0.5). So I would like to use own flash model in hysys or use expressions which well approximate Y- X values or use tabular K ( Y/X) values. But I dont know how it can be realized ? . I know that hysys is "open" system , I have good programming skills, but need the example of this kind of hysys extention. Please help me.

Thank in advance
Аlexander

[Padmakar Katre]

Dear,
Could you please tell us your understanding behind using Wilson thermo model for this system instead of any Equation of state. Addition to that if you can just upload the system sketch with other inputs, we can comment/help you further.

[avm58]

Dear,
Could you please tell us your understanding behind using Wilson thermo model for this system instead of any Equation of state. Addition to that if you can just upload the system sketch with other inputs, we can comment/help you further.

Hello.
Ammonia-water is very nonideal polar mixture, so any EOS presented in Hysysgives bad result in prediction boiling temperature.It is known that the better way in this case - use complexelectrolyte model, but Wilson, NRTL оr Uniquac also may be used.
I simulate gaseous phase as ideal(It is possible, because pressure is not high ~ 18 bar).

Initial data for the AARP simulation:

Condition at the bottom of regenerator :

T= 118C, P=18.2 bar , NH3 - 0.334 weight %

Condition of absorbtion :

T= 35C ( Supercooling 1.5С) , P=2.82 bar , NH3 - 0.42 weight %

Real coefficient of performance is0.5 (Value may be found in literature ;calculated without simulating system).

Hysys gives ~0.6.

[Chellani]

Hi,
You can try following
1. Change BIPs of Fluid Package. I couldn't understand from your explanation that whether you've already tried this or not. If not, you can do so as you already have experimental data.
2. Use Tabular Properties. Go to Tabular tab of Fluid Package window and enable tabular properties. You can provide K values in form of curves
3. Use COM Thermo / write you own extension and register it. I don't have a ready-made extension which I can share but it would be really time consuming for you as you don't only have to think about flash which seems to be your only problem.

[Padmakar Katre]

Dear,
Could you please tell us your understanding behind using Wilson thermo model for this system instead of any Equation of state. Addition to that if you can just upload the system sketch with other inputs, we can comment/help you further.

Hello.
Ammonia-water is very nonideal polar mixture, so any EOS presented in Hysysgives bad result in prediction boiling temperature.It is known that the better way in this case - use complexelectrolyte model, but Wilson, NRTL оr Uniquac also may be used.
I simulate gaseous phase as ideal(It is possible, because pressure is not high ~ 18 bar).

Initial data for the AARP simulation:

Condition at the bottom of regenerator :

T= 118C, P=18.2 bar , NH3 - 0.334 weight %

Condition of absorbtion :

T= 35C ( Supercooling 1.5С) , P=2.82 bar , NH3 - 0.42 weight %

Real coefficient of performance is0.5 (Value may be found in literature ;calculated without simulating system).

Hysys gives ~0.6.

Dear,
As far as Hysys is concerned, even PR or SRK will help you out to deal with Ammonia-Water System, since I used it in past projects. You can try with Sour option of PR/SRK i.e. Sour-PR and Sour-SRK. Let is know these simulation results after validations with the real data.

[avm58]

Dear,
Could you please tell us your understanding behind using Wilson thermo model for this system instead of any Equation of state. Addition to that if you can just upload the system sketch with other inputs, we can comment/help you further.

Hello.
Ammonia-water is very nonideal polar mixture, so any EOS presented in Hysysgives bad result in prediction boiling temperature.It is known that the better way in this case - use complexelectrolyte model, but Wilson, NRTL оr Uniquac also may be used.
I simulate gaseous phase as ideal(It is possible, because pressure is not high ~ 18 bar).

Initial data for the AARP simulation:

Condition at the bottom of regenerator :

T= 118C, P=18.2 bar , NH3 - 0.334 weight %

Condition of absorbtion :

T= 35C ( Supercooling 1.5С) , P=2.82 bar , NH3 - 0.42 weight %

Real coefficient of performance is0.5 (Value may be found in literature ;calculated without simulating system).

Hysys gives ~0.6.

Dear,
As far as Hysys is concerned, even PR or SRK will help you out to deal with Ammonia-Water System, since I used it in past projects. You can try with Sour option of PR/SRK i.e. Sour-PR and Sour-SRK. Let is know these simulation results after validations with the real data.

Hello,
Unfortulately, it is bad idea to use PR or SRK to simulate ammonia -water system. These models were created for prediction VLE in hydrocarbon systems and then were modificated for sour components (H2S, CO2, etc) -PR -sour and SRK-sour . See below difference in boliing temprature :

P=18.2 bar X Nh3 = 0.37 mol/mol T boil experimental = 114.8C
Calculated :
PR T=122.3C, PR-sour T=130C, SRK-Sour T=131.3C , Wilson ( with my BIP) T= 113.8C

As you can see using PR or SRK leads to bad result. The error is ~ 15C - it is not possible to simulate AARP ! Using Wilson gives better results, but COP ( coefficient of performance) value is bigger than real.The old method (manual calculation on the base of T-H graphs) gives COP close to real value ~ 0. 5

[breizh]

Hi,
A paper about Nh3/water system.

Breizh

[Padmakar Katre]

Dear

I did simulations for the NH3 scrubbing/absorption by water in past with PR and found the results matching with the operational data, even in sour water stripper I used PR. If you share your simulation file, I may check and see how I can help you in this regard.

[Qasim Hussain]

Respected Sir
I am Student of Associate engineer in Mechanical Technology. Recently I started study on absorption refrigeration system using ammonia-water and hydrogen. This system has no moving part i-e pump or compressor. I want to calculate heat transfer in each section of refrigeration system generator, condenser, evaporator, absorber and also the working and flow calculation of bubble pump used in the system.
Currently I have problem to calculate the heat produce during absorption of ammonia in water. I required formula for this calculation and produce heat quantity during absorption of unit mass of ammonia in water at different temperatures (40 t0 45 c) and pressure (10 to 20 bar).
Please provide me chart, parameters related to my calculation. And also the working and flow calculation of bubble pump used in the system.
I will be thank full to you for this kind.
Qasim Hussain

Respected Members, i am student of associate engineer in mechanical technology. Cruntly

[PaoloPemi]

you need a package able to fit both VLE and enthalpy data for ammonia-water,
as suggested by Padmakar S Katre you may select some variant of Peng Robinson,
I use Prode Properties which includes a extended version of Peng Robinson model with parameters calculated to fit VLE, Enthalpy and density data of pure fluids, these are generally quite accurate,
your simulator may incude some other variants,
for VLE you need to enter proper BIPs,
Prode Properties library gives -0.43 and -0.136 (Kij, Lij) for extended Peng Robinson (see your simulator for details)
For water-ammonia mixture you may consider also the procedure included in Refprop which is based on ASHRAE values.

Edited by PaoloPemi, 09 July 2012 - 02:44 AM.