12 replies to this topic

[doodek]

Dear All,
I'm looking for a proper fluid package for refrigerants like R134a, R245fa. Which package (available in Hysys) should be used to ensure correct calculations? In my model (Organic Rankine Cycle) these fluids are in two phases: gaseous and liquid. I use NRTL, but I'm not sure if it's correct selection.

I would be grateful for your suggestions.

Best regards,
doodek

Edited by doodek, 04 July 2011 - 02:45 AM.

[PaoloPemi]

I am not familiar with your simulator and I don't know if for these components there are specific parameters based on Virial Equations (say BWR) or something like Refprop which produces very accurate results. Unless there are specific motivations (say multicomponent mixtures with strong polar chemicals) I would not adopt a model like NRTL to describe liquid phase activity, you may investigate if your simulator manages R134a as a pure component, in that case the state (gas or liquid) depends from temperature being above or below saturation temperature calculated by a specific correlation, for pure components VLE most simulators adopt regular model, for density, enthalpy, entropy etc. see the above suggestions or perhaps you may use a generic equation of state.

[abhi_agrawa]

Doodek,

Data for both the fluids is available at NIST. I would suggest that you get the data from there and validate the thermodynamic model. My feeling is that BWRS should give good results, if the parameters are available. So do verify the accuracy of HYSYS predictions.

Hope this helps,
Abhishek

[abhi_agrawa]

Doodek,

Since you are trying to model a refrigeration system, try to get a good match on the following, if available in NIST Database:

- Latent Heat
- Heat Capacity
- Saturation Temperature and Pressure
- Vapor density
- Liquid density

If you can get a reasonable match for the above properties, you can assume that you have a decent match on other properties also.

hope this helps,
Abhishek

[Technical Bard]

Since these refrigerants are essentially non-polar, PR or SRK should suffice, but as mentioned by the other posters, check the results against NIST or DIPPR data to verify.

[doodek]

Thanks for all your answers. They are very precious for me.
I have found an information about PR and SRK that these models are great for light gases and refrigerants (it confirms Technical Bards opinion). I'm going to verify that.
But in the same publication the author writes: "PR and SRK cannot handle two phase water". PR in Hysys supports the widest range of operating conditions and the greatest variety of systems. My question is: if the PR can be used to model water-steam cycle in power plant? Or I have to use NBS steam tables?

Regards,
doodek

[PaoloPemi]

in most cases you need to solve VLE and enthalpy, for pure fluids (meaning only one component in a stream) most simulators do use the DIPPR (or equivalent) corrrelation for saturation pressure, this is accurate but of course NBS (or IAPWS) tables are better.
For mixtures things are more complex as fugacities or activities are calculated according EOS or equivalent methods.
For enthalpy if you select a equation of state (PR , SRK etc.) the simulator starts from ideal gas (or liquid) state enthalpy calculated with correlations (DIPPR or equivalent) then calculates the departure (at real conditions) with the selected EOS, again for water NBS (or IAPWS) tables are more accurate.
Hoping this helps (a bit)

[Technical Bard]

Thanks for all your answers. They are very precious for me.
I have found an information about PR and SRK that these models are great for light gases and refrigerants (it confirms Technical Bards opinion). I'm going to verify that.
But in the same publication the author writes: "PR and SRK cannot handle two phase water". PR in Hysys supports the widest range of operating conditions and the greatest variety of systems. My question is: if the PR can be used to model water-steam cycle in power plant? Or I have to use NBS steam tables?

Regards,
doodek

I would not use PR or SRK for water alone - just use the steam tables. If you have water with a hydrocarbon, PR or SRK predict the VLE fairly well, but the enthalpies of the water can be flaky. Using Lee-Kesler enthalpy methods will reduce the error.

[doodek]

Hello again,
Thanks for all information. I hope that is my last question: I have checked various data for water and refrigerants and... all looks good apart from enthalpy. For example: water, temperature 15 C, phase fraction 1 (vapour). Hysys predicts enthalpy -2.413e05 kJ/kgmole, according to NIST: enthalpy 43 kJ/kmol. What's going on? Maybe standard state convencion isn't the same?
Thank you in advance.

doodek

Edited by doodek, 13 July 2011 - 08:21 AM.

[pavanayi]

Doodek,

On your question regarding enthalpies, Are you concerned with Enthalpy values (absolute) or ΔH? What is the reference enthaply for Hysys and what is that for NIST?

To clear the confusion, try defining two states and calculate ΔH using both (Hysys values and NIST values). That might clear the doubt.

There is a very useful article related to misconceptions in applying thermodynamics with simulation packages that was uploaded by another member in a similar post somewhere earlier. You may find it in search.

Edited by pavanayi, 14 July 2011 - 09:59 AM.

[Technical Bard]

You are correct. The NIST data and HYSYS use a different reference datum. You can only compare delta H.

Hello again,
Thanks for all information. I hope that is my last question: I have checked various data for water and refrigerants and... all looks good apart from enthalpy. For example: water, temperature 15 C, phase fraction 1 (vapour). Hysys predicts enthalpy -2.413e05 kJ/kgmole, according to NIST: enthalpy 43 kJ/kmol. What's going on? Maybe standard state convencion isn't the same?
Thank you in advance.

doodek

[doodek]

Ok, thank you so much.

[hshethna]

There is new property package in HYSYS called RefPROP. You should use that. That is meant for refrigerants. It is available in HYSYS V7.2 and V7.3