[dhinesht88]

Hi all,
I tried using Aspen Plus/Properties to extract the thermal and transport properties of a molten salt mixture (KCl-0.67+MgCl2-0.33 mole fraction) for 450oC to 900oC where the fluid is in the liquid phase. The predicted results from Aspen using different EoS (ELECNRTL, ENRTL-RK, SYSOP0, VANL-2, VANL-HOC, VANL-NTH, VANL-RK, Peng-Robinson, PR-BM etc.) were too far from the literature values for density, viscosity, thermal conductivity and heat capacity at constant pressure. Example: density of the salt mixture while defining the salts as conventional fluids were in the range of 450-850kg/m3 as opposed to 1650kg/m3 per the literature @450oC. Whereas selecting the components as solid gives the density of about 4000kg/m3.
Do I need to specify any binary interaction manually to get the properties right?
Please advice on this.
Thank you.

[Enthalpy]

Welcome, dhinesht88!

The density of 450-850kg/m³ is so much wrong that it must result from a big data error. One example would be if your software has no data at all for the density of molten KCL, takes zero then, so it sums the volumes of both salts but sums only the mass of MgCl₂ and gets for the mixture about 1/3 of the density of KCl.

Warning: I don't know Aspen Plus.

[dhinesht88]

Thank you for your response.
Actually Aspen Plus calculates the properties based on the standard Equation of States (EoS) such as Peng Robinson etc., hence i do not see any problem mishandling of big data. I tried several EoS and i do not see any improvements in the prediction. I am not sure whether it is appropriate to use those EoS for predicting the thermodynamic properties. If its possible, do we need to tune any constants to get proper results..

[silverz89]

Hey Dhinesht88,

Check this paper out if you haven't already: https://www.degruyter.com/downloadpdf/j/zna.1976.31.issue-7/zna-1976-0715/zna-1976-0715.pdf

Discussing using the Tait EoS for molten salts at higher pressures. I'm not exactly sure how much input you can make in Aspen regarding the parameters of the EoS, but this group seemed to achieve a decent precision modeling molten salt densities up to pretty high pressures using their approach. Might get a little trickier when you're modeling a molten salt mixture, but i liked the idea of working with a model that at least has assumed one fixed parameter for this class of fluid and a fairly straightforward temperature relationship for the other parameter.

One thing the paper mentioned is the molten salts tend to have lower isothermal compressibility than other fluids. This might be a potential source of error in the Aspen when modeling as a conventional fluid. 

[Enthalpy]

Did you try to let Aspen Plus predict the data for molten KCl alone, MgCl₂ alone, or 99%+1% mixtures? This will give you hints to what goes wrong.

Oh, and with "big data error" I meant "big error in the data". When you get a density 3× too low, it's not a matter of fine tuning, of interactions, or of model choice. The error is probably simpler.

I just have a kind of intuition that one data is completely missing (for instance the density of molten KCl), and that Aspen Plus takes zero if it doesn't find it.

[silverz89]

If you're still stuck you can maybe try walking us through what variable parameters you input into Aspen to set up the model (EoS, experimentally determined densities @ various states, phase, etc) and what parameters you had options to input but left as default or the software automatically filled in.