Basic sample script to compare isothermal properties of Ammonia

[ffraile]

Hi!
I am trying to prepare a basic Notebook to compare the output of the library with the Isothermal properties of Ammonia, here is the link:

https://colab.research.google.com/drive/1q5_LhMmJ4o5SdNyf0RRFquV4Coq4W4b7?usp=sharing

But the results are not as expected. This is a code snippet that summarizes what I am trying:

from thermopack.saftvrmie import saftvrmie
#Initialize for NH3
eos = saftvrmie('NH3')

enthalpy_vap_saft  = eos.enthalpy(ref_temp, ref_press*1e6, [1], eos.VAPPH, dhdt=True) # Specific enthalpy (J/mol), and optionally differentials
print(enthalpy_vap_saft)

enthalpy_liq_saft  = eos.enthalpy(ref_temp, ref_press*1e6, [1], eos.LIQPH, dhdt=True) # Specific enthalpy (J/mol), and optionally differentials

Can you please help?

[vegardjervell]

It is somewhat unclear from the code snippet here what output you are expecting, but the reference data used by ThermoPack is found here, and appears to be different from the reference used in the source you linked in the notebook.

Any discrepancy in absolute value for the enthalpy between ThermoPack and the other values in your notebook likely come from the difference in reference state.

[ffraile]

Thanks very much for the quick response! Appreciated. I am trying to assess if ThermoPack provides similar values to the source linked in the Notebook, which i believe corresponds to the same source (NIST). The enthalpy values that I get are negative and I cannot make sense of the data. I was expecting to get a similar value for the same reference point. Perhaps this code snippet makes it a little bit clearer:

from thermopack.saftvrmie import saftvrmie
#Initialize for NH3
eos = saftvrmie('NH3')

# Reference point extracted from NIST
ref_temp = 300 #300K
ref_press = 0.001 #0.001MPa
ref_enthalpy = 28.990 #28.990 KJ/mol
ref_phase = eos.VAPPH


enthalpy_vap_saft,  = eos.enthalpy(ref_temp, ref_press*1e6, [1], eos.VAPPH) # Specific enthalpy (J/mol), and optionally differentials
# Value of enthalpy_vap_saft is -45875.93633017853
enthalpy_liq_saft  = eos.enthalpy(ref_temp, ref_press*1e6, [1], eos.LIQPH) # Specific enthalpy (J/mol), and optionally differentials
print(enthalpy_vap_saft)
assert round(enthalpy_vap_saft/1000, 1) == round(ref_enthalpy, 1)

[vegardjervell]

Ok, I understand. The difference you are seeing comes from the NIST webbook calculator using a different reference value / reference state than the recommended value used by ThermoPack. It is admittedly not entirely clear to me where the reference values in the NIST webbook calculator come from.

In any case, unless you are interested in chemical reactions, the reference enthalpy is irrelevant, in which case you can compare values computed using NIST webbook and values from ThermoPack by

• Either comparing computed enthapy differences (rather than absolute values at a point)
• Or translate the ThermoPack values to the other reference

For the first case, when you compare enthalpy differences, the reference state cancels out, such that values computed from ThermoPack should coincide quite closely with the values from NIST webbook, given that you are using an appropriate EoS.

For the second case, you can use set_enthalpy_of_formation

T_ref = 298.15 # Reference temperature used by ThermoPack
p_ref = 1e5 # Reference pressure in Thermopack
H_ref_other = 28.822 * 1e3 # Enthalpy from other source
eos.set_enthalpy_of_formation(1, H_ref_other) # Set H_ref for component 1

T2 = 350 # Some other temperature
p2 = 8e6 # Some other pressure
H2, = eos.enthalpy(T2, p2, [1], eos.LIQPH) # Gives value comparable to other source at (T2, p2, LIQPH)

If you are interested in doing calculations for chemical reactions: I've double checked, and found that the absolute enthalpy of ammonia computed using ThermoPack coincides closely with reported enthalpies of formation here, here, here, and here, so I would trust the ThermoPack value to give reasonable reaction enthalpies, rather than the value from NIST webbook calculator (as mentioned, it is somewhat unclear to me where their reference value comes from).

[vegardjervell]

@ffraile, I'm closing this as not-planned and moving the thread to the discussions page, as the question seems to have been cleared up, and to not be an issue with thermopack. Don't hesitate to post on the discussion if you have more questions :)