R513a data + formula problem in excel

[gsavell121]

Description

I am quite new to Refprop and using the excel integration. Apologies if I have made a basic error, or if my thermodynamic understanding is lacking.

I am trying to look up the temperature and pressure of R513a based on known enthalpy and density values. If I put the values directly into refprop, I mostly seem to get reasonable results. However, when I use excel, large portions of the data return errors.

There is one data point in particular I am interested in at the moment which is returning an error - see steps to reproduce below.

I have tried to investigate the problem by plotting graphs both in REFPROP and Excel. As far as I can see, there seems to be a small data issue in REFPROP for this mixture around the 200kg/m3, 30°C, 270j/g point. Additionally, for some reason excel also cannot return large sections of the T-D or P-D plots between 250-700kg/m3. However, I'm unsure if problems are related. Please see additional information + attached excel file below.

Steps to Reproduce

1. Using the refprop excel formula: '''=@RefProp("T","R513a","HD","SI with C",245,257,0)'''

**Expected behavior: returns "20.978" (°C) ** as per using the point lookup facility in REFPROP, and which seems reasonable

**Actual behavior: returns "#[DHFLSH error 4] One or more inputs are out of range; Pressure below zero: P = -22.8283 MPa." **

Versions

REFPROP Version: 10
Operating System and Version: Windows 10 Enterprise
**Access Method: Excel (formula wizard) **

Additional Information

(also see attached excel file: R513a data error query.xlsx)

To try to find the problem, I have used excel to plot Temperature vs Density for different Enthalpies - there seems to be a data point error at ~200kg/m3, 30°C, 270j/g point. Also, the results between 250-700kg/m3 are somewhat random - sometimes errors come up and sometimes erroneous results come back.

There seem to be similar issues with the Pressure vs Density graphs. The data point error persists - see the odd data point in the Refprop plot on the 30°C line at around 140kg/m3:

[ianhbell]

I think the problem is reference states. Default reference states in Excel and REFPROP GUI are different for mixtures, which is why when you put in density and the enthalpy, you get a different output. If you were to start off with properties without reference state dependence (e.g., temperature and pressure), from those calculate density and enthalpy in both places, and then calculate outputs given density and enthalpy, that should give you the same answers.

[gsavell121]

Thanks for the quick help - I'll have a look at the reference states and see about calculating from other properties to begin with.

Presumably the the odd data point in the R513a data is a different issue, or could that be caused by the same thing?

[ianhbell]

Let's first fix the reference state issues. I think that is 99% of it, then also you might need to enable the saturation splines in Excel: usnistgov/REFPROP-wrappers#106

[EricLemmon]

It could also be a bug that we fixed some time ago. If the reference state issue doesn't work, send Ian an email and he'll send you the beta version: Ian.Bell@nist.gov

[gsavell121]

Enabling saturation splines in Excel seems to have fixed most of the issue. I now get nice complete curves on the graphs and the formulas return sensible numbers - with the exception of one strange missing data point, now at around 300kg/m3:

[gsavell121]

Regarding refence states, I haven't been able to figure out the correct changes in the VBA - I must be missing something because Excel is stubbornly returning the same results what ever settings I put in. If you can spot my issue I'd appreciate it! Thanks.

The steps I have followed are:

1. Set i = 1 rather than 2:
   

2. Uncommented the 3 lines, and changed the option to 'ASH' for example:
   

3. Saved, closed, and restarted Excel

4. Verified that the VBA changes persisted and that the addin is still looking at this copy of the .xla

5. Recalculated the formula

[EricLemmon]

Have you tried the beta version yet? I believe this is fixed with that version. There is also an alpha version that could be tried if the beta fails.

[EricLemmon]

Has this issue been resolved?

[gsavell121]

Hi Eric - not really, sorry. I've tried the Beta version, and I'm still having problems with both the strange data point in R513a and changing the reference states in Excel.

[EricLemmon]

I've lost your email address, can you send it again to eric.lemmon@nist.gov