R-515B Density Uncertainty

[MEC-RnD-NO]

Hello,

Does REFPROP publish uncertainties for the model of R515B? Specifically density uncertainty in subcooled liquid.

R515B is a mixture (91.1% R1234zee, 8.9% R227ea), and so for I have just been making this mixture myself in excel since it is not a predefined mixture in REFPROP. I use =@FluidString10(A1:B1,A2:B2,"MASS") as a substitute for my cell containing my refrigerant, where A1 contains R1234zee, B1 contains R227ea, A2 contains 0.911, and B2 contains 0.089.

It is my understanding that R-515B is modeled in REFPROP using the mixture formulation (using the HMX.BNC file) based on the pure-fluid equations of state for R-1234ze(E) and R-227ea. With that said, how does the model compare to experimental data? Does R515B experimental data exist that the REFPROP model could be compared to?

I spent some time looking myself without luck. I saw this study mentioned here #358 (comment) but there is a hole in the data where the R515B mixture would lie.

If no R515B dataset exists, do you have a ballpark guess for what density uncertainty in subcooled liquid region may be?

The experiment I plan to run is sensitive to density uncertainties >~0.1%

[marciahuber]

Interaction parameters for the binary pair R-1234ze(E) and R-227ea were determined in this paper from 2023 https://doi.org/10.1063/5.0135368
You can email REFPROP@NIST.GOV and request an updated HMX.BNC file with these parameters, and we also can send a .MIX file for R515B.
The manuscript details the data used and gives comparisons with the data.

[MEC-RnD-NO]

Awesome that's exactly what I needed, but now I have a follow up issue.

I put the new HMX.BNC and R515B.MIX in their appropriate locations. Now, I get the following error when selecting any predefined mixture (not just R515B)

[SETUP error 102] Error in file or premature end of file

This same error is returned in excel when using any predefined mixture, as well as when I make a mixture myself using the method I described in my original post.

[marciahuber]

Well...that's because I sent the wrong file. :( It crashes for any mixture, which is not particularly helpful. Will try again.

[MEC-RnD-NO]

Got it working thank you,

I noticed the data for R-1234ze(E)/227ea from https://doi.org/10.1063/5.0135368 used a mole fraction range of 0.33 - 0.67 R-1234ze(E).

Taking the formula of R515B: R-1234ze / R-227ea (91.1% / 8.9%), which is expressed as weight percent, and converting it to mole fraction (R-1234ze(E): 114.04 g/mol, R-227ea: 170.03 g/mol):

The mole fraction of R515B is 0.939 R-1234ze(E).

This is outside of the 0.33 - 0.67 range used in the study. Knowing this, do you believe that the new HMX.BNC will still give me the density AARD of 0.03% and a maximum deviation of 0.5% described in the paper when I use it for R515B?

[HightopRaven]

In my opinion the short answer to your question is yes.
"The density measurements cover the full density range and are reproduced with an AARD of 0.03% and a maximum deviation of 0.5%." (§4.5 on page 16 of https://doi.org/10.1063/5.0135368)

Yes the composition of R-515B is an extrapolation beyond the range of compositions used for experimental measurements in the referenced publication (per Table 12 on page 18 of https://doi.org/10.1063/5.0135368). But consider that the mixture model will converge to each of pure fluid density values as the mole fraction for that component approaches 1.0. The resulting uncertainty at an intermediate composition (between measurement points) is not trivial to estimate. At 0.68~0.74 mole fraction of R-1234ze(E) the density uncertainty is likely larger than it was at 0.67 mole fraction where measurements were made. Likewise at a 0.939 mole fraction of R-1234ze(E) the density uncertainty is likely larger than for pure R-1234ze(E) where measurements were made. If you look at the contents of the R1234ZEE.FLD file, it cites the equation of state source as https://doi.org/10.1007/s10765-016-2040-6 and then summarizes that work as follows:
? The uncertainty in density in the liquid and vapor phases is 0.1% from 200 K to
? 420 K at all pressures. The uncertainty increases outside of this temperature
? region and in the critical region. [...]

You could also look into the uncertainty of the binary pair measurement data here: https://doi.org/10.1021/acs.jced.3c00104