Inconsistent pH and `H[+1]` concentration in selected presets

[SuixiongTay]

Description

• The ValueError is reproducible for presets: flue_gas, leachate, and plating.
• Removing H[+1], OH[-1], and H2O(aq) from the preset solutes resolves the ValueError for flue_gas and plating.

Error message

ValueError: Cannot specify both a non-default pH and H+ at the same time. Please provide only one

ToDo(s)

• Check the inconsistencies in yaml and json processing

[rkingsbury]

@SuixiongTay - question about this. I noticed that all the newly added presets in #317 include the charge_balance as a parameter, and I can't figure out how that got there. I want to confirm you used `Solution.to_file(<yaml_filename.yaml>) to generate the presets.

The error should not be thrown as long as the H[+1] concentration and the pH are consistent with each other (see here)

I thought I had tracked down the cause of this as a subtle issue related to volume (and I've made some improvements to from_file as a result), but it still isn't resolving the issue, so I want to make sure there isn't something going on with the source of the file itself.

To be clear, the errors are small. For example in the difference between the two is 2.8645520 vs. 2.8644211, but I still can't figure out why there's a discrepancy at all.

[SuixiongTay]

Ah yes that's an error on my end, since I wrote the yaml files manually. I will get this corrected.

[rkingsbury]

OK I didn't realize that. It's very important to use the built-in methods to create the yaml files - otherwise the from_file code doesn't know what to do with the contents. I suspect that if you re-generate them using from_file and the latest GitHub, that will solve the error.

Since the bug in this issue might be caused by the source of the files, I'll go ahead and merge #347 .

[rkingsbury]

I'm going to assume #350 resolved this. Let me know if the problem re-occurs and I'll re-open.

[rkingsbury]

OK, it appears this problem still occurs for selected presets. We will need to investigate in more detail how those presets were generated, because a close inspection of the data shows that their H[+1] concentration is not consistent with the pH property, and that's why the error is occurring - these two values are not within tolerance of each other.

This internal inconsistency should be impossible in a Solution because pH is calculated directly from the H+ concentration (see this code).

Because this issue seems confined to only 3-4 presets, we can tackle it on a longer timeline (not necessary for next release).