Limitations for low initial concentrations

[plechacek]

Hello,

I am using PHREEQC to evaluate radium sinks in groundwater at a field site. I have several models evaluating radium co-precipitation with sulfate and carbonate minerals, and sorption to illite, and Fe/Mn oxides. Are there any limitations or considerations when using low initial aqueous radium concentrations (3.9-51 x 10^-16 M)? I.e., are there any errors or concerns with accuracy when modeling such a low range of concentrations?

[dlparkhurst]

I'd be careful and do a lot of testing. Sometimes concentrations are censored at 1e-14, but I'm not sure exactly when it might happen.

If you have problems, you can probably increase concentrations by several orders of magnitude and correct back to the real concentrations in post processing. This also may be a way to test your calculations.

[plechacek]

Thanks! I ran the models again with initial Ra concentrations three orders of magnitude higher. I only had one model result that was different, which was sorption to HFO and HMO for my sample with the lowest initial Ra concentration. Based on that it appears concentrations started to be censored somewhere between 3.9*10^-16 and 4.1*10^-16 M. I have corrected back to the real concentrations post processing for that sample.

[dlparkhurst]

I forgot that there is an undocumented option to set the level of censoring. By default it is 1e-25. You can set it lower as follows.

Code: [Select]

KNOBS
-min_total 1e-30

It hasn't been test too much, so I don't know if it will cause other problems.