Jump to content
RockWare Support Forum


  • Content Count

  • Joined

  • Last visited

Everything posted by Stefan

  1. In connection with another activity I had a look back in an older forum topic of mine and the following conversation by Email with Tom Meuzelar that was not held via the forum because of its very detailed content. By doing so I wondered about one of the comments of a developer that was added and that I obviously have overlooked before: If he is using a dataset with "activity model: pitzer" in the header lines, he should be aware there are known issues with that method, as implemented in EQ3/6 some years ago. Most significantly, in order to get correct results, he needs to include by hand in the database zero values for whatever ternary parameters are not explicitly set there, but are employed in his calculation. Are these “issues” relevant to EG3/6 only (where I am already aware of them). Otherwise, if such ternary parameters have to be set to zero also in GWB datasets, this should be emphasized somewhere in the documentation of GWB. In particular it would be interesting to know what GWB does if such parameters are not set to zero. Kind regards, Stefan
  2. Hi Tom Thank you for this clever trick. I thought about adding NaCl myself. But, I did not believe that 1 nmol would be enough for adjusting a charge imbalance that is ten times higher. It proofs how important it is to make use of the charge balance option of GWB. (However, it seems that the mismatch between step 3 and step 4 is worth a look by the programmer.) Kind regards, Stefan
  3. Hi Tom Sorry, I was a bit careless when writing down my questions. Furthermore, for clarity I should have added also the react file for the second case which shows the results I wonder about. This file is now added (SiO2_test2.rea). The only difference to the first file is the species Si2O3(OH)4-- that is now not suppressed. The problem occurs between the calculations c and d. Step #100 of calculation c does not coincide with step #0 of calculation d. Regards, Stefan P.S. I am using React 7.0.6. SiO2_test2.rea
  4. Hi Tom, thank you for your quick reply. I have not made a plot of charge imbalance errors up to now. (By the way: How can I create such a plot?) I just had a look on the listed molalities of the individual species in the output files. For case 1 (only the first polymeric species considered) all is fine. When unsuppressing the second polymere species it starts fine with output “a”. But when reaching the end of the calculation (last entry in output “d”) the listed molalities are far from summing up to charge balance. Regards, Stefan
  5. Hello, I am trying to model the dissolution of amorphous silica at high pH (addition of NaOH). To improve the dataset of Reardon 1992 I added some polymeric Si species from the literature. This was successful in terms of calculated Si concentrations compared to literature data. Unfortunately, with increasing reaction progress the charge balance gets progressively lost. I have added an extract of the data file as well as an input script. It is running well as it is. But when unsuppressing the second polymeric species the described error occurs. It is getting worse with additional polymeric species (not included in the data file). Thank you for any help. Best regards, Stefan SiO2_test.dat SiO2_test.rea
  6. I created a batch job for about 100 individual react input files. When starting the job several error messages occurred after a couple of react input files have been calculated: “Warning: Failed to center progress window.” followed by “Warning: Failed to update progress window.” Independent of the first two messages “React stop: SetIcon – winfo. React is ending.” sometimes appears and causes the termination of react. I used the “pause” command within the batch job to identify a possible corrupt react file. I recognized that the error messages occurred only when I clicked too quickly on the “OK” button to proceed with the next react file. Bay waiting one or two seconds before clicking “OK” the react files run without trouble. All of the react jobs worked fine when called separately. What is going on? How can I regain the comfort of the batch job? (I am using react, release 7.0.6) Best regards, Stefan
  7. Stefan

    React: error messages

    Hi Tom Thank you for your support. Your answers help me unterstanding these kinds of error mesages. Script 1 works fine with charge ballancing on SO4-- for the second part. There seem to be some little secrets that one has to learn with time and/or by asking a more experienced user via the support forum. Regards, Stefan
  8. By calculating simple salt systems sometimes error messages occur, that unfortunately are not self-explanatory (for me). Similar problems have been addressed within the user forum before but the answers do not apply for the systems shown in the examples below. Example Script 1: reset data = thermo_hmw.dat balance off temperature = 25 1.0000 kg free H2O K+ = 2e-8 molal Mg++ = 1e-8 molal SO4-- = 2e-8 molal react 25 mol Schoenite go pickup entire go The first part works well. But picking up the entire system as base for further calculations leads to the following error message: *N-R didn't converge after 999 its., maximum residual = Largest residual(s): Resid Resid/Totmol Cbasis -------------------------------------------------------- SO4-- -51.16 0.1124 171 -------------------------------------------------------- By replacing “pickup entire†by “pickup fluid†it works well (but solids are lost as expected with this command). It seems that there is some conflict with minerals swapped into the basis. But if so, how to solve this? A similar problem arises when using another script that should lead to approximately the same result. Example Script 2: reset data = thermo_hmw.dat balance off temperature = 25 1.0000 kg free H2O Mg++ = 1e-8 molal SO4-- = 1e-8 molal swap Schoenite for K+ 25 free mol Schoenite go This results in the following message: Ionic strength out of range. Largest residual(s): Resid Resid/Totmol Cbasis -------------------------------------------------------- SO4-- -8.029e+009 1 9e-009 Mg++ -4.014e+009 1 9e-009 What puzzles me most is the message “Ionic strength out of rangeâ€. As one can see from the first part of the first script this system has an ionic strength of about 8. With other setups one may easily create systems with an ionic strength quite double that value. Can anybody help me understanding this kind of error messages and tell me how to deal with them? Regards, Stefan
  9. Hello I have trouble keeping compatibility of thermodynamic data between datasets for different geochemical codes. Thus, I would like to know some details on the Pitzer formalism incorporated in GWB. Obviously, the Aphi parameter can be individually set for a specific dataset (and for the different temperatures herein) by the debye huckel a (adh) parameter. For the thermo_hmw dataset you take adh(25°C) = 0.5092. This corresponds with a value for Aphi of 0.3909 what when rounded is the value from Bradley and Pitzer 1979 (J. Phys. Chem 83/12) for 25°C and saturated pressure. In the original HMW 84 paper 0.39 was used. Looking in Archer & Wang 1990 ( J. Phys. Chem. Ref. Data 19/2) Aphi should be 0.39148. On page 351 in Craigs “green book†a value of 0.3920 is documented for calculating the Etheta values. Further, the values provided for Aphi in the thermo_hmw dataset agree with those of Bradley and Pitzer 1979 at 0, 25 and 60°C. At higher temperatures small deviations exist. Theses differences in Aphi (at 25°C and other temperatures) are small and thus leading to small differences in the resulting activity coefficients. For the application on “large scale real world problems†one can neglect this. But by trying to build a database with compatible thermodynamic values this makes the difference. Thus, I have two questions: 1) What is the value for Aphi used in GWB (or in different parts of GWB) for calculating activity coefficients following the Pitzer formalism? 2) What is the function behind the temperature dependency of Aphi in GWB datasets and - I assume - all other values listed in the same way for different temperatures? Best regards, Stefan
  • Create New...