FAQ: QUESTIONS ASKED BY NOVICE USERS (c) Australian Geological Survey Organisation 1999-2002 (c) Yuri Shvarov 1999-2002 Last updated: 3 May 2002 This file does not substitute for the comprehensive HCh documentation. It only provides quick answers to some questions asked by the package users. Please see HCh documentation for further details. Your own contribution to this file in terms of additional questions will be very much appreciated. LIST OF QUESTIONS 1 HCh 1.1 GENERAL QUESTIONS 1.1.1 What does HCh stand for? 1.1.2 What are main areas of HCh application? 1.1.3 Can HCh be used to carry out mass transfer (reaction path) calculations? 1.1.4 Can I use HCh for the construction of conventional activity-activity diagrams (e.g., fO2-pH)? 1.1.5 What is the HCh T-P Range? 1.1.6 What is the HCh calculation range in terms of the ionic strength? (NEW) 1.1.7 Where can I get the summary of the HCh specifications? 1.1.8 What are advantages/disadvantages of HCh compared to EQ3/6 in the field of "conventional" modelling in hydrothermal geochemistry? 1.1.9 What are limitations of the free energy minimisation approach employed by HCh from the practical point of view? 1.1.10 Can I run HCh on a Macintosh computer? 1.1.11 How do I get assistance in using HCh? 1.2 HCH INTERFACE 1.2.1 What is the best way to get started for a devoted Windows/Mac user? 1.2.2 Are there any plans to upgrade HCh to a special Windows version? 2 UNITHERM USAGE 2.1 UNITHERM DATABASE 2.1.1 What is the guaranty that the included UNITHERM database has no incorrect entries? 2.1.2 How to prevent inadvertent database changes made by a user? 2.1.3 How can I add a new species to the UNITHERM database? 2.1.4 Why do I have a wrong Clapeyron slope for a UNITHERM mineral with phase transitions? 2.1.5 How to extend the default database? 2.1.6 How to create and use my own customised database? 2.1.7 What other models for aqueous species can I use apart from those of Helgeson and Ryzhenko? 2.1.8 Why can't I read some extended database references? 2.2 UNITHERM REACTION MODE 2.2.1 How do I write a new reaction? 2.2.2 Why do I have apparently wrong pK values? 3 MAIN USAGE: PROBLEM DEFINITION & PROCESSING RESULTS 3.1.1 What is the best way to get started with modelling? 3.2 CREATING & EDITING FILES 3.2.1 Why do I have a message on database errors, but all the HCh programs seem to work properly? 3.2.2 What is the best way to choose system components (species)? 3.2.3 Can I add new input substances to an existing Blank file? 3.2.4 What is the best way to define my rock composition? 3.2.5 Can I use mineral names to define my rock composition? 3.2.6 Can I use brackets in mineral formulae? 3.2.7 Can I fit a long mineral formula in a short input window? 3.2.8 What is the best way to define my fluid composition? 3.2.9 How can I delete "Initial values" from *.BL or *.IN files? 3.3 PROCESSING GIBBS RESULTS 3.3.1 Within MAIN, how can I copy a GIBBS result file to another directory? 3.3.2 How can I obtain fO2/fH2 values from my results? 3.3.3 What is the best way for plotting the modelling results? 4 RUNNING GIBBS 4.1 CALCULATING EQUILIBRIA 4.1.1 Why do I have an error message "Restrictions incompatible"? 4.1.2 How can I restrict the volume of mineral assemblage for my isothermal-isochoric system (T, V = const)? 4.1.3 Can I calculate models with isoenthalpic boiling? 4.1.4 What factors affect the equilibrium calculation time and how can I speed up my calculations? 5 EQUILIBRIUM MODELLING (MAIN + GIBBS) 5.1.1 How can I define pH and Eh of aqueous phase for computing Eh-pH stability diagrams? 5.1.2 What is the easiest way to calculate solubilities of minerals? 5.1.3 Can I calculate metastable equilibria using HCh? -------------------------------------------------------------- 1 HCH 1.1 GENERAL QUESTIONS 1.1.1 What does HCh stand for? HCh stands for HydroChemistry. But you can calculate absolutely "dry" systems as well. 1.1.2 What are main areas of HCh application? - Hydrothermal geochemistry - Hydrochemistry and hydrogeology - Experimental geochemistry (aqueous) 1.1.3 Can HCh be used to carry out mass transfer (reaction path) calculations? Yes, HCh is IDEALLY suited for MASS TRANSFER calculations. In fact, such calculations are probably the main area of the HCh application. You should keep in mind, however, that overall non-equilibrium PROCESS is modelled using local-equilibrium approach without any provisions for modelling of the process kinetics. Speaking of the REACTION PATH modelling the answer depends on its definition. You cannot model a selected (specified) sequence of mass-transfer calculations that follow predefined phase or reaction boundaries. However, employing a sequence of mass-transfer calculations (e.g., "titration" models), it is possible to establish the direction of the chemical PROCESS and the sequence of main reactions leading to a new equilibrium state. Limitations imposed by the local equilibrium approach would still apply. 1.1.4 Can I use HCh for the construction of conventional activity-activity diagrams (e.g., fO2-pH)? No, you cannot use HCh for this purpose directly. However, the results of HCh calculations may be plotted on such diagrams. For a concise review of the relationship between mass- transfer calculations and activity-activity diagrams please refer to T. Wolery (1992): http://www-ep.es.llnl.gov/www- ep/esd/geochem/EQ36manuals/eq36pkg.pdf (page -4-). 1.1.5 What is the HCh T-P Range? In UNITHERM, Gibbs free energies are tabulated up to 500øC and 5kb. But it is stated in the package specifications that the possible calculations involving an aqueous solution cover the temperature range of 0 - 1000øC and pressure range of 1 - 5000 bar at water densities exceeding 0.35 g/cm3. Are the calculations of HCh valid to 500 or 1000øC or does the upper temperature limit vary phase by phase according to the conditions pertaining to the basic data for each phase? (a) A T-P grid defined in UNITHERM is not related to the range of HCh calculations -- but the mentioned density limitations should be met. For example, you cannot calculate equilibrium composition for an aqueous phase at 500øC and 500 bar where dH2O = 0.26g/cm3. The calculations involving aqueous phase are limited by the MHKF model for aqueous species. You can modify the UNITHERM T-P grid at any time by using the UNITHERM command (DefTab). Just follow the syntax of the default table to create your own T-P grid. You can also modify the default T-P grid by the UNITHERM /TAB command-line option. (b) The upper temperature limit for a phase (e.g., mineral) will not limit calculations. It will limit stability only of this particular phase. 1.1.6 What is the HCh calculation range in terms of the ionic strength? (NEW) Activity coefficients of charged aqueous species are calculated by HCh according to the extended Debye-Huckel equation. As the Debye-Huckel equation does not work properly at high ionic strength, HCh cannot check ALL POSSIBLE states of your system. To avoid overflow in the Debye-Huckel procedure during calculations for concentrated solutions, the free energy minimisation code of HCh (GIBBS) artificially restricts the value of the ionic strength by 10. Thus, you should keep in mind the following: 1. If the GIBBS output contains the aqueous phase with ionic strength in excess of 10, THE RESULT IS INCORRECT! Please use other geochemical programs that support calculation of activities of aqueous species in brines. 2. If the GIBBS output contains the aqueous phase with ionic strength less than 10, the equilibrium should be found correctly. We believe that the free energy is a convex function, and the local minimum should be found in the reasonable area. 3. If the GIBBS output does not contain the aqueous phase at all, but it is present in your System file as a possible phase, this result can be either correct or not depending on properties of brines that GIBBS cannot calculate. 4. You are not affected by these problems if your System file does not include the aqueous phase. Note that definitions "correct" and "incorrect" in the current context refer exclusively to your thermodynamic model but not to the model agreement with nature. 1.1.7 Where can I get the summary of the HCh specifications? Please refer to your User's Guide (page 48), GUIDE*.TXT file, or the WWW page http://www.agso.gov.au/geochemistry/HCh/hch_www_spec.html 1.1.8 What are advantages/disadvantages of HCh compared to EQ3/6 in the field of "conventional" modelling in hydrothermal geochemistry? Presently there are no users with comparable experience in using both packages. However, William McKenzie (Univ. of California) contributed a preliminary answer to this question. We assume that you have at least partial access to the HCh documentation (see Question 1.1.10). To learn more about modelling codes, and EQ3/6 in particular, browse the following WWW sites: http://www-ep.es.llnl.gov/germ/WR-codes.html and http://www-ep.es.llnl.gov/www-ep/esd/geochem/eq36.html#Manuals The main differences in the usage of the HCh and EQ packages result from the different algorithms employed to calculate equilibrium compositions of chemical systems. EQ3/6 employs a method based on equilibrium constants, whereas HCh uses the free energy minimisation technique. Bethke (1996), Zeleznik and Gordon (1960, 1968) and Brinkley (1960) have argued that both methods are computationally and conceptually equivalent. However, they result at least in different ways of setting your modelling problems. EQ3 is an equilibrium aqueous speciation model. EQ3 solves for equilibrium by using the log Ks, aqueous concentrations of species, activity coefficient models, and mass and charge balance constraints. The resulting set of equations (including non-linear ones) is solved by the Newton-Raphson method. To calculate speciation using EQ3, you need to specify basis species (e.g., Na+, Cl-, SiO2(aq), etc) representing all elements in the system. The chosen basis species are used to write reactions for the rest of aqueous species and for saturation of minerals. The log K for each reaction is calculated at a given pressure at 6 temperatures (25-300øC); log Ks at other temperatures are interpolated using the calculated constants. The free energy minimisation code of HCh (GIBBS) calculates equilibrium composition of chemical systems "directly" without having to specify basis species, write reactions, and calculate log Ks. GIBBS solves for equilibrium by using mass and charge balance constraints, free energy of components (species) at standard state conditions, and activity coefficient models. Using HCh, you can specify any temperature-pressure path for you calculations. In case of EQ3 you have to use a different data base for each pressure or set of basis species. EQ6 is mass transfer model. After the equilibrium state for the aqueous solution is determined by EQ3, EQ6 titrates a small amount of reactant (e.g., a mineral) into the solution, and the new equilibrium state is again determined by EQ3. The equilibrium system continues to evolve by continuing to titrate small amount of the reactant into the solution until the fluid reaches saturation with the reactant or the reactant is totally consumed. Aqueous composition, mineral precipitation, mineral dissolution are determined at each step of the titration. In turn, HCh has its own tools for mass-transfer calculations. Processes like titration or mixing can be modelled by HCh very easily and effectively using the means provided by the Control menu. EQ3/6 includes some provisions for kinetic calculations, including "real-time" calculations. In contrast, the HCh usage is restricted only to local- equilibrium models. As to the databases, HCh can be easily made compatible with EQ3/6 if necessary (at least using SUPCRT92/96 SPRONS files). 1.1.9 What are limitations of the free energy minimisation approach employed by HCh from the practical point of view? Not all local-equilibrium processes can be directly modelled using the free energy minimisation method. For example, you cannot directly model processes taking part under constant volume, or isoenthalpic processes. See answers to Questions 4.1.2 and 4.1.3. 1.1.10 Can I run HCh on a Macintosh computer? To the extent of our knowledge, HCh has been successfully run on PowerMac G3 under MacOS 8.0 using SoftWindows 2.0 (make sure your math coprocessor is turned ON!). No problems should be expected with better hardware and DOS/Windows emulation programs. To make sure that the software runs on your system, try to download and run the HCh DEMO version from http://www.agso.gov.au/geochemistry/HCh/hch_www_demo.html 1.1.11 How do I get assistance in using HCh? Please refer to the following files included in your distribution package: Installation: README.TXT General information: GUIDE*.* Running examples: EXAMPLES.TXT Distribution and licensing: LICENCE*.TXT Disclaimer: LICENCE*.TXT Latest information: WNEW*.TXT Frequently asked questions: FAQ.TXT (this file) You may also consult the manual "HCh: a software package for geochemical equilibrium modelling. User's Guide. Australian Geological Survey Organisation, Record 1999/25". This text is an extended hard-copy version of the GUIDE*.TXT file. The manual can be ordered from the Sales Centre of the Australian Geological Survey Organisation (sales@agso.gov.au) You can get some immediate help during your HCh working sessions. In the main UNITHERM or MAIN windows you can always press to obtain a general reference on the available commands, keys, and menus. Normally, at the bottom screen lines you can also find a tip(s) on a currently possible action(s). 1.2 HCH INTERFACE 1.2.1 What is the best way to get started for a devoted Windows/Mac user? I am a devoted Windows/Mac user who is not accustomed to the DOS environment or DOS shells. As a result, I am finding using the program interface a little bit frustrating. What is the best way around it? PLEASE NOTE THAT AT THE BOTTOM SCREEN LINES THERE IS NORMALLY A TIP(S) ON A POSSIBLE ACTION. In the main UNITHERM or MAIN windows you can always press to obtain a general reference on the available commands, keys, and menus. 1.2.2 Are there any plans to upgrade HCh to a special Windows version? Development of the 32-bit Windows version of HCh is currently in progress. Please contact us at the end of 2002 re the anticipated release date and pricing. 2 UNITHERM USAGE 2.1 UNITHERM DATABASE 2.1.1 What is the guaranty that the included UNITHERM database has no incorrect entries? The default UNITHERM database has been used at the Department of Geochemistry, Moscow State University for the last 10 years for research and industry-oriented projects. All data were typed in and checked "manually". However, see the DISCLAIMER in your LICENCE*.TXT file. 2.1.2 How to prevent inadvertent database changes made by a user? Inadvertent changes made by the user can be blocked by setting the Read Only attribute for any of the database files. See DOS help for the ATTRIB command. 2.1.3 How can I add a new species to the UNITHERM database? The data can be added interactively in a UNITHERM session. Use or commands to activate the UNITHERM input mode. If you are working under Windows, you can use Windows Copy and Paste commands to facilitate data entry. To add a new or alternative data for an already existing species you may find useful the UNITHERM Copy command, . By duplicating the record, you may reduce a possibility of mistypes. Just edit the copied record using the command. Starting from the HCh version 3.4 you can use the UNITHERM command-line option /IMPort to load data from the text file DATA.TXT. This file is intended for data interchange between different program packages and databases. 2.1.4 Why do I have a wrong Clapeyron slope for a UNITHERM mineral with phase transitions? Many mineral data in the default UNITHERM database come from the compilation of Helgeson et al. (1978). The format of mineral data seems to be completely consistent with that of Helgeson, especially if I use the J<->cal switch . Why do I have an apparently wrong value for the Clapeyron slope (dP/dT) for minerals with phase transitions (quartz, for example)? Please note that UNITHERM uses the reciprocal of the Clapeyron slope expressed as dT/dP*1000, rather than the "true" Clapeyron slope dP/dT. Just view your mineral record again, paying attention to the exact name of the relevant record field. The dT/dP representation is more convenient from the computational point of view. 2.1.5 How to extend the default database? If I want to add a new mineral/phase to the default UNITHERM database, where would I preferably obtain the relevant thermodynamic data? Generally, the database is YOUR OWN responsibility. Presently, there are a number of quite comprehensive compilations available (e.g., Robie & Hemingway, 95; SUPCRT92; Holland & Powell, 98; etc). The choice of the particular database will depend on your PARTICULAR PURPOSES, PROBLEMS, and REQUIRED ACCURACY. Databases themselves are generally tuned for particular application areas or for particular chemical systems. Please note that the result of simple "data mixing" sometimes might be disastrous-you should make sure that the data are reasonably consistent. 2.1.6 How to create and use my own customised database? Your own customised database can be created in a number of ways. In most cases it is expedient to copy the initial UNITHERM database (files UT.DIR, UT*.BIN & UT*.REF) into a separate directory, and modify the database copy. If you want to keep only a part of the initial records, start UNITHERM from the new database directory with the /SCOPY command-line option. Now you can select the components that should be kept in the new database. The database compiled from this selection will be saved in the WT.DIR and W*.BIN. files when you finish your UNITHERM session. To make the new database usable, copy these files into UT.DIR and U*.BIN files over the initial UT*.* files. After that you can work with UNITHERM as usual. Refer to Question 2.1.3 how to add new database components. Use the same command-line option (/SCOPY) if you want to create your own database starting from scratch. Just deselect all the database components in your /SCOPY UNITHERM session, and copy the empty database on exit. Again, the new database will be saved in the WT.DIR and W*.BIN files. Refer to Question 2.1.3 how to add new database components. You can use SEVERAL versions of UNITHERM databases in your everyday work. Just (a) keep your relevant database files (UT*.*) in separate directories (not the UNITHERM program!), and (b) start UNITHERM (UT.BAT) from an appropriate database directory. However, make sure to specify the proper PATH to the appropriate database for MAIN and GIBBS in a MAIN session. Follow the steps summarised below: Menu Bar -> Options -> Set Paths -> ...UNITHERM database -> Enter Save this setup in the preferred directory: Menu Bar -> Options -> Save Setup -> Choose the preferred directory -> Enter 2.1.7 What other models for aqueous species can I use apart from those of Helgeson and Ryzhenko? You may include in your database aqueous species defined through pK of exchange reactions using the "complex" (Ryzhenko-Bryzgalin) format of the UNITHERM database. (1) For example, you can use an assumption that pKdiss NaHS (aq) is approximately equal to pKdiss NaCl (aq). This implies that pK of the exchange reaction NaHS (aq) + Cl- = NaCl (aq) + HS- is independent of T and P and equal to 0. In this case, the complex NaHS (aq) must be defined as follows: NaHS (aq) B.Ryzhenko model Coef. Basic species 1 NaCl (aq) 1 HS- -1 Cl- pK(298) 0.000 A(zz/a) 0.000 B(zz/a) 0.00 ---------------------- Calculated from the assumption: pK(NaHS) = pK(NaCl) (2) Using a similar approach, you can also use the method suggested by Gu et al. (GCA, 1994). This method implies that dGr(T,P) of an exchange isocoulombic reaction is approximately constant. In this case, temperature and pressure dependence of pKr(T,P) values will be represented by a simple equation: pKr = (298/T)*pKr(298,1), and an ambient known value of pKr can be used to calculate the reference pKr(298,1) value. For example, Zn(HS)2 (aq) can be expressed in terms of the exchange reaction Zn(HS)2 (aq) + 2Cl- = ZnCl2 (aq) + Cl- with pKr(373,PSat) = 10.43 (pKr(298,1) = 13.05). You can enter in the database the following record: Zn(HS)2 (aq) B.Ryzhenko model Coef. Basic species -2 Cl- 1 ZnCl2 (aq) 2 HS- pK(298) 13.054 A(zz/a) 0.00 B(zz/a) 0.00 ---------------------- Calculated from the approach by Gu et al. (1995) 2.1.8 Why can't I read some extended database references? I want to obtain an extended reference for a UNITHERM component (), but instead have unreadable strings of Windows symbols. Is this a program bug? No. Some of the extended references in the primary default UNITHERM database may be provided in Russian. They can be read in normal Russian characters in a conventional (non-Windows) DOS session or in a full- screen Windows session (). 2.2 UNITHERM REACTION MODE 2.2.1 How do I write a new reaction? Press enter when you have completed the reaction name. Now specify the stoichiometry of your reaction. For example, for the reaction 0.25 Magnetite + 1 H2S (aq) = 0.25 Pyrite + 0.5 Troilite + 1 H2O type in PyPoMt buffer Component Coefficient Pyrite 0.2500 Troilite 0.5000 H2O 1.0000 Magnetite -0.2500 H2S (aq) -1.0000 and press Esc. Note that the program calculates pK values (= -logK) of reactions. 2.2.2 Why do I have apparently wrong pK values? I tried to calculate pK (Delta G) values for a reaction involving a gas species, but they do not make any sense. Is this a program bug? Make sure that you are using the "1bar" reference state for gases to calculate the parameters of your reaction. Just start UNITHERM with the /1BAR command-line option. Otherwise, UNITHERM tabulates values g(t,p) = g(t,1bar) + RTln(P), and by default will calculate pK and Delta G values of your reaction on this basis. 3 MAIN USAGE: PROBLEM DEFINITION & PROCESSING RESULTS 3.1.1 What is the best way to get started with modelling? Refer to the "Equilibrium Modelling" section of the GUIDE*.TXT file or the HCh manual (especially, the provided flow chart). The experience has shown that it is a good idea to BROWSE THE HCH DOCUMENTATION, help windows, and the provided examples, even if you are extremely reluctant to do so! 3.2 CREATING & EDITING FILES 3.2.1 Why do I have a message on database errors, but all the HCh programs seem to work properly? When I create a new System file I get a message from MAIN that the UNITHERM database contains errors. However, MAIN seems to work normally, and UNITHERM cannot find any database errors. What can be wrong? You may have violated the HCh naming conventions for aqueous species. When MAIN reads your database to create a new System file, it parses the chemical formulae of all database components to obtain the list of the chemical elements and species stoichiometry. The error message tells you that there are some formulae that cannot be parsed properly. However, these components are not shown by MAIN and they will not be offered for your selection in the list of the available species. This error can occur only for aqueous species, where the component names are represented by combinations of a conventional chemical formula and an optional comment. An optional comment must be separated from the chemical formula either by the space or comma. The error message tells you that you need to check the names of the aqueous species that you have added to UNITHERM. 3.2.2 What is the best way to choose system components (species)? Is it a good idea to include all the offered components in my new System file when I create it? From a point of view of an experienced user rather NO than yes. If you want to avoid computational problems, loss of computational time or strange results it is better to limit yourself to the components that are (1) likely to be stable or significant in your models; (2) come from reasonably consistent datasets. However, observing the condition (1) may require some preliminary calculations. Observing the condition (2) is the direct responsibility of a literate user. You will be able to EXCLUDE excessive components later through EDITING of your System file, but you will not be able to ADD new (or excluded) components to an existing System file. It might be a good idea to keep a reference copy of your initial System file with the longest list of components. 3.2.3 Can I add new input substances to an existing Blank file? You can add new input substances ONLY INSTEAD of previously defined substances (the total number of substances will stay the same). If you are not sure how to define the input for your system, you may want to reserve a few extra lines for the future use. Just enter a few dummy substances allowed by your chemical system within the bottom rows of your Blank file, for example: H O S * * * Substance ..... Unit * * * 0. H2O kg 1. H2S moles 2. H2S moles 3. H2S moles * * * End of file * * * To specify the real content of H2S for your system in the Input file, you will need to enter only one H2S value, and set the rest of the "H2S" fields to zeroes. 3.2.4 What is the best way to define my rock composition? What is the best way to define my rock composition in a Blank file for the purposes of geochemical modelling? Choose "...other substances" from the option list when you create a new Blank file. Then you can specify the composition of your rock either in terms of (1) oxides and chemical elements, or (2) minerals, in grams or moles. For example: H O K Na Al Si * * * Substance ..... Unit * * * 0. H2O kg 1. SiO2 g 2. K2O g 3. Na2O g 4. Al2O3 g * * * End of file * * * H O K Na Al Si * * * Substance ..... Unit * * * 0. H2O kg 1. SiO2 g 2. KAlSi3O8 g 3. NaAlSi3O8 g 4. KAl3Si3O12H2 g * * * End of file * * * When you create an Input file to prepare Control files for modelling, it will be often convenient to normalise your rock composition per 1 kg. 3.2.5 Can I use mineral names to define my rock composition? I want to define my rock composition as a number of minerals, can I just enter the appropriate mineral names in the provided fields? No. You must use an EXPLICIT CHEMICAL FORMULA of the mineral. For example, if you want to use quartz, type in SiO2. If you have started MAIN from Windows, you can also start UNITHERM and use it for further reference using Windows Cut and Paste capabilities if necessary. 3.2.6 Can I use brackets in mineral formulae? I want to define my rock composition as a NUMBER OF MINERALS. Can I use brackets in the mineral formula? You can (e.g., AlO(OH)), but they consume space in long formulae (see below). You can also use square brackets [ ] and the asterisk (*) (e.g., CaSO4*2H2O). 3.2.7 Can I fit a long mineral formula in a short input window? I want to define my rock composition as a NUMBER OF MINERALS, but my formula won't fit in the provided window. Is there any way around this problem? The provided input box allows you to enter a chemical formula of 14 characters. If you want to use a mineral/substance with a longer formula, you can "split" your mineral/substance in two. For example, epidote Ca2FeAl2Si3O13H (15 characters) can be represented within two separate lines as: 1. Ca2FeAl2O7H 2. Si3O6 3.2.8 What is the best way to define my fluid composition? What is the best way to define my FLUID COMPOSITION in a Blank file for the purposes of GEOCHEMICAL modelling? Choose "...other substances" from the option list when you create a new Blank file. Then you can specify your fluid composition in terms of any convenient and sensible substances: H O K Na Al Si * * * Substance ..... Unit * * * 0. H2O kg 1. CO2 moles 2. NaCl moles 3. HCl moles 4. H2S moles * * * End of file * * * We recommend you to AVOID definition of fluid compositions using COMBINATION OF THE "...OTHER SUBSTANCES" AND CHARGED AQUEOUS SPECIES (see Question 4.1.1). If you intend to enter a solution composition obtained by preliminary calculations with GIBBS (for example, if according to your model the initial solution should have an equilibrium composition), you can specify the solution composition in terms of chemical elements. At the same time, composition of rock can be better specified in terms of minerals or oxides. Using the "...other substances", you can combine a number of ways to set the bulk composition of your system. However, if you want to specify the chemical composition of your solution in terms of chemical elements, make sure to do it with the maximal accuracy. Otherwise you may face the problems addressed in Question 4.1.1. 3.2.9 How can I delete "Initial values" from *.BL or *.IN files? Copy your file to a file with a new name, delete the source file, and rename your copy back to its "parental" name. The initial values will be deleted. This operation is possible because initial values are not copied. 3.3 PROCESSING GIBBS RESULTS 3.3.1 Within MAIN, how can I copy a GIBBS result file to another directory? Unfortunately, you cannot do it. However, to ensure a convenient work environment, follow the steps summarised below: 1. Create your own working directory for a particular problem (or group of problems). 2. Start HCh from this directory. 3. Menu Bar -> Options -> ChDdir Request -> None -> Enter 4. Menu Bar -> Options -> Save Setup -> ...in the current directory -> Enter All your files and settings relevant for a particular problem will be easily kept together. See the HCh documentation about the HCH.BAT and UT.BAT files. 3.3.2 How can I obtain fO2/fH2 values from my results? Unfortunately, GIBBS does not output fO2 or fH2 values. Use log K values for the reactions O2(g) = O2(aq) log K1, and H2(g) = H2(aq) log K2 to recalculate your results, e.g.: log fO2 = log mO2 - log K1. However, make sure you use the proper reference state for gases (T, P = 1 bar). 3.3.3 What is the best way for plotting the modelling results? Write your results in a binary file (*.RE). Process it within HCh by means of the Result menu, and save the selected results as an ASCII *.REX file. The data will be ready for import by EXCEL and any other graphing package. Just follow the chain of options summarised below: Menu Bar -> Result -> Binary file -> Choose Result (*.RE) file -> Choose Cross sections (graph) -> Choose Horizontal -> ... -> Press (save) -> Choose All (autoselect) -> Choose table layout -> Edit a filename and/or press Now your data can be imported by EXCEL. Use the REX.XLS workbook to facilitate this import. 4 RUNNING GIBBS 4.1 CALCULATING EQUILIBRIA 4.1.1 Why do I have an error message "Restrictions incompatible"? The bulk system composition you have specified cannot be defined in terms of the possible system phases. Most likely something is wrong either with your speciation model or the way you have specified the bulk system composition. Go back and have a close look at your Control, Input, or System files. The most frequent cause of this message is the absence of the charge balance in the aqueous solution. 4.1.2 How can I restrict the volume of mineral assemblage for my isothermal-isochoric system (T, V = const)? Equilibria in an isothermal-isochoric system (T, V = const) can be calculated only by the minimisation of the Helmholtz free energy (the Gibbs free energy is a thermodynamic potential of a system only at constant T and P). GIBBS does not calculate the Helmholtz free energy. A fundamental consequence of the existence of the equation of state is that the system pressure is a DEPENDENT function of the equilibrium state at the specified temperature and volume, and cannot be set to an independent value. Likewise, for the given temperature and pressure, the volume is a function of equilibrium, and cannot be independently specified. The phase rule for the isothermal-isochoric system is different from the phase rule for the isothermal- isobaric system. Moreover, as was shown by Korzhinskii, for some systems under a fixed volume, different phases may be subjected to different pressures. All these factors make it impossible to solve such problems by the minimisation of the Gibbs free energy of the system. 4.1.3 Can I calculate models with isoenthalpic boiling? As GIBBS minimises the Gibbs free energy of the system, processes in isoenthalpic systems cannot be calculated directly. You need to estimate an isoenthalpic path for your system independently in terms of temperature and pressure. The resulting T-P curve can be specified in your Control file. 4.1.4 What factors affect the equilibrium calculation time and how can I speed up my calculations? There are many factors that affect the calculation rate for apparently similar problems. Clearly, the calculation time will depend on the set of the phases and components specified for your problem (the more phases and phase components, the longer the calculation time). However, the calculation time can dramatically vary for the systems with the same set of the chosen components. Here we discuss the most obvious factors pertaining to this situation. 1. The bulk system composition. This is one of the most critical factors affecting the calculation rate. For example, for hydrothermal systems involving a rock and an aqueous phase, calculations on rock-dominated systems are generally faster than for fluid-dominated systems. This is an intrinsic property of the minimisation algorithm employed by GIBBS. An unfavourable calculation case may occur when the bulk composition of your system is close to the saturation with an infinitely small amount of a new phase. For example, in a gas-water system the bulk composition of your system can match an equilibrium between the aqueous solution and an infinitesimal quantity of the gas phase. This situation may occur if you model isothermal-isobaric degassing of a fluid (e.g., according to an algorithm like [*] = [A] + k*[G], where k <= 1). If the amount of the gas phase at a new equilibration step is as small as the uncertainties of the mass balance equations, the equilibrium might not be calculated at all. 2. Redox equilibria. Calculations on systems with redox equilibria are inherently slower than on systems without them. We recommend you to avoid calculations of this type in closed systems without specifying sufficient quantities of redox buffers. For example, you may wish to calculate equilibria in the system H2O-H2S. If you include in the solution model only H2O, H+, OH-, H2S (aq), and HS-, you will encounter no problems- calculations will be fast. However, if you add other species such as O2 (aq), H2 (aq), HSO4-, and SO4--, the calculation time will increase dramatically. Without a redox buffer (i.e. sufficient quantity of a chemical element in different oxidation states of relatively comparable amounts), redox equilibria will be extremely sensitive to very small variations in concentrations of species in different oxidation states. If these variations are comparable to the uncertainties of the mass balance equations, the equilibrium might not be calculated at all. In geochemical modelling, try to avoid this problem by setting reasonable redox constraints from the very beginning. To this end you can explicitly specify redox buffers in your rocks (e.g., coexisting Fe-oxide + Fe- bearing silicate, or Fe2O3/FeO ratio). You should apply the same approach to the fluids that are not initially constrained by the fluid-rock equilibrium (e.g., by specifying total H2 (aq) or H2S/H2SO4 ratio). Of course, you cannot always avoid this problem-you might be interested exactly in the redox equilibria in the H2O-H2S system resulting from oxidation of H2S through reactions with water. If you are not interested in redox equilibria at all, and they will not be significant in your models, you might discard them COMPLETELY. For example, you might be interested in solubility of calcite in CO2-dominated aqueous solutions. In this case, make sure that the components of your system comprise elements only in the same oxidation state each (e.g., do not include H2 (aq), O2 (aq), and CH4 (aq) in your aqueous phase!). Some of the factors affecting the calculation time are related to temperature and pressure specified for your calculations. This relationship results in an apparent T-P dependence of the calculation rate. 3. Greater variety of low-temperature minerals. If your problem includes a large number of solid phases stable at low-temperatures, calculations for the low temperature region will be slower. The problem of choice of the stable mineral assemblage will require more time as more possibilities exist in this region. 4. Behaviour of electrolytes and activity models for aqueous species. In aqueous systems the increase in the calculation time might be related to calculation of the activity coefficients of aqueous species. Generally, higher temperatures and low pressures will promote association of aqueous complexes, lower values of ionic strength, and values of activity coefficients closer to unity. All these factors make the aqueous solution closer to ideal. The overall result is a faster calculation convergence. The exact calculation time will be affected by the exact models for the activity coefficients of charged and neutral aqueous species (e.g., the default GIBBS options vs /c=@1, /b=@3). For a fluid-dominated system the fastest calculation time should be expected for high-T, low-P region for activity coefficient models favouring association of ion pairs. You can speed up your calculations for aqueous systems in a "manual" calculation mode of GIBBS when using Blank or Input files. To this end, calculate a number of sequential equilibria in a row going along the defined temperature/pressure gradient. In this case GIBBS will use the previously obtained results for the aqueous solution as the initial approximation for further calculations, and the new results will be obtained faster. Note that to take the advantage of this mechanism you do not need to specify the GIBBS /s option. In the general case, you can speed up your calculations by setting the GIBBS /s option (compute sequentially). In this case the entire equilibrium phase assemblage obtained at the previous calculation step will be used as the initial approximation for further calculations. Without this option GIBBS estimates the starting phase assemblage by itself. Note that in most cases of sequential calculations it is expedient to go "down" the temperature; it makes the usage of the /s option more effective. 5 EQUILIBRIUM MODELLING (MAIN + GIBBS) 5.1.1 How can I define pH and Eh of aqueous phase for computing Eh-pH stability diagrams? Unfortunately, you cannot do it. Physically, pH and Eh are DEPENDENT variables and cannot define a state of equilibrium. Some chemical programs indeed allow you to specify these variables. However, strictly speaking their results are not correct, as a "conditional" equilibrium is not the true chemical equilibrium. To fix pH and Eh values you can only use buffer substances, like in real chemical system. 5.1.2 What is the easiest way to calculate solubilities of minerals? What is the easiest way to calculate solubility of minerals if I already know the stable mineral assemblage? You can use a system with the "perfectly mobile components" (PMC). In this case, GIBBS does not need to search for minerals that are already known to be stable. For example, if you want to calculate solubility of minerals in the Quartz-Calcite- Wollastonite assemblage, follow the steps summarised below: 1. Create your System file: a) From the list of the possible phases, select "Perfectly mobile components", "Individual phases", and "Aqueous solution". b) From the list of the chemical elements, select C, Ca, H, O, and Si. c) From the list of the PMC, select Quartz, Calcite and Wollastonite. d) From the list of individual phases, select all minerals. e) From the list of aqueous species, select aqueous species as usual. 2. Create your Blank file: a) Choose definition of PMC fugacities in terms of lg(P) or ln(P). b) Choose any option for definition of the bulk chemical composition. c) Define fugacities of the PMC as constant, and set their log values to 0. Specify the measurement units for water as kg. 3. Run GIBBS in the interactive mode: a) Specify your Blank file as a source file. b) Start GIBBS. Input temperature, pressure, and the amount of water (1 kg) -- and calculate equilibrium. If you obtain only the aqueous solution, it represents the solubility of your specified mineral assemblage. Precipitation of any new mineral(s) means that dissolution of your assemblage is incongruent. If you get the error message "Solution unlimited", your specified mineral assemblage is NOT stable. 5.1.3 Can I calculate metastable equilibria using HCh? The most general approach for calculating such equilibria is using artificial "User Elements". Using user elements you can specify the chemical substances that cannot be re-equilibrated. For example, thermodynamically unstable organic substances in natural waters can survive within long periods of geological time without any modification (see the EXAMPLE.TXT and relevant HCh files). In many cases, metastable equilibria can be calculated in a more straightforward way. For example, you can simply exclude from your system a potentially stable mineral, thus preventing it from precipitation (e.g., you can "stabilise" amorphous silica versus crystalline quartz). Another example: if you want to calculate composition of a natural water in equilibrium with the atmosphere (including atmospheric oxygen and nitrogen) and include all the relevant aqueous species from the UNITHERM database, you will obtain a solution of nitric acid with pH ~ 1. On the contrary, if you exclude from your system NO3-, you will be able to calculate the metastable equilibrium-aqueous solution with pH ~ 7.