Related Information




Tips & Tricks


page last updated: 28.11.2011

Getting Started

Getting started with thermocalc is often the hardest thing to do. This section provides a walk through of the process of drawing your first pseudosection. The walk through is divided into the following sections:

Before You Start

Using thermocalc is not difficult, but it is not always self-evident what to do or how to start. Before undertaking calculations on your own rocks we recommend that you look at the documentation, tutorials and course lecture notes, including making calculations on one of the provided example bulk rock compositions and recalculating published diagrams. We also recommend that you read the following paper, as well as recent papers using thermocalc on similar rocks to your own.

Powell, R, Holland, TJB, & Worley, B, 1998. Calculating phase diagrams involving solid solutions via non-linear equations, with examples using THERMOCALC Journal of Metamorphic Geology 16, 577–588.

To undertake your own calculations and to draw a phase diagram there are a number of programs and datafiles you will need. The required files can be downloaded as the thermocalc bundle.  As well as some documentation this bundle contains the thermocalc program, the internally consistent dataset, the drawpd program, as well as a small number of example datafiles for use. You also need to decide on what chemical system you will need to use, based on the rocks you are modelling and the problem you are trying to solve.

back to top

What System Should I Use?

Deciding on what system to use in your calculations is one of the first critical steps. Using the wrong system will often result in you having to re-do the diagrams in a more appropriate system. As a general rule, it is often better to use one of the smaller simpler systems like KFMASH if you are investigating a process in general rather than attempting to derive detailed P-T information from a suite of rocks. However, if you wish to derive quantitative P-T information from a suite of rocks, you will typically have to use one of the larger systems, and in many cases the largest system available.

There are some fairly basic logic steps that you can follow to make this decision.  Firstly, the system you use should be large enough to encompass the main compositional range of your rock. However, you may have to also consider elements that are only a minor constituent in your rocks if those elements are preferentially partitioned into one of your main minerals (eg Mn in garnet, Ti in biotite) as these can strongly influence the stability of these phases. Secondly, you need to consider what minerals you will/may need to consider to appropriately model your rock, it is useful to look at diagrams from the literature for similar rocks for guidance. It is worth noting that using a smaller system for the sake of simplicity can strongly influence the results (see White et al., 2007, JMG v25 p 511-527 for a more detailed discussion).

A range of different systems and mineral a-x coding can be found in the datafiles page.

back to top

What Diagram?

There are a number of different diagrams that can be constructed that show different things, but these can be reduced to three main types, petrogenetic grids, compatibility triangles and pseudosections. Of these, pseudosections are the most commonly used.

Petrogenetic grids/P-T projection

image of petrogenic grid

Mouse over image to enlarge

These show only the invariant and univariant equilibria and are independent of bulk rock composition. In some of the larger systems these grids are of little use as only a small number of stable univariant  reactions exist. In other systems, such diagrams may contain so many univariant and invariant equilibria that they are quite difficult to decipher. However, in many systems (e.g. KFMASH) grids can be constructed that show much useful information about the controlling low-variance equilibria. Typically, the grids for such systems have already been published and it is not necessary to reproduce them.

Compatibility diagrams

image of compatibility diagram

Mouse over image to enlarge

Triangular compatibility diagrams, such as the commonly-used AFM diagram for metapelites, are a useful way to represent mineral composition for divariant, trivariant and quadrivariant equilbria at fixed P and T. Such diagrams have to be reduced to 3 components to be used as the apices (e.g. Al2O3-FeO-MgO) by projecting from “in excess” phases (e.g. quartz, muscovite & H2O). The apices do not have to simply be the basic oxides, and more complex apices can be used. Likewise, the projecting phases can be both pure endmembers and solid solution phases.


image of pseudosection

Mouse over image to enlarge

Pseudosections are multivariant phase diagrams that are constructed for a fixed bulk composition or for a compositional vector. P-T pseudosections using a fixed bulk composition are the most commonly used, and typically the best place to start. P-x and T-x pseudosections may be useful for comparing the evolution of a suite of samples that more or less monotonically vary in their composition. Pseudosections are the most commonly constructed type of diagram.

While P-T grids, compatibility diagrams and P-T and P/T-x pseudosections are the most commonly calculated, there are other types of diagrams that can be drawn that may be less automated.

back to top

Setting Up

Having chosen a system, the minerals to include and what diagram to draw, you will need to set up your datafiles for your calculation. Firstly you need a folder with all the input files and the software. Below is an annotated example of a folder with the required files.


More information on input files can be found in the input files section of the documentation

To run thermocalc using these files several scripts in the tc-prefs and the script file have to be used. This is largely a matter of switching on or off scripts in the prefs and script files. Example script files that can be used as templates for each type of diagram and calculation can be found in the datafiles page.

The key scripts needed to get things running here are:

dataset     55          % dataset is tc-ds55
calcmode    1           % mode 1 calcs ie phase diagrams
scriptfile  ptpstute    % defines the script file to use

in the tc-prefs file


axfile      NCKFMASHp

in the thermocalc script file

to run thermocalc efficiently there are a number of other scripts in the script file related to the type of calculation you are undertaking. Again, consult the input files, output files and scripts documentation.

back to top

Bulk Rock Compositions

A critical step in calculating a pseudosection is picking a bulk rock composition that represents a volume of equilibration in a rock. Obviously, deciding on a volume of equilibration is a somewhat subjective process. In high-T rocks it is common to use XRF analyses as bulk rock compositions, with the size of the sample chosen for crushing determined by petrographic analysis of the sample. Other methods such as quantitative x-ray maps of thin sections, or point counting combined with EPMA analyses. Once you have a bulk rock composition, which is normally in wt%, it has to be converted to mol% and reduced to your model chemical system.

Reducing the measured bulk rock composition to the model system requires some care and can be done in several ways. A more detailed discussion of this can be found in the tips and tricks section (coming soon). This section includes discussion on estimating Fe3+ contents and choosing how to estimate H2O. Simply ignoring a component is the simplest approach, however, this can introduce bulk composition aberrations if that component combines with others in a phase you are not modelling. For example, simply ignoring CaO and Na2O in a plagioclase-poor rock will affect the model SiO2 and Al2O3 proportion.

back to top

Opening thermocalc

On both the Mac and PC, thermocalc requires that all the input files and the program are in the same folder, the same applies to drawpd where the input and prefs files must be in the same folder as the program. Thermocalc operates in different ways on the Mac and PC. On the PC, thermocalc can be simply double-clicked to run. The recent versions of  thermocalc for Mac run in the unix shell program called “terminal” (which is hidden away in the utilities folder inside the applications folder).  To run thermocalc on Mac the following steps are required:


1. open terminal

screenshoot open terminal

The window should look something like that above. You can adjust the window size to suit.

2. you have to now tell Terminal what directory (folder) thermocalc is in. This is done by typing “cd” followed by a space, and then dragging the folder icon onto the terminal window, then hit return. Terminal will now list that directory, and the terminal window will look something like  

screenshot change directory image


3. to run thermocalc, type “./tc330” for non-intel macs and “./tc330i” for intel machines and hit return. This should bring up the standard thermocalc statements and questions such as:

screenshot run command mac

back to top

Starting your own Diagram

Getting started on a new diagram (ie when you don’t know what the final diagram will look like) is not always easy, and choosing which equilibria to start with requires a reasonable understanding of the likely equilibria to exist for a particular bulk rock composition.  In some rock compositions (e.g. metapelites) there are some key equilibria that will be generally always be seen (e.g.  bi + sill = g + cd) in some form depending on what system you are using (typically this exists as a narrow divariant or trivariant field of coexisting bi-sill-g-cd-q-ksp-liq ±pl±ilm±ru±mt…..). In other rock compositions this is not always so. You could try and guess some equilibria from what minerals you have in you rock, but this is not always reliable.  The most stable (i.e. lowest Gibbs energy) assemblage can be calculated using the dogmin script, but as thermocalc is not primarily set up as a Gibbs energy minimisation program, this process may not work if thermocalc could not calculate the most stable equilibria.

From personal experience, the most reliable way to start a diagram on a new bulk rock composition (at least while you are new to thermocalc) is to build a T-x diagram from an existing  diagram that you or someone else has constructed (i.e. known equilibria) to your new diagram.  To make this process relatively simple, it is best to use a bulk composition for the known equilibria that is as close to your sample as possible.  Once you have built up a series of equilibria across the diagram you can use these equilibria as a starting point for your own P-T pseudosection. It is typically best to trace some low variance equilibria across the diagram as calculating where the mode of phases goes to zero is typically easier and more reliable than trying to guess where a new phase comes in.  A tutorial on this approach can be found in the tutorials page.

back to top