The major shortcoming of computer modelling is that changes to the model can easily be made and thus the effect of say a 10% error in one of the component concentrations can very readily be investigated. Similarly, stability constant data may be varied over a range in order to see if this has any effect on the speciation profile. This type of operation also serves to identify those species which are of major importance. Last but not least, it is very quick and can easily be repeated.