generally there is a critical point where the dew and bubble line intersect, for many mixtures critical point differs from cricondenbar (max VLE P) and cricondentherm (max VLE T) but there are many exceptions to these rules, see
according my experience it should be so but ther are cases where the usual thermodynamics (SRK, PR and all related modifications) predict strange curves with one, two or no critical points, with some mixture two liquid phases are predicted, don't know how accurate the models are in these cases.
frank45, if the iteration technique in the model is not converging correctly, you could get multiple answers, its kinda like the square root of a number, there are 2 square roots.
not sure about the exact reasons (may be dcasto is correct) but there are cases where simulators can produce wrong result, recently for a pipeline study we have compared different tools (with same EOS, parameters etc.), at low pressures they all agree but going up near critical we got different fractions of condensate or even wrong phases (for example liquid where it should be vapor) etc.
Anton
does anyone know of some "official" review/comparison about convergence in critical regions etc. etc. ? I have experienced similar problems (different results) and I am not sure that in all cases our reference process simulator does produce reliable results...