The impact that unrecognised differences in the chemical patterns of Galactic
globular clusters have on their relative age determinations is studied. The two
most widely used relative age-dating methods, horizontal and vertical, together
with the more recent relative MS-fitting method, were carefully analyzed on a
purely theoretical basis. The BaSTI library was adopted to perform the present
analysis. We find that relative ages derived using the horizontal and vertical
methods are largely dependent on the initial He content and heavy element
distribution. Unrecognized cluster-to-cluster chemical abundance differences
can lead to an error in the derived relative ages as large as ~0.5 (or ~6 Gyr
if an age of 12.8 Gyr is adopted for normalization), and even larger for some
extreme cases. It is shown that the relative MS-fitting method is by far the
age-dating technique for which undetected cluster-to-cluster differences in the
He abundance have less impact. Present results are used in order to pose
constraints on the maximum possible spread in the He and CNONa elements
abundances on the basis of the estimates - taken from the literature - of the
Galactic globular clusters relative age dispersion obtained with the various
relative age-dating techniques. Finally, it is shown that the age-metallicity
relation found for young Galactic globular clusters by the GC Treasury program
is a real age sequence and cannot be produced by variations in the He and/or
heavy element distribution.