A reanalysis of chemical abundances in galactic PNe and comparison with theoretical predictions

New determinations of chemical abundances for He, N, O, Ne, Ar and S are derived for all galactic planetary nebulae (PNe) so far observed with a relatively high accuracy, in an effort to overcome differences in these quantities obtained over the years by different authors using different procedures. These include: ways to correct for interstellar extinction, the atomic data used to interpret the observed line fluxes, the model nebula adopted to represent real objects and the ionization corrections for unseen ions. A unique `good quality' classical-type procedure, i.e. making use of collisionally excited forbidden lines to derive ionic abundances of heavy ions, has been applied to all individual sets of observed line fluxes in each specific position within each PN. Only observational data obtained with linear detectors, and satisfying some `quality' criteria, have been considered. Such observations go from the mid-1970s up to the end of 2001. The observational errors associated with individual line fluxes have been propagated through the whole procedure to obtain an estimate of the accuracy of final abundances independent of an author's `prejudices'. Comparison of the final abundances with those obtained in relevant multi-object studies on the one hand allowed us to assess the accuracy of the new abundances, and on the other hand proved the usefulness of the present work, the basic purpose of which was to take full advantage of the vast amount of observations done so far of galactic PNe, handling them in a proper homogeneous way. The number of resulting PNe that have data of an adequate quality to pass the present selection amounts to 131. We believe that the new derived abundances constitute a highly homogeneous chemical data set on galactic PNe, with realistic uncertainties, and form a good observational basis for comparison with the growing number of predictions from stellar evolution theory. Owing to the known discrepancies between the ionic abundances of heavy elements derived from the strong collisonally excited forbidden lines and those derived from the weak, temperature-insensitive recombination lines, it is recognized that only abundance ratios between heavy elements can be considered as satisfactorily accurate. A comparison with theoretical predictions allowed us to assess the state of the art in this topic in any case, providing some findings and suggestions for further theoretical and observational work to advance our understanding of the evolution of low- and intermediate-mass stars.