Phenolic group shows chemical shift delta value below or upto 10 in PMR spectrum due to intermolecular hydrogen bonding in specific solvents but it shifts sometimes upto 15. What can be the reason/s for that?
Did you try to run your NMR at different temperatures? Tautomers are really great suggestion. You might also consider solvent composition (water or protic solvents content, pH, ionic strength, etc.). I can imagine many factors, and of them are not really obvious (e.g. non-stabilized chloroform tend to decompose over time and release HCl, so pH might sharply decrease).
For mobile protons (those that can be rapidly exchanged like O-H of acids, phenols, alcohols...) the observed chemical shift is the time-averaged shift. To get a sharp signal(s), there two limiting situation: (i) very fast and (ii) very slow exchange rate. In situation (i) the shift will move according to the conditions of H-bonding, pH, T, ... One possibility is to "fix" the proton in one situation, for example use of DMSO(d6) (a strong H-bond acceptor) as solvent, i.e. in large excess, gives almost only the [RO-H...DMSO] species and a more stable shift. pH increase or decrease will give fast excjhange but the signal will be that of (or close to) the added acid or of the water formed by neutralization of O-H. Note that in intermadiate situation (i) and (ii) the signla bradens , and may even disappear because it is too broad and hidden ion the baseline noise. In fact, I am just summarizing parts of the very nice review indicated by Jurgen.