A quick google uncovered this old paper from 1926 in which the equilibrium is plotted over pH 4-8 (see fig. 3). It would appear that the enol form predominates above pH 5.8.
While I follow the concept, I know little about the methods used and cannot give an authoritative interpretation of the results myself. If I had to explain why the enol form is the major form of the pyruvate ion, on first inspection I would probably go for an intramolecular hydrogen bond between the OH and the carboxylate (which will not be present in the keto form). How significantly stabilising this would be in aqueous (protic) solution though, I am not sure.
I am a little pressed for time today, but I think I found a relevant reference:
Keto-enol equilibria in the pyruvic acid system: determination of the keto-enol equilibrium constants of pyruvic acid and pyruvate anion and the acidity constant of pyruvate enol in aqueous solution
Y. Chiang , A. J. Kresge , P. Pruszynski
J. Am. Chem. Soc., 1992, 114 (
, pp 3103–3107
DOI: 10.1021/ja00034a053n
It looks to me that pK
E for pyruvate is about 5 and for pyruvic acid is about 3.2.