For a 'scavenger problem' known in the literature the exact many-particle solutions obtained by the authors for recombination of excess electron in solution are compared with the approximations commonly used to interpret experimental data. Distinctions related to an important role of time correlations are analyzed. The differences are shown to be great enough to lead to qualitatively incorrect description of the experiment. In particular, it is shown that the superposition approximation can decrease essentially the ultimate probability of geminate recombination in the presence of scavengers. This, in turn, may give rise to considerable errors in the determination of geminate pair parameters. (c) 2006 Elsevier B.V. All rights reserved.