A Survey of Charge‐Balance Errors on Published Analyses of Potable Ground and Surface Waters

Abstract

AbstractAlthough the charge balance for a single chemical analysis of a water's electrolytes is not a reliable gauge for the accuracy of that analysis, the percent charge‐balance error (%CBE) becomes more credible as a means for evaluating analytical technique when applied to groups of analyses.Just how good are charge balances for chemical analyses of potable ground and stream waters in mainline geological journals? Starting with each journal's first volume, a search found 68 articles in six journals (Appl. Geochem., Chem. Geol, Geochim. Cos. Acta., Ground Water, J. Hydrol, and Water Resources Res.) that had what was deemed to be “complete” analyses of individual samples' major ions. Analyses of brines and hydrothermal solutions were excluded as were “composite” averages. A total of 1,062 %CBEs were computed and tabulated, and the average was 3.99%$ 6.56 (1σ).The average %CBE (by journals) ranges from 1.55% to 9.34%. Quality of %CBE is slightly better for articles published after 1970. Not surprisingly, low ionic strength samples are prone to charge‐balance errors > 10%. Of the 1,062 charge balances calculated, 612 were positive and 450 were negative. Eight of the 68 articles had analyses that were evenly balanced between positive and negative CB errors. There are 21 articles in which the majority of analyses had negative charge‐balance errors. Thus a clear majority (39) of articles had analyses in which there were more positive than negative charge‐balance errors.Systematic error is likely for situations where the analyses in an article are predominantly positive or predominantly negative. The likely reason for the former occurrence is lab determination of alkalinity from the nonacidified field aliquot. Failure to filter samples may cause the latter situation—especially in carbonate terranes. Systematic laboratory errors involving salt standards and dilutions can also yield charge‐balance errors for a suite of analyses that are skewed to either positive or negative errors. The care with which the analyst did his/her job may also impact charge‐balance errors.