Gravel-corrected Kd values

Citation
Di. Kaplan et al., Gravel-corrected Kd values, GROUND WATE, 38(6), 2000, pp. 851-857
Citations number
14
Categorie Soggetti
Environment/Ecology,"Civil Engineering
Journal title
GROUND WATER
ISSN journal
0017467X → ACNP
Volume
38
Issue
6
Year of publication
2000
Pages
851 - 857
Database
ISI
SICI code
0017-467X(200011/12)38:6<851:GKV>2.0.ZU;2-F
Abstract
Standard measurements of solute sorption to sediments are typically made on the <2 mm sediment fraction. This fraction is used by researchers to stand ardize the method and to ease experimental protocol so that large labware i s not required to accommodate the gravel fraction (>2 mm particles). Since sorption is a phenomenon directly related to surface area, sorption measure ments based on the <2 mm fraction would be expected to overestimate actual whole-sediment values for sediments containing gravel. This inaccuracy is a problem for ground water contaminant transport modelers who use laboratory -derived sorption values, typically expressed as a distribution coefficient s (Kd), to calculate the retardation factor (Rf), a parameter that accounts for solute-sediment chemical interactions. The objectives of this laborato ry study were to quantify the effect of gravel on strontium Kd and Rf value s and to develop an empirical method to calculate gravel-corrected Kd value s for the study site (Hanford Site in Richland, Washington). Three gravel c orrections, Kd(gc) values, were evaluated: a correction based on the assump tion that the gravel simply diluted the Kd(< 2mm) and had no sorption capac ity (Kd(gc),(g=0)), a correction based on the assumption that the Kd of the intact sediment (Kd(tot)) was a composite of the Kd(< 2mm) and the Kd(> 2m m) (Kd(gc,g=x)), and a correction based on surface area (Kd(gc,surf)). On a verage, Kd(< 2mm) tended to overestimate Kd(tot) by 28% to 47%; Kd(gc,g=x) overestimated Kd(tot) by only 3% to 5%; and Kd(gc,g=0) and Kd(gc,surf) unde restimated Kd(tot) by 10% to 39%. Kd(gc,g=x) provided the best estimate of actual values (Kd(tot)); however, Kd(gc,g=0) was appreciably easier to acqu ire. Although other contaminants will likely have different gravel-correcti on values, these results have important implications regarding the traditio nal approach to modeling contaminant transport which uses Kd(< 2mm) values. Such calculations may overestimate the tendency of gravel-containing sedim ents to retard contaminant migration.