Membrane Behavior of Two Backfills from Field-Constructed Soil-Bentonite Cutoff Walls

Two soil-bentonite cutoff-wall backfills obtained from construction sites, one in New Jersey and one in Delaware, were tested for the existence of membrane behavior. Both backfills were designed as a mixture of dry bentonite (3–4% by dry weight) and the locally excavated soil blended with bentonite water slurry to provide slumps ranging from 100 to $150\mathrm{mm}$ (from 4 to $6\mathrm{in}.$ ). The results of the membrane tests indicate that both backfills exhibit membrane behavior. Further, the magnitude of the membrane behavior increases with decreasing void ratio. However, the magnitude of the increase in membrane behavior in these construction-site backfills was lower than that previously reported for model backfills prepared in the laboratory. The difference in the membrane behavior is attributed, in part, to a lower percentage of clay in the construction-site backfills relative to the model backfills. Nonetheless, based on the measured membrane efficiencies for the two field-constructed backfills, the total liquid flux $\left(q\right)$ through the cutoff walls can be expected to be reduced relative to that in the absence of membrane behavior $\left(q_h\right)$ by 1–10% for the cutoff wall in Delaware and 7–8% for the cutoff wall in New Jersey, depending on the void ratio. Thus, the results of this study suggest that membrane behavior in field-constructed cutoff walls can be significant, depending on the void ratio and the clay content of the backfill.