Flow past Bench-Scale Vertical Ground-Water Cutoff Walls

Abstract

An experimental study was conducted to evaluate factors affecting flow rates past soil-bentonite (SB), geomembrane (GM), and composite geomembrane-soil (CGS) vertical cutoff walls. Intact walls and walls containing defects (inadequate keys, windows, and poor joint seals) were studied. For intact cutoff walls, CGS walls had the lowest flow rate, followed by SB and GM walls. CGS walls typically had flow rates as much as 100 times lower than comparable GM walls. Flow rates increased by a factor of 2–160 when the GM walls contained defective joints or the SB walls contained permeable windows. For all wall types, an effective key was required to achieve low flow rate past the wall. For GM walls, under seepage was not effectively controlled when the GM was placed in direct contact with the aquitard. Flow rates for GM walls placed in direct contact with the aquitard were nearly equal to the aquifer flow rate without a wall. Better control of underseepage was possible with SB and CGS walls placed in direct contact with the aquitard because the soft bentonitic backfill conformed to the surface of the aquitard. Hydraulic conductivity of unsealed joints in GM walls was estimated from flow rates past GM walls where a portion of the joint was unsealed. These hydraulic conductivities ranged between 1.8 × 10