Seepage Cutoff Walls: Lessons Learned, Implemented, and to be Implemented

A seepage cutoff wall is one of the most common seepage mitigation measures in embankment dams and levees. Specifications for seepage cutoff walls are mainly performance-based, where criteria for determining performance are usually short-term hydraulic conductivity and/or strength parameters based on laboratory testing of field collected samples. However, long-term performance of the seepage cutoff walls could be dependent on many design, in-situ, and construction related factors, which may not necessarily be predicted by short-term laboratory-based performance benchmarks. A recent post-construction investigation of conventionally-constructed (i.e. open-trench method) slag cement-cement-bentonite (SCCB) walls gave an opportunity to understand long-term characteristics of SCCB walls. Key observations include: a) variable as-built in-situ properties across a cutoff wall depth due to potential initial consolidation of solid particles and other factors, b) potential for development of shear stress zones along the side walls resulting in longitudinal cracks, and c) mixing with in-situ soils. Recent levee improvement projects successfully adopted higher initial strength and higher minimum solid contents criteria, which could reduce the negative impacts incurred by delayed slurry set-up and initial settlement. The initial set-up of the open trench cement slurry cutoff walls may also be important in horizontally-mixed deep mixing method (DMM) cement slurry cutoff walls. In specifications for the DMM walls, hydraulic conductivity criterion is highly emphasized and strength criterion is often under-emphasized by requiring strengths significantly lower than the lateral earth pressures. As a result, a combination of a lower hydraulic conductivity criterion with a lower 28-days strength criterion can be achieved with very high water-cement ratio (such as 3.1:1). This very soft, high water content soil-cement-bentonite mix may result in an initial compressible element. Recent specifications for the DMM cutoff walls require an initial 7-days strengths (similar concept as open-trench SCCB cutoff walls), which can be achieved with use of a reasonably lower water-cement ratio and may result in potential for less compressible elements. Also, 28-days strength requirement needs to consider the lateral load that the wall would experience and long term resiliency. In open-trench soil-bentonite (SB) cutoff walls, even though backfill samples meet the specified hydraulic conductivity criteria, the potential for long-term internal erosion of the SB cutoff wall materials into a filter-incompatible aquifer or plucking remains high, if the binding materials are erodible fines. While relaying on an Atterberg limits-based criterion of backfill materials is not a practical measure, requiring a minimum bentonite content is a reasonable approach to ensure an SB cutoff wall indeed has bentonite as the binding agent to ensure long-term resilience of the wall. Improvements have been made to specifications to incorporate lessons learned from recent investigations of past projects, current construction practices, and past performance that may improve long term performance.