The International Information Center for Geotechnical Engineers

Vertical Impermeable Barriers (Cutoff Walls)



4.            Case Histories

4.1.      Geomembrane Wall

Table 14 briefly summarizes several case studies involving the use of geomembranes in vertical barriers. Among the cases listed is a cutoff wall encompassing a Great Lakes chemical plant to prevent the spread of contaminated groundwater, previously described by Burnette and Schmednecht (1994). The wall was constructed through layers of sand, gravel and cobbles by means of the vibrating beam installation method. The 4 ft wide, 80 mil thick, HDPE panels were inserted into a 5 inch wide trench supported by a clay/cement slurry which was up to 35 ft in depth. The average rate of production for this project exceeded 1100 square feet per day.

Another case history summarized in Table 13 consisted of a 21-mile geomembrane wall constructed to contain drilling fluids in Alaska’s North Slope oil fields. The wall primarily consisted of a double lined system consisting of both HDPE, reinforced polyurethane and reinforced PVC geomembranes. The geomembrane was rolled out in a 12 inch wide trench with the roll ends seamed by thermal fusion. The wall, typically 10 ft in depth, was keyed in to the permafrost. The wall was constructed in temperatures reaching as low as -44°C and production rates were typically 500 ft for a 10 hr sheet.


Table 13:  Summary of Case Studies Utilizing Geomembranes as Cutoff Walls (after Koerner & Guglielmetti, 1995 as presented by Thomas & Koerner, 1996)

table 13


4.2.      Sheet Pile Wall/Drain Combination

In the early 80’s McDowell & Associates designed a containment system which incorporated the combination of a gravity drain and a downgradient sheet pile wall to stop the spread of contaminants into an adjacent river. For most of the length of the system, the sheet pile wall was keyed into a low permeability silty clay but deep sand deposits, believed to be from a past river channel, were occasionally encountered where the sheet pile wall was left hanging. A study which involved the measurement over time of both the river level and the readings from several piezometers was conducted and it was concluded that there was some potential for seepage into the river. To overcome this potential seepage, pumping wells were installed in specified locations to further lower the water table of the contaminated site. In this project the cutoff wall was primarily used to manipulate the groundwater flow rather than for direct containment. The flownet shown in Figure 14 was constructed to show an area were there was no potential for contaminant seepage to the river in a zone where the sheet pile wall was not keyed into the silty clay. (McDowell, 1983)

figure 13

Figure 14: Flownet for Containment System Designed by McDowell & Associates (McDowell, 1983)


4.3.      Slury Wall Common Mistakes

The following brief case histories taken from Benson (2002) are examples of some common ways a cutoff wall can be insufficient. The first case consisted of a soil bentonite wall with a hydraulic conductivity of 5x10-9 m/s utilized to isolate a lagoon from the surrounding groundwater in the western United States. The wall was to be keyed into the underlying bedrock. The bedrock consisted of claystone which was overlain by a relatively thin sandstone layer. Excessive leakage quickly became apparent and pump tests revealed that the 1.9 km wall was leaking approximately 1000 m3/day, which was 100 times the expected value. Further investigation revealed that 48% of the wall was keyed into the relatively pervious sandstone instead of the relatively impervious claystone. The reasons for the lack of a proper key came down to lack of proper site investigation prior to construction and the fact that the engineer for the job was only on site for the first day. A previous study by Techavises (1998) described the effectiveness of a cutoff wall for varying gaps between the wall and the “impervious layer”,the relationship can be seen in Figure 15.

figure 14

Figure 15: Wall Effectiveness as a Function of Gap Size between the Wall and the “impervious layer” (From Tachavises, 1998 as presented by Benson, 2002)


The next case consisted of a soil bentonite cut-off wall to isolate a plume emanating from an uncovered landfill. The stratigraphy consisted of approximately 16 m of granular material underlain by 10 m of thick clay till.  The wall was constructed with conventional slurry wall methods and the backfill was mixed using the soil excavated from the trench. The backfill was to have a hydraulic conductivity greater than 1x10-9 m/s, fines content greater than 30% sand content less than 15%. During construction sand contents reached as high as 41% despite specifications and warnings from the engineer. When the engineer would not certify the wall, a drilling program was conducted to verify the integrity of the wall. The program discovered the presence of windows with permeability’s as low as 2x10-6 throughout the wall where sand content exceeded 20%. Because the engineer would not sign off on the wall, the contractor was forced to fix the wall through the process of deep soil mixing. Another study by Techavises (1998) analyzed the effectiveness of a slurry cutoff wall with high permeability windows, the results can be seen in Figure 16.

figure 15

Figure 16: Effect of High Permeability Windows on Wall Performance (From Tachavises, 1998 as presented by Benson, 2002)


The third case consisted of a project where segments of a soil bentonite slurry wall were being replaced by a cement bentonite wall that had to have a hydraulic conductivity less than 1x10-8 and compressive strengths exceeding 175 kPa at a hazardous containment site. The reason for the replacement was that the liquid contaminants were apparently too aggressive for the soil bentonite wall. Sampling of the wall was conducted by making cylinder molds to be tested after seven days and at a later date. The first set of samples did not meet the strength requirements so core samples were taken to determine if further curing strengthened the wall. Testing showed that neither the hydraulic conductivity nor the strength was adequate.  Further core samples were taken with even higher hydraulic conductivities and lower strengths and a test pit was eventually dug. It was then suggested that disturbance to the samples may have contributed to the unacceptable test values.  Further testing on the cylinder samples which had been held to test at a later date affirmed that the previous samples were most likely disturbed, the test results were all less than the specified hydraulic conductivity and greater than the specified strength. In this case a large amount of time effort and money was spent due to inadequate sampling methods.



Add comment

NOTE: The symbol < is not allowed in comments. If you use it, the comment will not be published correctly.

Security code
*Please insert the above-shown characters in the field below.

The Corporate Sponsors: