SEALABLE JOINT STEEL SHEET PILING FOR GROUNDWATER CONTROL AND REMEDIATION: CASE HISTORIES

David Smyth¹, Robin Jowett² and Murray Gamble³

Abstract

The Waterloo Barrier^TM steel sheet piling (patents pending) incorporates a cavity at each interlocking joint that is flushed clean and injected with sealant after the piles have been driven into the ground to form a vertical cutoff wall. The installation and sealing procedures allow for a high degree of quality assurance and control. Bulk wall hydraulic conductivities of 10^-8 to 10-¹⁰ cm/sec have been demonstrated at field installations.

Recent case histories are presented in which Waterloo Barrier^TM cutoff walls are used to prevent off-site migration of contaminated groundwater or soil gases to adjacent property and waterways. Full enclosures to isolate DNAPL source zones or portions of contaminated aquifers for pilot-scale remediation testing will also be described. Monitoring data will be used to demonstrate the effectiveness of the Waterloo Barrier^TM in these applications.

Introduction

Groundwater remediation most commonly involves groundwater extraction by wells or drains, with subsequent treatment of the contaminated water at surface. This pump-and-treat approach can be effective for the control and containment of groundwater but, as indicated by Mackay and Cherry (1989), it generally requires long-term operation and is seldom effective for restoring aquifers in contaminant source zones.

With the recognition of the limitations and inefficiencies of pump-and-treat, there have been strong incentives to develop alternate approaches and new technologies for groundwater remediation. Mackay et al. (1993) and Cherry et al. (1996) describe various methods by which contaminant source zone isolation and plume containment might be achieved. As shown in Figure 1, these methods include source zone isolation using low permeability cutoff walls, long-term hydraulic control of the plume emanating from the source zone using an active pump-and-treat system, or in situ treatment of contaminants emanating from the source zone using permeable reactive walls or Funnel-and-Gate^TM systems (patented). Hydrogeological conditions and contaminant characteristics will govern the feasibility of these approaches at a given site, but the potential for the use of vertical barriers in source zone isolation and plume containment is apparent.

Renewed interest in recent years in the development and application of barrier wall technologies has been described by Mutch et al. (1994). This has led to advancements such as improved hydraulic performance, enhanced installation methods and extended capabilities at depth or in difficult soil conditions. Given the variety of barrier technologies and the construction techniques available, it is reasonable to assume that there will be technical and cost advantages of using particular barrier walls in different situations. The remainder of this paper describes the development, testing and application of sealable joint steel sheet piling (Waterloo Barrier^TM) for cutoff wall construction.

Waterloo Barrier^TM

In the late 1980’s researchers at the University of Waterloo (UW) required a secure means of isolating portions of an uncontaminated sand aquifer for experimentation involving the controlled release of DNAPL chemicals. Several conventional cutoff wall technologies were investigated and found to be prohibitively expensive for the small scale, closed cell installations or insufficiently watertight. Thus, other construction options for the test cells were explored.

Steel sheet piling has seldom been used in environmental applications due to an unacceptable amount of leakage through the interlocking joints (McMahon et al., 1995). A method of sealing the joints of conventional steel sheet piling to limit the leakage was devised and the construction of a number of test cells using this prototype version of the Waterloo Barrier^TM proceeded at Canadian Forces Base Borden in Ontario.

As shown in Figure 2, the Waterloo Barrier^TM consists of conventional steel sheet piling with a modified interlock. A sealable cavity is incorporated into the interlock between adjacent sheet piles as the sheet is manufactured. Product specifications are presented in Table 1. The barrier can be installed using conventional pile driving techniques. A foot-plate at the base of the cavity reduces the build-up of compacted soil and the entry of obstructions during driving. After driving, the entire length of each cavity is jetted clean using pressured water or air. The integrity of each cavity can be evaluated for imperfections or blockages using the jetting hose, or using more sophisticated techniques such as downhole fibre optic video equipment. Following the cleaning and inspection of the cavities, the low permeability sealant can be emplaced from bottom to top in each cavity.

^*Available June, 1997

Since 1989, more than 20 test cells have been installed by UW for field research purposes at Borden and another site in southwestern Ontario. These have ranged in dimensions from 1 by 3 to 9 by 9 m, and have extended to depths ranging from approximately 3 to 15 m. Several of these cells have been constructed with concentric double walls, a configuration which facilitates rigorous hydraulic testing. Figure 3, after Starr et al. (1992), shows a schematic diagram of such a double-walled cell in which a hydraulic test was undertaken. The cell extends to a depth of 14.7 m through a surficial sandy aquifer into an underlying aquitard. The sealable cavities were injected with a bentonite slurry. For the test, the water level in the moat bounded by the two walls was maintained at a constant level. At the start of the test, the water level in the internal cell was raised by approximately 1 m relative to the natural water table in the vicinity of the cell. As the test proceeded, the decline in water level in the internal cell was monitored with time. Corrections were made to these levels to account for losses by evaporation.

In applying an analytical solution to assist in interpretation of the data, it was assumed that the underlying aquitard was impermeable and that all leakage from the internal cell occurred laterally through the barrier wall. In reality, some vertical leakage into the underlying aquitard would have occurred, so this assumption would result in an overestimation of the hydraulic conductivity of the barrier. As shown in Figure 4, the bulk hydraulic conductivity of the cell wall was calculated to be 6 x 10^-9 cm/sec. Similar tests in other cells, including those sealed with organic polymers, resulted in bulk hydraulic conductivities ranging from 10^-8 to 10^-10 cm/sec.

Commercial Applications

The Waterloo Barrier^TM became available commercially in late 1993 and has been used to provide subsurface containment and control at 18 sites across North America. With application in a wide variety of site conditions, it has been necessary to develop a number of joint sealants to meet project requirements. Issues that may be considered in sealant selection include sealant/contaminant compatibility, the presence of unusual groundwater chemistry such as high salt content, the ability of the sealant to withstand anticipated differences in hydraulic head across the barrier, the effects of wet/dry and freeze/thaw cycles on grout integrity, the removability of the sealant for temporary installations, permeability characteristics, pumpability and thermal expansion characteristics, design life of the system and cost. The types of sealants available include clay-based grouts such as bentonite and attapulgite, cement-based grouts modified with expanding agents, epoxy polymers, urethane polymers, and mechanical inflatable packers.

The ability to document quality assurance and quality control (QA/QC) during the construction process is a desirable attribute of cutoff walls used for environmental applications. With commercialization, QA/QC was another aspect of the Waterloo Barrier^TM technology that needed to be further developed. As potential leakage through the wall is limited to the sealed joints, QA/QC procedures can be focused on joint integrity prior to sealing and the sealing operation itself. Vertical alignment of piles monitored during driving, and the flushing and probing of the sealable cavities provides documentation regarding the ability to inject sealant the full length of the cavity. Records of grout volume, pumping time and starting depth provide assurance that the entire cavity has been sealed.

Project costs are site specific and are dependant on many factors such as project size, location, profile thickness of the sheet pile, type of sealant used, installed depth and driving conditions. Overall costs including mobilization, materials, pile installation, joint flushing and sealing, and a QA/QC program and report generally range from $160.00 US to $ 270 US per square metre ($15.00 US to $ 25.00 US per square foot) of barrier wall.

Waterloo Barrier^TM projects undertaken to date have varied considerably with respect to purpose, size, geological conditions and special requirements, but the barrier has proven to be a robust and versatile system for groundwater pollution control. It has been used to control the flow of contaminant plumes to enhance in situ or pump-and-treat remedial measures, to isolate source zones of contamination in the subsurface and adjacent to waterways, to isolate zones of contamination in soils where dewatering and excavation operations were undertaken, and to improve the performance and efficiency of a soil vapor recovery system in the unsaturated zone adjacent to a landfill. In the following section, representative projects are described in more detail.

Source Zone Isolation, Industrial Facility, Washington

At a large industrial facility in Washington, approximately 5,500 sq. m of Waterloo Barrier^TM was installed in three enclosures around zones of known subsurface contamination. The barrier was driven to depths ranging from 8 to 15 m in soils that included fill materials, sand, silt and clay with some gravel and organic matter. Installation procedures were disrupted by the occurrence of buried wooden pilings requiring some pre-drilling to facilitate driving of the steel sheeting. A second complication of the installation process, which was overcome by the flexibility of the Waterloo Barrier^TM system, was the need to align portions of the cells around existing utilities and beneath parts of operating buildings. The latter entailed barrier installation through the floors of buildings with limited overhead clearance. An attapulgite-cement grout was used to seal the interlock cavities.

Chemical monitoring conducted over the two year period since the barrier was installed has failed to detect the presence of contaminants outside the enclosures. Although the implementation of the ultimate corrective actions at the site has yet to occur, future options include enhanced isolation of the source zone areas within the cells with active pumping, or the removal of a section of barrier and its replacement with a permeable reaction curtain to treat in situ the controlled discharge from the source zone area.

Venus Mine Site, Yukon Territory, Canada

A barrier wall was constructed for the Department of Indian and Northern Affairs to isolate waste rock and tailings at a former mine site from an adjacent surface water system . The waste area was largely bounded by bedrock and natural soils. The barrier was installed to replace a dyke of natural soils along a portion of the perimeter. The barrier was approximately 250 m in length, extending to a depth of about 7.5 m. The dyke and overburden material was primarily a silty clay. A cementitious grout was used for sealing the cavities. Subsurface soil conditions were favorable for sheet pile installation, but unusual aspects of the project were the remote location of the site and associated issues pertaining to mobilization, the influence of the cold climate and temperature variations, the high dissolved solids content of the groundwater, and the sealing of the barrier to the bedrock surface.

The general purpose of the barrier was to reduce groundwater flow and the associated subsurface transport of dissolved contaminants including arsenic and zinc derived from the weathering of the waste rock and tailings from the impoundment towards an adjacent lake. A second function of the barrier was to enhance the stability of the dyke and reduce the surface erosion of the tailings. Additional features of the remedial scheme include a cap to limit infiltration of precipitation and erosion of the impoundment surface, and a decant system for surface water drainage.

The remedial system was installed in September 1995. Preliminary findings of the surface water monitoring program indicate a reduction in the level of contaminants of 80% for arsenic and a 50% for zinc. Monitoring over a longer term will be required to determine the full effect of the barrier wall and capping system on the surface water quality in the lake.

Former Kitchener Landfill, Kitchener, Ontario

In the mid 1970’s, methane gas was identified on the residential properties adjacent to a former landfill site. At that time, some gas control systems were installed at the site perimeter. In the late 1980’s a decision was made to upgrade the facilities to improve the effectiveness of methane gas containment. Site investigation determined that a perched water table in the landfill would interfere with conventional vertical extraction wells. A Waterloo Barrier^TM cutoff wall was incorporated into the design to provide a barrier to both water and landfill gas, and to direct the gas to horizontal collection lines located above the water level.

In June 1995 approximately 3,125 sq. m of Waterloo Barrier^TM was installed in two segments along the boundaries of the landfill that adjoined residential neighbourhoods. The barrier was driven to depths ranging from 4.5 to 9.5 m and extended through the unsaturated zone to beneath the seasonally low water table. Soil conditions consisted of fine sand to silt, but excavation of buried refuse was required at several locations along the wall path. Barrier installation was achieved using a crane-mounted vibratory hammer. The crane was operated from the margins of the landfill mound. Much of the barrier alignment was on the slope created by the landfill mound and engineered cover. Access to the toe of the slope was restricted in the vicinity of the residential neighbourhoods. The sealant used for the interlocks was a cementitious-based grout.

Since installation of the barrier wall, a vacuum has been maintained above the water level, indicating that the gas from this area is being collected and that the Waterloo Barrier^TM is functioning as designed.

Lowry Air Force Base, Colorado

In November, 1995 a pilot scale Funnel-and-Gate^TM system was installed in the path of a TCE plume at Lowry Air Force Base. Waterloo Barrier^TM was utilized to construct the rectangular treatment gate and the low permeability funnel that channels contaminated groundwater into the subsurface treatment zone. The system extended through a fine sand aquifer to the bedrock surface at a depth of about 5.5 m. A cementitious grout modified with silica fume was used to seal the interlocks of the funnel and the two connecting walls of the treatment gate. Up and down gradient sections of the gate were sealed temporarily with a prefabricated mechanical packer system consisting of an injectable grout line and a rubber bladder that was pressurized by air. This temporary seal was maintained while soil in the gate was excavated and the reactive media emplaced.

The gate was excavated to within 10 cm from the base of the piles and remained fully open for a period of at least 12 hours. During this time there was no visible leakage of groundwater water through the sealed joints. A small volume of water up to several cm deep collected on the floor of the excavation, but it appeared that the source of this infiltration was from below. A hydraulic head differential of about 3 m existed between the outside water table and the base of the excavation.

Once the gate was filled with reactive media the temporary mechanical seals were released and the up and down gradient sections of barrier wall were removed thereby opening the gate to allow the flow of contaminated water. The versatility of the Waterloo Barrier^TM in constructing irregular layouts on both large and small scale, its ability to form a watertight seal between the funnel and the treatment gate components, and its potential for removability make it particularly useful in Funnel-and-Gate^TM construction.

Advantages and Limitations of Waterloo Barrier^TM

Field applications have confirmed several advantages of the Waterloo Barrier^TM system. Installation is clean and rapid with minimal waste generation. The system is flexible and can be installed to satisfy customized layouts; for example, the alignment and installation procedures can be modified to accommodate existing infrastructure and natural features. Further, the barrier system, depending on selection of sealant, can be compatible with a wide range of contaminant conditions. QA/QC documentation of driving, joint flushing and sealing operations may be an advantage in complying with regulatory requirements. The barrier can function as a structural wall facilitating source zone removal by excavation. Hydraulic performance and integrity of the barrier system can be predicted based on the joint inspection and sealing documentation, and the hydraulic test results of numerous enclosures.

There are applications for which the Waterloo Barrier^TM may not be appropriate. The vibration and noise associated with pile driving equipment may be a problem in densely populated areas. Hydraulic pile driving equipment, however, can be used to reduce vibration and noise at some sites. Waterloo Barrier^TM installation is subject to the same limitations in terms of soils types and attainable depths as conventional steel sheet piling. In bouldery terrain and very dense unconsolidated sediments, the use of sheet piling will not be possible. Steel sheet pile applications are generally restricted to depths of less than 30 m or so. Although installation capabilities for sheet piling at depth can be enhanced by techniques such as the use of water jets at the leading edge of the pile as driving occurs, or by pre-drilling along the wall path, these will add to project costs. At some sites it is necessary to seal the barrier system to bedrock. Although techniques have been developed for sealing the base of Waterloo Barrier^TM to underlying rock formations, special precautions will be necessary, and effectiveness of the sealing techniques may be difficult to confirm.

The Waterloo Barrier^TM technology has been available for less than three years and although development, field testing and applications to date have demonstrated its effectiveness in the control and containment of subsurface contamination, it is anticipated that the full capabilities including the advantages and limitations of the technology will become more clear as more projects are implemented.

References

Cherry, J.A. , S. Feenstra and D.M. Mackay. (1996) Concepts for the remediation of sites contaminated with dense, non-aqueous phase liquids (DNAPLs), In Dense Chlorinated Solvents and Other DNAPLs in Groundwater, (ed. J.F. Pankow and J.A. Cherry), pp. 475-512. Waterloo Press, Portland, Oregon.

Mackay, D.M., and J.A. Cherry. (1989) Groundwater contamination: Limitations of pump-and-treat remediation. Environmental Science and Technology, 23 (16), 630-636.

Mackay, D.M., S. Feenstra and J.A. Cherry. (1993) Alternative goals and approaches for groundwater remediation. In Proceedings of Workshop on Contaminated Soils: Risks and Remedies, pp. 35-47, Stockholm, Sweden.

McMahon, D.R. et al. (1995) Vertical Barriers: Sheet Piles. In Assessment of Barrier Containment Technologies, International Containment Technology Workshop, Baltimore, Maryland (ed. R.R. Rumer and J.K. Mitchell), pp.77-93

Mutch, R.D., R.E. Ash and N.J. Cavalli. (1994) Advancements in subsurface barrier wall technology, in Superfund XV, pp. 784-789, Washington.

Starr, R.C., J.A. Cherry and E.S. Vales. (1992) A new type of steel sheet piling with sealed joints for groundwater pollution control, In Proceedings of the 45th Canadian Geochemical Conference, pp. 75-1 to 75-9, Toronto.

Acknowledgements

The authors acknowledge the contributions of John Cherry (UW), Sam Vales (UW), Robert Starr (UW), Jack Hammill (Canadian Metal Rolling Mills) and Cam Wood and co workers at C3 Environmental to the development of Waterloo Barrier^TM technology. Research funds have been provided to UW by the University Consortium Solvents-in-Groundwater Research Program, which has been sponsored by The Boeing Co., Ciba-Geigy, Dow, Eastman Kodak, General Electric, Laidlaw, Mitretek Systems, Motorola, PPG Industries, United Technologies Corporation, and Canadian (NSERC) and Ontario (URIF) governments; and the Ontario Environmental Technologies Program. The support of site owners and consultants has been valuable in the technology transfer and acceptance of the Waterloo Barrier^TM.

1. Department of Earth Sciences, University of Waterloo, Waterloo, ON, Canada N2L 3G1
2. Waterloo Barrier Inc., P.O. Box 385, Rockwood, ON, Canada N0B 2K0
3. C3 Environmental, P.O. Box 188, Breslau, ON, Canada N0B 1M0