Return to Bulletin Contents.| Waterways Experiment Station | Structures Laboratory |
The U.S. Army Corps of Engineers operates and maintains a wide variety of hydraulic structures, including mass-concrete gravity dams, rock-fill dams with concrete facings, and roller-compacted concrete dams. Concrete appurtenances associated with such dams include intake towers, outlet works, and stilling basins. Located at over 600 project sites throughout the United States, these structures are subjected to a wide spectrum of environmental conditions. Also, the advanced ages of these structures, more than 40 percent of which are over 50 years old, increase the potential for concrete deterioration.
Many of these structures exhibit concrete cracking, which allows water intrusion into or through the structure. Water leakage through hydraulic structures can also result from poor concrete consolidation during construction, improperly prepared lift or construction joints, and water-stop failures. When leakage rates through cracked or deteriorated concrete and defective joints become unacceptable, repairs are made. Conventional repair methods generally consist of localized sealing of cracks and defective joints by cementitious and chemical grouting, epoxy injection, or surface treatments. Even though localized sealing of leaking cracks and defective joints with conventional methods has been successful in some applications, in many cases some type of overall repair is still required after a few years. Consequently, the potential for geomembranes in such repairs was evaluated as part of the Corps' Repair, Evaluation, Maintenance, and Rehabilitation (REMR) Research Program.
Various configurations of geomembranes have been used as impervious synthetic barriers in dams for more than 30 years. Generally, membranes are placed either within an embankment or rock-fill dam as part of the impervious core or at the upstream face of embankment, rock-fill, and concrete gravity dams. In recent years, geomembranes have been increasingly used for seepage control in a variety of civil engineering structures, including canals, reservoirs, storage basins, dams, and tunnels. Geomembranes have also been used successfully to resurface the upstream face of a number of old concrete and masonry dams, particularly in Europe.
A review of geomembrane applications (McDonald 1993) indicated that the success of these systems in arresting concrete deterioration and controlling leakage in dams, canals, reservoirs, and tunnels and the demonstrated durability of these materials are such that these systems are considered competitive with other repair alternatives. With a few exceptions, geomembrane installations to date have been accomplished in a dry environment by dewatering the structure on which the geomembrane is to be installed. Dewatering, however, can be extremely expensive and in many cases may not be possible because of project constraints. A durable geomembrane system that could be installed underwater to minimize or eliminate water intrusion and leakage would be an economical alternative for repair of a variety of hydraulic structures. Consequently, research was initiated to develop a procedure for underwater installation of geomembrane repair systems.
A two-phase contract to develop the system was awarded to Oceaneering International, Upper Marlboro, MD, and CARPI/USA, McMurray, PA, based on their respective expertise in underwater construction and geomembrane systems for dam rehabilitation. In Phase I, a conceptual design for the underwater repair system was developed based on research, material testing, and detailed evaluation of individual components and procedures. The constructibility of the design was demonstrated in Phase II through successful underwater installation of the system on a simulated concrete structure.
The objective of this phase of the study was to perform research, material testing, and evaluation of individual components and techniques required to facilitate successful underwater installation of membranes and to develop a procedure for underwater installation on the upstream face of a dam. Work in this phase included developing design criteria, surveying available materials, conducting material testing, and evaluating materials and assembly techniques. Material testing was conducted, when applicable, in accordance with standardized tests. However, other even more valuable information was collected with nonstandardized tests, namely with multiaxial, large-scale tests or tests that were intended to simulate conditions likely to be encountered during actual installation. Testing was conducted on drainage materials, membrane materials, anchorage profiles, gaskets, anchor bolts, and surface repair compounds.
Various types and thicknesses of geomembranes were tested to determine their conformability, burst resistance, and puncture resistance in the presence of a very rough substrate (Figure 1). Samples of membrane were placed in a pressure vessel that was sealed and pressurized to a maximum pressure of approximately 150 psi (1 MPa). Samples of membrane that did not rupture during pressurization were subjected to the maximum pressure for 24 hr. The specimens were then removed from the pressure chamber and inspected. A sample of reinforced polyvinyl chloride (PVC) after testing is shown in Figure 2. Obviously, the membrane conformed to the very irregular substrate without puncturing.
The mechanical fastening system that secures and seals the membrane system to the surface of the structure also received considerable attention in the design phase. The stainless steel profiles must be flexible enough to conform to the substrate, yet stiff enough to ensure continuous compression of the gasket without an excessive number of anchor bolts. The performance of both chemically grouted and mechanical anchors installed under submerged conditions was evaluated. A profile and gasket conformability test is shown in Figure 3. In this test, a 1-in. (25-mm) -thick, open-cell neoprene gasket is being compressed by a 1/4-in. (6-mm) -thick stainless-steel profile with anchor bolts on 12-in. (305-mm) centers.
The geomembrane system designed for underwater installation on the upstream face of a dam consists of a high-density polyethylene (HDPE) geonet drainage layer, and a PVC geomembrane backed with geotextile reinforcement, anchored and sealed around the perimeter and along vertical splices (Figure 4 and Figure 5). Development of the system is described in detail by Christensen et al. (1995).
A PVC geocomposite consisting of a geomembrane backed with nonwoven geotextile reinforcement was selected over the other available membrane materials because of its superior qualities with respect to constructibility, mechanical performance, durability, and prior use. HDPE geonet with preferential flow is a suitable drainage medium behind the membrane should a drained system be installed. The drained water can be discharged downstream through the structure or directly into the reservoir. Stainless-steel anchor bolts were selected to secure the perimeter profiles and vertical splice profiles to the concrete structure. Stainless-steel flat-bar profile sections with a minimum thickness of 1/4 in. (6 mm) were selected. Unless site-specific conditions dictate otherwise, the gasket should be open-cell neoprene, medium hardness, with a channel-shaped cross section.
The objective of this phase of the study was to demonstrate that the conceptual design could be practically installed underwater and that it provides a reliable barrier to moisture intrusion. The constructibility demonstration is described in detail by Marcy, Scuero, and Vaschetti (1996) and summarized in the following. The conceptual design and the constructibility demonstration are also summarized in a 9-min video report (REMR-CS-5).
The demonstration required a test structure that simulated a concrete hydraulic structure in need of repair. In an effort to make the constructibility demonstration comprehensive, the test structure was designed and built with features that replicate possible situations which could complicate the underwater installation of the geomembrane system. These features included rough surfaces, complex corners, depressions and protrusions, a V-shaped notch representing a construction joint, and various holes simulating discrete leakage points. The concrete structure was designed and constructed in the configuration of an L-shaped wall as shown in Figure 6.
A vacuum manifold was incorporated into the wall. The manifold creates a suction behind the membrane to simulate different hydrostatic heads and to test the efficiency of the system. The manifold is connected to 1-1/2 in. (33-mm) holes in the concrete which simulate points of discrete leakage through the structure.
After a successful installation in the dry (Figure 7), the wall was lifted with a 60-ton crane and lowered into the test tank to a depth of 20 ft (6.1 m). Multiple installations were performed underwater. The profiles were used as templates for the anchor-bolt holes. Holes were drilled with a hydraulic hammer drill, and the bolt holes were cleaned with water and a plastic brush. Three types of anchor bolts were installed: torque-set wedge bolts, chemical anchors which use a two-part epoxy, and chemical anchors which use two-part epoxy and a glass encapsulated resin cartridge. Underwater epoxy was applied to smooth the rough concrete at the perimeter. The geonet drainage layer was positioned and secured to the wall with small expansion anchors. The gasket was placed over the anchor bolts along the perimeter, and the membrane sections were rolled down the face of the wall. Bolt holes were punched in the membrane by tapping the membrane over the bolts with a hammer. A second gasket layer was placed between overlapping membrane sheets at the vertical splices and perimeter seal. The profiles were placed on the wall and the anchor bolts were torqued to 35 foot-pounds.
After all of the bolts were tightened, water was evacuated from behind the membrane using a hydraulic ejector. The combined effort of the pressure depression behind the membrane and the water depth resulted in a hydrostatic head of approximately 40 ft (12.2 m) of water. Two weeks after the vacuum was shut off, the membrane remained tightly conformed to the wall (Figure 8), indicating that seepage through the repair system was extremely slow. During one of the underwater installations, five anchor bolts that used a combination of two-part epoxy and a glass-encapsulated resin cartridge were used. These five bolts loosened as the nuts were tightened. Failure was later attributed to the installation technique.
The system was tested to determine the effect of the defective bolts. As the ejector evacuated water behind the membrane, the membrane conformed tightly against the wall. With the ejector shut off, the membrane remained tightly conformed for approximately 2 hr. With the suction reapplied, divers were able to locate a small leak near the defective bolts by injecting dye into the water near the bolts. The defective bolts were removed and replacement bolts were installed underwater. When the nuts were tightened, an efficient seal was achieved. This installation demonstrated that the system is repairable as well as constructible.
Results of the underwater installation dealt with two basic issues:
From the standpoint of installation feasibility, the underwater test demonstrated that ease of installation depended on the roughness of the substrate and the geometry of the structure. In rough areas, detailed procedures were required to ensure good perimeter sealing, while on fairly smooth surfaces, installation of all components was easily accomplished. Experience in the dry had already shown this, but environmental conditions underwater amplified the problems associated with difficult features. This test mirrored experience in dry installations and showed that additional care is required to ensure good perimeter sealing when installations are performed in the more challenging underwater environment.
The research team believed that particular geometries of the structures, such as the complex corners, should be treated with a prefabricated sheet. Such scenarios will have to be addressed for each installation. Structures with complex shapes, such as intake towers, may require prefabricated membrane pieces to reduce installation time. Protrusions and depressions may constitute a design issue if they are very sharp. Experience in the dry, however, has proved that such irregularities can be adequately addressed with additional transition layers of nonwoven, needle-punched geotextiles.
Testing the system revealed that seepage through the repaired area was very slow. Even where five adjacent anchor bolts failed, leakage was slow enough to make detection of the leak difficult to notice even when dye was injected at the point of leakage. Although the leakage rate was not measured, the research team believed that it was slow enough to be negligible with respect to the requirements of most concrete hydraulic structures. The use of a drained system helped to locate and rectify the leak.
The successful underwater installation of the membrane repair system demonstrated the feasibility of the system. Although results of the demonstration were more qualitative than quantitative, it is evident that the system is constructible and will perform acceptably when designed and installed correctly.
Compared to dewatering of a structure for repair, a geomembrane system that can be installed underwater minimizes the impact of the repair on project operations such as hydropower generation, and recreation. Also, the underwater repair system eliminates the potentially adverse environmental impacts associated with dewatering of many structures.
Pending availibiity of funding, current plans are to demonstrate the constructibility of the underwater repair system as a prototype structure. Candidate structures or appurtenances are being solicited. Anyone with a potential application for a repair of this type should contact Jim McDonald at (601) 634-3230.
Christensen, J. Chris, Marcy, Matthew M., Scuero, Alberto M., and Vaschetti, Gabriella L. (1995). "A Conceptual Design for Underwater Installation of Geomembrane Systems on Concrete Hydraulic Structures," Technical Report REMR-CS-50, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Marcy, Matthew M., Scuero, Alberto M., and Vaschetti, Gabriella L. (1996). "A Constructibility Demonstration of Geomembrane Systems Installed Underwater on Concrete Hydraulic Structures," Technical Report REMR-CS-51, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
McDonald, James E. (1993). "Geomembranes for repair of Concrete Hydraulic Structures," The REMR Bulletin, 10(4), 1-6, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Comments