Geomembrane sealing systems for dams

By Robert Koerner and John Wilkes

Abstract

The 2006 ICOLD Bulletin on Geomembrane Sealing System for Dams supersedes and greatly expands upon two earlier versions, published in 1981 and 1991. The 250 dams incorporating geomembranes cited in this new Bulletin include: 174 fill, 43 concrete, 32 roller compacted concrete, and 1 hybrid. Of the total, 115 (46%) are in Europe, 47 (19%) are in China, 38 (15%) are in the United States, and the remaining 50 (20%) are scattered in other countries and locations. Since most of the experience gained to date is in Europe, the European Working Group on Geomembranes prepared the Bulletin.

This article attempts to summarize the 300-page report and is subdivided according to the same chapter structure contained in the Bulletin. Since this Bulletin is a status report of the current state of the practice, there are no recommendations, per se, those being left to readers for their own conclusions.

Introduction

For more than 45 years, geomembranes have been used to provide waterproofing on various types of large dams. Table 1 provides information on the 250 dams cited and detailed in this Bulletin. Most have been installed in the dry, but some success at underwater sealing has also been achieved.

By far the largest proportion of these dams have used factory-manufactured polymeric geomembranes. When such geomembranes incorporate other types of geosynthetic materials, such as geonet drains and geotextile cushions, filters, or separators, a “geomembrane sealing system (GSS)” is achieved and will be referenced as such throughout the paper. As just stated, the waterproofing geomembranes used in these dams are generally polymeric, although 22 are bituminous, and 4 are in-situ types. Since in-situ types are no longer used and bituminous types are currently quite rare, the focus in the Bulletin is almost entirely on polymeric geomembranes made from various types of resins and their related formulations.

Materials, testing, and durability

Most of the geomembranes cited in this Bulletin are made from thermoplastic polymers (those that can be thermally welded in the field) with polyvinyl chloride (PVC) being most widely used. Table 2 presents these statistics. It should be recognized, however, that all polymers are actually formulations containing the designated resin (from which the name is derived), additives (mainly antioxidants), colorants (often carbon black), and some fillers. The geosynthetic literature is abundant on details of manufacturing quality control and manufacturing quality assurance. The Bulletin provides a detailed comparative behavior of the various geomembranes. Also seen in Table 2 is that both exposed and covered situations are encountered and in approximately equal proportions.

Testing of the manufactured geomembrane is covered in the Bulletin in the context of providing the desired information for various situations. Both ISO and ASTM standards are identified. Individual tests for the following situations are described:

• quality control during manufacture
• identification testing
• performance testing
• compliance testing

Regarding the important issue of geomembrane aging and its in-service durability, a critical issue is whether the geomembrane is exposed or covered with soil, rock, or concrete. The former is much more critical due to ultraviolet exposure and the usually high accompanying temperatures. Discussion on this topic is addressed in the Bulletin. In this context, however, it should be noted that the oldest exposed geomembrane installation in 1974, and covered in 1960, strongly suggests that the proper resins and formulations are currently available in the context of anticipated dam lifetimes.

Space in this paper precludes further discussion; however, the geosynthetics literature is abundant on this subject.

Loads applied to geomembrane sealing systems (GSS)

Table 3 illustrates various dam configurations and their respective GSS locations for 236 of the dams cited; the remaining 14 are special cases. As can be envisioned, mechanical loads are certainly to be anticipated. Addressed in the Bulletin are the following:

• gravity load
• subgrade differential settlement
• puncture loads
• wind uplift
• reservoir waves
• ice in the reservoir
• uplift from water and air

Also addressed are various physical, chemical, and biological agents that can be important on a case-by-case basis. They include the following:

• heat
• ultraviolet radiation
• water ingredients
• biological activity (e.g., microorganisms)
• vegetation
• fauna
• vandalism

Each of these topics are discussed in the Bulletin, but to a limited extent since each is a topic within itself.

Geomembranes for new and rehabilitated fill dams

Table 4 indicates that both new and rehabilitated fill dams (earth and earth/rock types) are commonly waterproofed using geomembranes on or in their upstream slopes. This is the most widely used GSS application reported in the Bulletin. When used on upstream slope of fill dams, cover is readily achievable usually with a geotextile cushion (typically a thick needlepunched nonwoven fabric) and then rock riprap or articulated concrete block mattresses on the surface. One hundred (100) cases are of this general type.

As seen in the table, a relatively high number of fill dams use GSS without being covered (44 cases), i.e., in an exposed condition. Confidence in geomembrane durability appears to be a factor as many of these dams are recent case histories. Of importance in this regard is that the geomembrane can be visually inspected as to its performance whenever the water level is lowered.

Internal GSS have been used in 18 large fill dams, most of them in China. The Bulletin gives numerous placement options, which are all cost effective in comparison to conventional core walls made from clay or asphaltic concrete. The internal GSS is also applicable to raising the height of existing dams.

In essentially all cases of geomembranes used with fill dams, a drainage layer is used behind the geomembrane, thereby creating a “system.” This drainage layer is often a geonet composite with two geotextiles thermally bonded to it on both sides. This allows easy placement on a steep slope or even vertically. Of course, interface stability is an important issue and direct shear testing is required. The inclined drainage system empties into a horizontal drainage gallery beneath the embankment and is discharged at the toe of the downstream slope. This drainage gallery is typically granular soil of adequate thickness and permeability so as not to mobilize excess pore water pressures beneath the downstream embankment portion of the dam. Many configurations are illustrated in the Bulletin.

Anchorage of the geomembrane is emphasized in a number of sections of the Bulletin. Clearly, at the 2 abutments and the foundation, leakage must be minimized. Concrete beams, or “plinths,” are common in this regard. These plinths are cast with connections of various types integral with their construction. They can be continuous (with the geomembrane welded to polymer inserts) or discrete (with anchor bolts, batten strips, and associated hardware to physically fasten the geomembranes). These are obviously critical design details.

It should be noted that if the strategy is for exposed geomembranes on slopes, wind and wave forces can be significant in shifting the geomembrane out of position. This will require intermediate anchorages throughout the slope. Several scenarios are presented in the Bulletin.

Lastly, the seaming of the geomembrane rolls together on the slope is arguably the most conventional of the design/construction issues. With thermoplastic geomembranes of the type presently used, welding of edges and ends of the rolls is straightforward. Hot-air welding can be done at any angle (even vertically) and can even be done with two parallel tracks, leaving an unbonded space between. This so-called “dual channel” seam allows for air inflation and is an excellent nondestructive test of the completed weld. If flaws are detected, conventional patches or cap strips are used and retested using the vacuum box method.

Geomembranes on concrete, masonry dams

Table 5 presents statistics on the use of geomembranes in the rehabilitation of poured concrete and masonry dams. (Roller-compacted concrete dams will be treated separately). In all cases except one, the geomembranes are exposed. It is also current practice always to include drainage (and cushioning) behind the geomembrane thus creating a GSS. There are no instances of using GSSs in newly designed concrete dams.

In most of the cases, a grout curtain is first constructed to minimize foundation leakage. A concrete beam (or plinth) is constructed over the grout curtain, which also serves to anchor the base of the geomembrane to provide proper waterproofing. A drainage system (very thick geotextile, geonet, geonet/geotextile composite, or geotextile/ geonet/ geotextile composite) is always used on the face of the dam to collect any seepage that bypasses the geomembrane. This drainage system leads to the dam’s existing drainage gallery, where it is collected, monitored, and released downstream. Many variations are illustrated in the Bulletin. The drainage system also serves as an antipuncturing layer.

Regarding attachment to the face of the dam, experience over the years has lead to linear, vertical anchorages of wide geomembrane panels being the preferred strategy. The spacing between vertical anchorages (typically 3-5m) and the type of anchorage system must leave the geomembrane panels adequately smooth and yet properly tensioned. Aesthetics are also a consideration, and black geomembranes are never used, whereas any color (often gray) can be formulated, providing that proper ultraviolet durability is achieved. The most common loads to be resisted by the exposed geomembrane panels are wind and uplift.

Geomembranes for roller-compacted concrete (RCC) dams

As illustrated in Table 6, there are 32 RCC dams which incorporate a geomembrane as the watertight element. Most cover the entire face of the dam; however, 5 of them cover the joints and cracks only.

As seen in Table 6, approximately half of the geomembranes are exposed and the other half covered. The exposed solution has many similarities to the concrete and masonry dam rehabilitation schemes presented in the previous section.

One such scheme is the Sibelon- CARPI System. The covered solution (known as the Winchester System) uses a composite geomembrane/geotextile, which is prefabricated onto a concrete panel. This panel is then used as the upstream forming system of the RCC dam, with the geomembrane facing the concrete as it is being placed. This allows the preformed concrete to act directly against the impounded reservoir and thus protects the geomembrane against ultraviolet exposure, puncture damage, and vandalism. It has been used on 10 dams in the United States.

Special cases

Special cases include waterproofing of joints on concrete dam faces for both new and rehabilitation situations. They sometimes are constructed underwater. The Bulletin presents Table 7 in this regard.

Important in regard to these special cases is the use of water stops (both embedded and external) in RCC dams. The geomembrane used must have excellent elasticity and flexibility, while also having the requisite durability. Good tear and puncture resistance is also required, and sometimes burst resistance as well.

Quality control (QC) and quality assurance (QA)

While proper QC and QA are obviously important, a separate chapter in the Bulletin was felt to be necessary in this regard. Furthermore, in dealing with geosynthetic materials, there are manufacturing issues as well as construction issues. In all, the 4 following interconnected quality issues, when properly practiced, lead to an acceptable project:

• MQC = manufacturing quality control at the factory
• MQA = manufacturing quality assurance at the factory
• CQC = construction quality control in the field
• CQA = construction quality assurance in the field

In this regard, one has ISO 9000 certification for the manufacturing considerations (both MQC and MQA); and certification for both installers (for CQC) and inspectors (for CQA) is available for field personnel.

The Bulletin is quite detailed in giving guidance to the type and number of tests for MQA and CQA personnel to require. Some generic specifications for different types of geomembranes are currently available. Installation of the geomembrane sheets/panels as well as fastening systems are described. Space in this paper precludes going into detail in this regard.

Regarding contracts, it is recommended to separate the waterproofing contract from the earthwork or concrete construction contract. GSS waterproofing projects require completely different work personnel and skill sets from the conventional dam contractors.

While there is a considerable amount of important text in the Bulletin in this chapter, a few points bear noting:

• A quality control plan must be submitted along with the bid documents.
• It is common to specify the total acceptable leakage from the upstream face or from each compartment.
• It is typical to require a warranty on materials and a warranty on installation.
• Quality, not cost, should be the major consideration in the selection process of a contractor supplying a GSS.

Robert M. Koerner, Ph.D., P.E. NAE, is emeritus professor of civil engineering at Drexel University and director of the Geosynthetic Institute, 475 Kedron Avenue, Folsom, PA 19033-1208. He is a member of Geosynthetics magazine’s Editorial Advisory Committee. John A. Wilkes, P.E., is president, CARPI-USA Inc., 2706 Ogden Road, Suite 3, Roanoke, VA 24014. This article is an edited, magazine version of the Koerner/Wilkes paper, which was presented at the USSD conference in March 2007. It appears in Proceedings—United States Society on Dams, 27th Annual Conference, Philadelphia, March 5–9, 2007, pp. 69–78.

Acknowledgements

The 2006 ICOLD Bulletin on Geomembrane Sealing Systems for Dams was prepared by the European Working Group, consisting of the following members: E. Aguiar Gonzalez (Balsas de Tenerife, Spain), P. Barkek (Swiss National Committee), M. Blanco Fernandez (Laboratorio Central De Estructuras Y Materials C.D.E.X., Spain), P. Brezina (Povodi Odry, Czech Republic), H. Brunold (Austrian National Committee), D. Cazzuffi (ENEL CESI, Italy), H. Girard (Cemagref, France), M. Lefranc (French National Committee), J. L. Machado do Vale (Portuguese National Committee), C. Massaro (Azienda Energetica Metropolitana Torino), J. Millmore (British National Committee), L. Schewe (German National Committee), A. Scuero (Italian National Committee), P. Sembenelli (Italian National Committee), G. Vaschetti (Italian National Committee), with the assistance of R. M. Koerner (Drexel University/ GSI, USA).