For almost 40 years, the generally accepted destructive test rate for geomembrane seaming has been one sample for every 500 feet (152 m) of seam. On large projects this leads to hundreds of holes cut into liners that engineers, manufacturers and installers work very hard to make leak proof. With a typical failure rate for most installers in the 1%–2% range, why do we continue to cut and patch perfectly good seams? With modern materials and equipment, is there a way to limit destructive seam testing?
Geomembrane seaming destructive sample origins
In 1987, Thomas Wright authored the Manual of Procedures and Criteria for Inspecting the Installation of Flexible Membrane Liners in Hazardous Waste Facilities (Wright et al.) under contract with the Hazardous Waste Engineering Research Laboratory of the U.S. Environmental Protection Agency (EPA). This appears to be the origin of the one-seam sample in 500 feet (152 m) frequency. Subsequent articles on the topic refer to this work as the industry standard for installation practices and testing. Interestingly, the actual recommendation in the text refers to a minimum of one test per seaming crew per day and only taking destructive samples when there is an insufficient number of construction quality assurance (CQA) inspectors to observe each seaming crew fulltime or when the results of testing nondestructive samples indicate poor seam quality. A review of literature at the time points to using judgment and performance of the installation crew to base sampling frequency on a project-by-project basis. Wright’s work in 1987 only states that one test every 500 feet (152 m) of seam is “normally required” (Wright et al.) and is not an absolute.
Considering current-day practices, the industry, its products and its installation methods have matured. However, the concept of using judgment and adjusting seam sampling based on performance has been widely replaced by prescribed values.
A geomembrane installation cannot be completed without some number of repairs due to welder burnouts, “tee” seams or nondestructive test failures among other normal installation conditions. However, randomly cutting destructive seam samples based on a seaming distance may add unnecessary geomembrane holes and, if a deficiency is found, there is no data to indicate its extent. At best, a destructive sample/test passes and confirms that it was not needed in the first place. At worst, a seam is subjected to several additional destructive samples/tests to try to identify the extent of the seam failure.
Pilot project and field study evaluation background
Sequoia Services was retained by the owner as the general contractor for two projects that included geomembrane seaming. Wood Environment & Infrastructure Solutions, the design engineer and Construction Quality Assurance (CQA) firm for both projects, collaborated with the owner to determine how to implement new technologies to limit destructive testing for future, larger-scale projects. The team met with regulators, material manufacturers and installers to understand how to achieve this goal. The team recognized that monitoring and evaluating geomembrane seaming parameters using data acquisition welding machines would enable pinpointing seam defects. Furthermore, seam contamination could be reduced thereby improving seam quality by using taped geomembrane edges. The resulting goal was to evaluate these new technologies in practice aiming for destructive testing only as the data indicated a need.
Two pilot projects and field studies were conducted. The project-specific details are summarized in Table 1.
Project design and specifications
Data and quality-based geomembrane—seaming
Wood designed the projects, preparing construction drawings, technical specifications and the CQA plan. Technical specifications for the high-density polyethylene (HDPE) geomembranes included some key differences from a standard geomembrane specification, including requiring taped geomembrane edges to limit seam contamination and requiring the use of data acquisition welding (DAW) machines to continuously monitor double-track fusion seaming performance.
The technical specifications required fusion welding machines equipped with data acquisition capabilities to measure, record and display seaming temperature, speed and pressure as well as to display voltage. The technical specifications defined the required tolerances for each seaming parameter presented in Table 2.
Consistent with ordinary geomembrane installation, the geosynthetic installers established the target seaming temperature, speed and pressure appropriate for the geomembrane material and ambient weather conditions.
The technical specifications and CQA plan defined procedures for data/quality-based field seam evaluation instead of prescribing destructive seam sampling and testing at a frequency of one sample/test per 500 feet (152 m). The data/quality-based field seam evaluation requirements stated that:
Destructive test samples shall not be collected at a prescribed frequency.
Destructive test samples will be collected based on review of geomembrane seam DAW reports where temperature, speed and pressure values were outside of the defined tolerance.
The CQA engineer/personnel may elect to collect destructive test samples based on visual observation of seaming operations and/or observed seam quality.
The geosynthetics installer shall submit geomembrane DAW reports to CQA personnel at the end of each working day or, at the latest, the beginning of the next working day.
Taped geomembrane edges
AGRU America Inc. developed its Clean Seam technology to support these projects. The taped edge affixes a 6 inch (15.2 cm) wide tape to the smooth edge of the geomembrane roll using a proprietary process that leaves no residue on the liner surface when removed. The roll is manufactured with the tape on the top of the sheet on one edge and on the bottom of the sheet on the other edge (see Figure 2). When the rolls are deployed, the taped edges align as adjacent panels are overlapped or “shingled.” Just before the welding machine seams the panels, the tape is removed, revealing a clean geomembrane surface free of dirt and moisture and ready for welding (see Figure 3).
For both sites, geomembrane installation began each day with trial seams conducted in the morning and after the lunch break before beginning production seaming. Installation personnel established and set the target seaming temperature, speed and pressure appropriate for the conditions that day and set the specified parameter tolerances in the welding machine. In addition, the welding technician set the data recording frequency according to the project specifications.
CCS was able to install all four layers of geosynthetics on the project in two days. CCS used Leister GEOSTAR G7 LQS double-track fusion wedge welding machines to seam the geomembrane. The Leister units were set to record a data point every 4 inches (10.2 cm) and if a defined tolerance was exceeded, the machine recorded data on a 2-inch (5.1 cm) frequency until the readings returned to the project-specified range.
CCS seamed approximately 520 feet (158 m) of double-track fusion seam on the primary geomembrane and approximately 550 feet (168 m) on the secondary geomembrane. In addition to the machine display, the equipment allowed CCS personnel to monitor seaming using a tablet PC (connected to the welding machine by Wi-Fi).
Due to site constraints, the preferred deployment direction did not always allow the taped edges to shingle properly toward the pond sump. Given the small size of the project, some of the panels were rotated by hand to allow the taped edges to mate properly. Additional area was prepared to be able to deploy the geomembrane from the opposite side of the pond and alleviate the need to rotate the sheets.
Each welding machine automatically assigned sequential seam numbers (e.g., 1, 2, 3). Therefore, the first trial seam began with seam number 1 and production seams started with the next sequential seam number after trial welding (e.g., 3 or 4), depending on the number of trial seams prepared. At select times during the day, the geosynthetic installer transmitted DAW reports electronically to CQA personnel in portable document format (.pdf). DAW reports were comprised of the following information:
Seam summary record: One-page report showing four seams (per summary record), including project information, welding machine make/serial number, seam number, seam start and end times, and the set and measured temperature, speed, and joining force (referred to as pressure) at the seam start and end times.
Graphical report: One page showing the temperature, speed and pressure plotted versus the data measurement position (distance) and color coded to visually identify out-of-tolerance seaming parameters.
Detailed report: Multiple-page seam record reporting each data measurement position (distance), temperature, speed, pressure, latitude and longitude. Out-of-tolerance seaming parameters were reported in red text to visually distinguish them. The minimum and maximum temperature, speed and pressure, as well as the specified welding limits (tolerances), are summarized at the end of each detailed report.
Learning from the Site 1 installation, which showed consistent speed and temperature, but out of tolerance pressure with acceptable results, the pressure envelope was expanded from ± 20 pounds force (89 N) to ± 50 pounds force (222 N) with the speed and temperatures remaining the same. ESI used Demtech VM-20/A (with Pro Data) double-track fusion wedge welding machines to seam the geomembrane. Unlike the Leister machine that reported data recoding intervals in inches (cm), the Demtech equipment recorded data in time intervals. ESI set the data recording frequency to one recording every second.
ESI seamed approximately 35,601 feet (10,851 m) of double-track fusion seam on the 40 mil (1 mm) HDPE geomembrane over 10 workdays.
The processing pad was designed to drain radially with a high point in the center. This configuration did not allow the taped edges to be used with the desired panel shingle to allow down-slope drainage to the perimeter channel. Due to the fine particle size of subgrade soils and the limitation on using the taped edge on the major seams, CQA personnel collected eight destructive samples to confirm seam strength without the taped edge. (See Figure 5.) The eight destructive samples passed. In comparison, a typical installation based on the current industry standard would have required 70 destructive samples.
The Demtech equipment allows the welding technician to input the seaming data so the weld/seam number can be customized at the time of installation instead of requiring post-processing by the CQA firm. ESI marked the liner with paint each time a welding tolerance was exceeded, which simplified identifying and visually evaluating locations with parameters outside of specified tolerances. In addition, an exception report was produced that only identified out-of-tolerance areas for the day’s installation instead of reporting all data for each seam. The reporting can be adjusted to provide all data for each seam, but the project specification and regulatory reporting requirements did not dictate showing all data. ESI transmitted DAW reports electronically to CQA personnel in portable document format (.pdf) at the end of each installation day.
The seam weld report is a multipage report produced for each seam where a welding parameter exceeded tolerances. The first section contained project information, welding machine make/serial number, seam number, seam start and end times, and the set and measured temperature, speed, and joining force (referred to as pressure) at the seam start and end times. The second section contained seam data, which showed a graphical representation of the seam for each parameter followed by a seam detail report providing all out-of-tolerance parameters and their location on the seam.
Data and quality-based geomembrane seaming—evaluation
Wood Environment & Infrastructure CQA personnel reviewed DAW reports in conjunction with non-destructive seam test results and field observations. Fusion seams were non-destructively tested using the air-channel method (ASTM D5820) consistent with standard practices and technical specification/CQA plan requirements. Non-destructive test results indicated all seams passed test requirements. DAW reports for Site 1 indicated that seaming speed and temperature remained stable, constant and were within specified tolerances. DAW reports indicated that seaming pressure fluctuated above and below specified tolerances. Interestingly, for Site 2, the pressure and temperature were constant, while the installer needed the ability to vary the speed to avoid burnouts on the thinner 40 mil (1 mm) geomembrane.
During trial welding, personnel observed that the welding machine “set” pressure, before the machine was clamped onto the geomembrane, was higher than the “welding” pressure displayed after the machine was clamped onto the geomembrane. The welding pressure was approximately 70 pounds-force/foot (311 (kN/m)) less than the “set” pressure. Therefore, the seaming pressure tolerances set in the welding machine were adjusted to account for the set-to-welding seaming pressure change.
Where DAW reports indicated seaming parameters outside of the defined tolerances, CQA personnel located and observed the seam quality. Acceptable seam quality was based on observing consistent welding machine track indentations, straight alignment, consistent geomembrane surface appearance and cleanliness.
DAW reports at Site 1 and observations indicated that the majority of out-of-tolerance data were low pressures recorded at anchor trenches, seam ends and wrinkles; and high pressures recorded at geomembrane tee seams. The geosynthetic installer and CQA personnel attributed low pressures to technicians holding the welding machine by its handle and lifting, guiding or directing the machine across anchor trenches and wrinkles. High pressure records at tee seams were attributed to the welding machine passing over three layers of geomembrane.
Seam weld reports at Site 2 revealed that pressures were constant with the Demtech machines; however, ESI needed more flexibility in the specified speed tolerance window to account for increasing sheet temperatures as the day progressed. The project specifications were adjusted following additional trial seam testing to ±2.5 feet/min (0.76 m/min).
Finally, both sites experienced some difficulties with welding parameters being set in the welding machines as a percentage of the specified value, instead of an absolute value. This resulted in approximate ranges and tolerances for the project. Trial welds showed that often the specified tolerances could be exceeded up to three times the project value before seam strengths were affected. This factor of safety built into the welding process diminished the need for absolute tolerances on the welding parameters.
Conclusions and recommendations
Some have questioned the need for and value of long-established industry standard geomembrane design and installation CQA practices requiring geomembrane destructive seam sampling and testing at a frequency of one sample per 500 feet (152 m). Practice consistently shows that the vast majority of destructive test results pass project requirements.
The authors had the opportunity to evaluate two new approaches to liner installation through these pilot projects. These projects aimed for destructive testing only as the data indicates a need by using taped geomembrane edges to reduce seam failures due to contamination, and by using automated welding technology to measure and record seaming parameters.
Based on the pilot project results, the authors make the following conclusions and recommendations:
DAW seaming specifications can be identified before the project begins; however, field trial seams should be conducted to affirm the actual welding window for parameters prior to production seaming. This gives the installer the reasonable latitude to adjust installation techniques to match the site environmental and material conditions while also providing the engineer confidence that the parameter ranges still satisfy the project requirements.
Up-front coordination is required to make sure variable tolerances, report setup and dissemination are coordinated. Beginning the welding process after these details are resolved can prevent delays during installation and result in higher confidence interpreting the data results.
The installer must be aware of and plan for the direction of geomembrane panel deployment to match the tape on the top and bottom edges of the geomembrane.
The authors are confident that the data and quality-based geomembrane seaming pilot projects resulted in a higher quality geomembrane installation relative to standard industry practices. The risk to extended installation time and costs from destructive testing were reduced because the data and quality-based geomembrane field seaming framework resulted in no destructive seam sampling/testing for Site 1 and significantly reduced sampling/testing for Site 2.
Though potential cost and time savings exist, they were not quantifiably verified. The increased cost of taped-edge geomembrane and data acquisition welding machines may be offset by less destructive testing. The authors also note that competitive bidding may be limited because there are limited geosynthetic installers using DAW welding machines and manufacturers making taped-edge geomembrane.
Evan Andrews is the chief engineer for Sequoia Services, a heavy civil contractor located in Greensboro, N.C. He was formerly a lead engineer for Duke Energy in its landfill programs group and a regional manager for solid waste consultant Joyce Engineering. He is a professional engineer in six states and serves as a liaison to the International Geosynthetics Society, North American Chapter from the CCR industry.
Ken Daly is a principal civil engineer serving as Wood Environment and Infrastructure’s solid waste/CCR practice lead based in Charlotte, N.C. He has over 26 years of experience in engineering design, permitting and construction of solid waste management facilities. Since 2007 he has served as the lead engineer, project manager and its program manager for numerous CCR landfill and impoundment closure projects.
All figures courtesy of the authors.
Wright, T. D., Held, W. M., Marsh, J. R., and Hovater, L. R. (1987). Manual of procedures and criteria for inspecting the installation of flexible membrane liners in hazardous waste facilities. U.S. Environmental Protection Agency. https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=910243S8.TXT.
Geomembrane destructive testing
Contractor: Sequoia Services
Engineer: Wood Environment & Infrastructure
Installers; Chesapeake Containment Systems (CCS); Environmental Specialities International (ESI)
Productsz: AGRU HDPE Geomembrane with Clean Seam technology; Leister Geostar Welder; Demtech VM 20/A with Pro Data
ACKNOWLEDGEMENTS: NC Department of Environmental Quality (NC DEQ); International Association of Geosynthetics