Patch extrusion welding as a geomembrane failure mechanism

Geomembrane performance: Lessons learned—Part 2

By Garrett Q. Dutson

Introduction

This case history examines the geosynthetic construction, use, failure, and repair of an evaporation pond at an EnergySolutions facility in Clive, Utah.

A high priority is placed on management of water that has contacted waste stored in this pond. A geosynthetic investigation into the root cause of repeated excessive leaking in the pond is discussed. Two failure mechanisms were discovered and are also reviewed:

• damage due to maintenance activity.
• poor patch extrusion welding technique.

The poor patch extrusion welding consisted of a large number of sharp, upwardly-turned hardened extrudate that appeared to have worn against an overlying HDPE liner and created raised bumps and tears in the overlying liner. It appears that the combination of this unsatisfactory geosynthetic lining welding technique, coupled with the pond’s operation and maintenance, created a preventable geomembrane failure mechanism.

In conclusion, subsequent successful repairs of the 1995 evaporation pond and lessons learned are discussed.

Site overview

EnergySolutions is an international nuclear services company that operates a low-level radioactive and hazardous waste disposal site near Clive, Utah, in Tooele County, west of Salt Lake City.

Management of water that has contacted waste is a high priority at the Clive facility and is known as “contact water.” Support facilities and evaporation ponds were constructed at the Clive site for water management.

The ponds are governed by a Groundwater Quality Discharge (GQD) Permit issued by the state of Utah. The permit specifies the construction, operation, and monitoring requirements for all Clive facilities with the potential of discharging pollutants that could move directly or indirectly into groundwater.

Pond overview

The Clive facility has eight evaporation ponds used for water management. An evaporation pond originally constructed in 1995 is the subject of this case history and its location is shown in Figure 1.This pond is located within the waste handling area and is used to store and evaporate contact water that can also be reused for dust suppression in that area.

The pond’s dimensions are 58m × 66m × 2.7m (190ft × 215ft × 9ft), with a capacity of nearly 5.3 million liters (1.4 million gallons) of contact water at its permitted 1.2m (4ft) depth limit. The climate is arid to semiarid, with average annual precipitation usually less than 0.2m (8in.) and a total evaporation generally around 1.3m (50in.), (Montgomery Watson, 2000).

This particular pond was originally constructed on a foundation of native clay materials overlaid with three layers of geosynthetics. A leak-detection system was constructed throughout the pond using geonet for drainage located between two 60-mil HDPE liners: a lower “secondary” containment liner and an upper “primary” containment liner.

All layers are sloped to a collection sump that includes a HDPE access pipe housing a submersible pump and leak-detection equipment. Any contact water extracted from the leak-detection system is discharged back onto the surface of the primary liner.

The GQD permit specifies a particular allowable leakage rate for the pond’s primary liner that, when exceeded, triggers a contingency response plan. The permitted allowable leakage rate of the pond at the time of this case study was approximately 613 liters/day (162 gallons/day).

Pond issues and repair history

Following installation in 1995, the pond served water management needs until late 1998 when the upper, or primary, liner (the 1995 primary liner) was accidentally torn during the removal of several pieces of aerator equipment.

The record indicates that the pond was leaking as much as 2,650 liters/day (700 gallons/day) but does not indicate any other reasons for the leaking. The 1995 primary liner was subsequently repaired in the summer of 1999.

The 1999 repairs consisted of the addition of a new primary 60-mil HDPE liner (the 1999 primary liner) placed directly over the torn and leaking 1995 primary liner and the removal and replacement of the HDPE access pipe and leak-detection equipment (Comstock, 1999).

The pond was then placed back in service until October 2006 when site personnel—in an effort to clean the pond—pumped it out, placed it out of service, and attempted to remove sludge by shovel. Correspondence states that while removing the sludge, damaged areas were identified.

The correspondence further states that subsequent cleaning and precipitation events over the pond led to excessive leakage of 1,090 liters (288 gallons) in February 2007 (McCandless, 2007). Thereafter, a determination was made to further investigate the cause for the excessive leakage and to establish a repair plan of the pond.

In April 2008, preliminary field inspections of the 1999 primary liner surface were performed by site personnel, and a CQA/QC plan was developed detailing the required inspection and repair procedures. Accordingly, in May 2008, an in-depth investigation into the root cause(s) of the pond’s excessive leakage rate was initiated.

The investigation focused on the increased leakage rate, which followed the initial manual cleaning of the 1999 primary liner surface. The investigation was performed by a joint inspection team consisting of site and regulatory personnel and an independent geosynthetics quality assurance company (QA inspector). During that investigation the QA inspector identified and documented nearly 190 combined potential defects in the 1995 and 1999 primary liners.

The inspection personnel characterized surface punctures as the result of two different “phenomena.”

1. The first condition was the product of the October 2006 hand-shovel damage during pond sludge maintenance where the edges of the shovel blades caught an edge of existing patches or bead repairs, lifting extrudate and tearing the 1999 primary liner. Figure 2 shows an example of the shovel damage to an edge of a patch in the 1999 primary liner.
2. The second condition involved raised lumps in the surface of the 1999 primary liner caused by mounded extrudate on the surface of the underlying 1995 primary liner. (See Figures 4 and 5) for an example of the poor patch extrusion welding technique found on the 1995 primary liner as it was uncovered during the May 2008 investigation.It is discussed in detail below (Wienecke, 2008).

Following discussion and agreement with the Utah Department of Environmental Quality, the known defects were cut out rather than repaired due to the large number of defects discovered. Additionally, this provided a flatter, more suitable subgrade surface for a new lining system. Photos of the defective repairs (Figures 3a and 3b) show cutout areas varying in size, on average, from 0.3 to 0.6m (1–2ft) in diameter (Wienecke, 2008).

It was also determined that due to the risk of damaging the underlying clay foundation, and to avoid the risk of introducing subsurface contamination from Clive facility operations to groundwater, all existing liner systems would be left in place following removal of the known defects and would be covered with a new liner system.

The defective areas were then cut out and removed to expose the 1995 primary liner and, in some cases, the underlying geonet and 1995 secondary liner (Dutson, 2008a). In addition, a small cut, less than 5cm (2in.) long, was discovered on the underlying geonet during field preparations for the repair of the pond and during removal of a poor extrusion bead weld on both the 1999 and 1995 primary liners.

It was discovered upon removal of the affected portion of the geonet that the 1995 secondary liner had a scar of approximately 2.5cm (1in.). A vacuum box test of the scar verified that a pinhole through the 1995 secondary liner had been created.

Following the procedures of the CQA/QC plan, the secondary liner surrounding the pinhole was removed and the extent of damage to the underside and surrounding clay foundation was assessed and sampled for contamination. The clay foundation was tested and determined to be intact and free of excessive moisture and from contamination. It was subsequently repaired with powdered bentonite and the area was re-patched with geonet and replacement 60-mil HDPE liner as required (Dutson, 2008a).

It was further demonstrated by visual inspection that the poor patch work on the 1995 primary liner had penetrated upward into the overlying 1999 primary liner, causing the bumps originally discovered and contributing, along with the shovel damage, to the leaking liner and ultimate pond failure. It was also demonstrated by visual inspection that the poor 1995 primary liner extrusion patch work did not penetrate downward into the underlying geonet, that the underlying 1995 primary liner surface was flat, and that the geonet appeared to have acted as a cushion between the 1995 secondary and 1995 primary liners.

No damage was found in the geonet, except for the 2.5cm scar described above (Dutson, 2008a).

The poor patch extrusion welding employed during the 1995 primary liner installation was determined to be a key failure mechanism and is of prime importance in this case history. The poor patch extrusion welding coupled with inadequate extrusion bead grinding before the 1999 primary liner installation are believed to have combined with natural thermal expansion and contraction forces to contribute to the failure of the 1999 primary liner.

What was alarming during the investigation was the large number of poorly completed extrusion welds that had created sharp upward-turned extrudate, characterized by a hardened, upwardly-directed shaped edge. (See Figures 4 and 5 for an example of the discovery of this poor patch extrusion weld as it affected the 1999 primary liner.)

This extrusion weld failure mechanism was identified during preliminary inspections during April 2008 and was thereafter exposed in multiple locations during the official inspection in May 2008 (Dutson, 2008a).

The QA inspector concluded that during extrusion patch welding of the original 1995 primary liner, the installers finished their extrusion welds by lifting the extrusion welding machine up and off of the welded 1995 primary liner in an upward sweeping motion that left elevated, hardened extrudate.

The QA inspector further pointed out that in several locations, the 1999 primary liner had stretched, or plastically flowed, over the protrusions, eventually causing either an impediment to cleaning that the shovels caught and punctured or that the extrudate itself caused a puncture (Wienecke, 2008).

During installation and repair of the 1999 primary liner, the installer and quality control/quality assurance personnel failed to note the inherent problems that these hardened, raised surfaces would pose to the 1999 primary liner surface. It appears that the initial installation contractor could have avoided these poor extrusion welds by pressing down the tip of the extrusion gun upon completion of the weld, as was observed during the 2008 repairs.

Moreover, it appears that the 1999 repair contractor could have avoided this failure by grinding down the welds or by placing a geosynthetic barrier, such as a nonwoven geotextile, over the surface prior to placement of the 1999 primary liner.

Geosynthetic repairs and testing

Once the root cause of the failure was addressed and the 1999 and 1995 HDPE liner surfaces were cleaned, new geosynthetics, including an 8oz nonwoven geotextile and a new 60-mil HDPE liner, were installed directly over the existing liner and tested by the HDPE installation contractor in late May and early June 2008 in accordance with the CQA/QC plan.

The new 8oz nonwoven geotextile was selected to act as a cushion to provide a softened surface for the new liner surface. It also provided a filtration layer for flow between the new liner surface and the 1999 primary liner to accommodate hydraulic continuity to the 1995 secondary liner (Wienecke, 2008).

The geotextile was placed directly over the prepared floor of the 1995 pond and extended approximately 0.3m (1ft) up each sideslope. (See Figure 6 for a picture of the prepared 1999 primary liner surface and deployment of the geotextile.)Care was taken to remove dust and debris during this operation and the QA inspector verified that the installation was performed as required.

The new 60-mil HDPE liner (the 2008 primary liner) was then installed over the geotextile and fusion welded to the existing 1999 primary liner at the top of the pond’s perimeter berms. (See Figure 7 for a picture of the nearly completed 2008 repairs [Dutson, 2008a].)

Following installation and testing, the pond was filled for a required four-week hydrostatic tightness test that started in late June 2008. The water level was set at the required 1.2m depth mark to ensure that the pond was at least at half-capacity (at the permitted level) in accordance with the CQA/QC plan.

The permitted high-water alarm depth for the pond is 1.5m. No additional “contact water” or wastewater was added or removed during the test and daily readings of the leak-detection flow rate were taken and average daily leak rates were calculated.

Water was pumped from the leak-detection system during this testing on only three occasions. Two of these occasions were during manual operation of the leak-detection pump at the beginning of the hydrostatic test. The third discharge was automatically performed for a required leak-detection pump test.

The hydrostatic test was successfully concluded in late July 2008 (Figures 8 and 9).The pond successfully passed the hydrostatic test when the calculated volume of the average daily leak rate for each week of testing was shown to be all well below the permitted 613 liters/day rate.

The average daily leak rate for the four weeks ranged from 0 liters/day to 80 liters/day. Following leakage testing, the contact water was left in the pond for operational use. It was observed during testing that approximately 0.3m of contact wastewater in the 1995 pond evaporated—a rate deemed acceptable based on historical data (Dutson, 2008b).

Today, the 1995 Evaporation Pond is performing as designed and in accordance with all permit requirements (Figure 10).Pond readings have remained well within the allowable leakage rate and no additional leaks have been noted. The 2008 geosynthetics repairs are working as designed.

Lessons learned

A valuable lesson learned from this 1995 Evaporation Pond history is the importance of observing all geosynthetic patch work for potential failure mechanisms, such as poorly completed extrusion welds, before performing any repair work and to plan for them if they are suspected in a repair.

The use of geosynthetic materials, such as geotextiles, can also be employed when cost effective, to provide a solution for numerous repairs and to provide additional protection. Moreover, proper extrusion welding technique, including finishing each extrusion weld with a downward push against the welded liner surface, should be specified on each geosynthetic project.

Finally, proper geosynthetic maintenance techniques, if needed, must be employed. This could include the use of a slurry and pump to remove sedimentation on top of a liner surface in lieu of shoveling.

In conclusion, perhaps the best advice for geosynthetic pond cleaning is to leave the surfaces alone and keep all sharp objects off the surfaces. When a pond becomes full of sediment it may be time to replace at least the upper lining surface. Such practices should ensure a longer pond facility life.

Garrett Q. Dutson, M.ASCE, P.E., Esq., is the site engineer for the EnergySolutions facility in Clive, Utah.
See also Geomembrane Performance: Lessons Learned—Part 1, “Case histories of exposed geomembrane performance” by Abigail Beck in the June/July 2011 issue of Geosynthetics.

References

Comstock Jr., David (1999), 1995 Evaporation Pond Leak Detection System Upgrade As-Built Report, Envirocare of Utah Inc., Salt Lake City, Utah, 1-8.

Dutson, Garrett Q. (2008a), As-Built Certification Report for the 1995 Evaporation Pond Repairs Project, EnergySolutions Inc., Salt Lake City, Utah, 1-5 and Attachments.

Dutson, Garrett Q. (2008b), As-Built Certification Report for the 1995 Evaporation Pond Repairs Project Addendum–Hydrostatic Test Results, EnergySolutions Inc., Salt Lake City, Utah, 1-4 and Attachments.

McCandless, Sean (2007), Subject: Evaporation Pond Allowable Leakage Rate Monitoring, correspondence with Utah Radiation Control Board, Nov. 14, 2007.

Montgomery Watson (2000), Evaluation of Water Management Alternatives for the Envirocare Facility, Clive, Utah, Section 1-3 and Appendix B–Stormwater Runoff and Evaporation Estimates.

Wienecke, Christopher J. and Repola, Kerry M. (2008), 1995 Evaporation Pond Repairs Construction Report for Geosynthetic Materials, Advanced Terra Testing Inc., Lakewood, Colo., 1-31 and Attachments.