Geocomposite cover for sulfide tailings

By Bruno Herlin, P.Eng.

Abstract

In autumn 2006, the Ontario Ministry of Northern Development and Mines tendered a construction project to provide a soil cover over the North Impounded Tailings (NIT) area at the abandoned Kam Kotia Mine site in east-central Ontario, Canada. The soil cover would effectively impede the entry of water and oxygen into the high sulfide tailings, substantially reducing acid generation and metal leaching effects from within the tailings.

The final design incorporated waste rock, sand, gravel layers, a geosynthetic clay liner (GCL), clay, and granular cover soils. During the fall of 2006 and winter of 2007, Hazco Environmental & Decommissioning Services implemented the construction of this design. Overall, 800,000m² of GCL was deployed and covered. Terrafix Environmental Inc. undertook deployment of the GCL under the supervision of Hazco and Earth Tech Engineering. This article summarizes the design and construction of the composite soil cover system installed at the Kam Kotia site.

Introduction

The Kam Kotia Mine in Timmins, Ontario, operated intermittently from the 1940s until 1972, producing copper, zinc, and gold. Following closure, approximately 3 million metric tonnes (MT) of acid-generating sulfide tailings and 500,000 MT of acid-generating waste rock were left on the surface at the site. These waste materials have since evolved to become significant sources of acid rock drainage (ARD) and metal leaching (ML), which has had a significant impact on the surrounding environment.

The Ontario Ministry of Northern Development and Mines (MNDM) implemented a multi-staged rehabilitation program at the site to mitigate the ARD-ML effects of the waste deposits. Several phases have been completed to date.

This paper will examine only the composite soil cover system installed during the fall of 2006 and winter of 2007.

The MNDM currently collects and treats ARD runoff and seepage and operates a high-density sludge (HDS) treatment plant to treat ARD-affected runoff. Wardrop Engineering Inc. and SENES Consultants Ltd. were retained to design a dry soil cover for the North Impounded Tailings (NIT) in 2004 and 2005. This design was tendered for construction in 2006 and was awarded to Hazco Environmental & Decommissioning Services.

The design goal for the project was to provide a dry soil cover to minimize the infiltration of water and also limit the ingress of oxygen into the tailings. This construction design will reduce the quantity, acidity, and metal loading of the leachate reporting to the site’s drainage collection system to the point where passive treatment technology could be implemented and the HDS plant could be taken out of service.

Design of the composite soil cover

Ontario’s MNDM commissioned Wardrop and SENES to design a soil cover over the NIT (Figure 1). Earth and rock borrow materials were characterized and preliminary designs were assessed using parameters calculated from standard geotechnical soil testing. Preliminary hydrogeological and geochemical models were run on four design options.

The two designs using compacted local clay and geosynthetic clay liners (GCLs) were found to be equivalent in performance and cost. A second round of detailed laboratory testing was done, including calculating void ratios, freeze-thaw permeability, water retention curves, and oxygen diffusion coefficients. After incorporating this data into the analyses, the optimal final design incorporated waste rock, sand, gravel layers, a GCL, clay, and granular cover soils.

Material testing and hydrological modeling

A comprehensive sampling, geotechnical testing program, and chemical analyses of the waste rock were undertaken to characterize the borrow materials available from the surrounding glaciofluvial (granular) and glaciolacustrine (clay) deposits. In addition to natural aggregate sources, the mine waste rock was also sampled and tested for use in the cover. This material testing and hydrological modeling program was conducted under the direction of Andrew Mitchell, P.Geo., formerly of Wardrop Engineering, and Jeff Martin, P.Eng., of SENES Consultants.

Additional detailed testing such as oxygen diffusion, freeze-thaw permeability, and moisture retention was conducted under the direction of Michel Aubertin at Ecole Polytechnique in Montreal. The complete hydrological modeling design and results, including the performance of four cover option models, was presented at the 58th Canadian Geotechnical Conference in Saskatoon.

Cover options and final design

The four cover options that were modeled by Wardrop and SENES are as follows, starting from the tailings surface upward:

1. 0.3m rock / 0.25m sand / 0.5m clay / 0.5m sand
2. 0.25m rock / 0.3m sand / 1.0m clay / 0.5m sand
3. cover incorporating a GCL with appropriate sand bedding and cover
4. cover incorporating a synthetic geomembrane (PVC) with appropriate sand bedding and cover

The final laboratory test program indicated that the silty clay from the site borrow pit was highly frost susceptible and that there could be an increase in the permeability of up to two orders of magnitude with repeated freeze-thaw cycling. This raised concerns for the longevity of a cover designed with clay as the sole water and oxygen barrier. To overcome this propensity of the clay, it was decided that a GCL would be incorporated into the final design. In addition to the GCL, some other final refinements of the design were incorporated to provide a more robust and durable installation.

The description of the final cover, from the tailings surface upward, is:

1. A 300mm-thick basal layer of crushed mine waste rock—a layer formed both an effective capillary break due to its coarse grain size distribution as well as adding structural stability to the design. In addition, using the waste rock in this manner provided an opportunity to deal with this acid-generating waste as part of another element of the mine site rehabilitation, which precluded needing to cover the waste rock pile in a later phase of work at an additional cost.
2. A 300mm layer of granular fill—a layer that completed the required thickness for an effective capillary break as well as providing a suitable subgrade for synthetic liner installation.
3. Next, a polypropylene-coated GCL forms the water and oxygen barrier to effectively isolate the tailings from the ingress of water and air into the tailings mass from above. The polypropylene-coated product was selected in the final design, since it has an order of magnitude lower hydraulic conductivity than the figures assumed in the modeling, which adds conservatism in the design at little additional cost.
4. A 300mm layer of silty clay, placed directly over the GCL to ensure full hydration throughout the service life of the cover. In addition, the lower permeability of the clay acts as a secondary barrier in addition to the GCL, enhancing the oxygen barrier.
5. A 500mm layer of granular soil that isolates the clay and GCL from physical disturbance and also provides a store-and-release function to mitigate the effects of sustained high precipitation or drought. The thickness of this layer was increased from the preliminary designs to provide greater protection from frost and root penetration. In addition, it provides the required confining stress on the GCL and enhances the durability of the cover.
6. A 100mm layer of organic mulch and topsoil, obtained from the removal of overburden from the clay source, will be turned into the upper 50-75mm of the granular layer to provide a growth media for surface vegetation.

The final design incorporated into the cover to provide greater resistance to frost-induced disruptions and a layer of the silty clay available locally was incorporated into the sequence to provide continual hydration of the GCL, which is essential to maintaining low gas permeability in the comparatively thin bentonite clay layer afforded by the product.

New GCL: polypropylene-coated

A geosynthetic clay liner containing a polypropylene coating was used to provide a unique hydraulic property. This product, which has been available since 1999, adopts merging a typical textile coating procedure to that of a needlepunched GCL. Originating from the textile industry, this process yields a composite clay geosynthetic barrier (GBR-C)/geosynthetic clay liner (GCL) product with hydraulic properties and physical performance that make it well-suited to many new design approaches. The product includes a polypropylene coating applied to the woven geotextile side of a GCL, providing a low permeability typical for a geomembrane at 5 x 10^-13cm/sec (ASTM E96).

Properties/test methods on polypropylene-coated GCL

The Canadian firm Sageos/CTT Group, under the supervision of Earth Tech Engineering of Winnipeg, did testing of the polypropylene-coated GCL for this project. Hydraulic testing on the coated GCL is a difficult task in the traditional permeameter due to the lower flow characteristics of this GCL, attributable to the polymer membrane coating. Whereas a typical non-coated GCL will yield permeability values on the order of 3 x 10^-9cm/sec under 35 kPa effective confining stress and 14 kPa head pressure, testing in accordance with the Hydraulic Conductivity Test Method ASTM D5084, polypropylene-coated GCLs have shown to force side-wall leakage to occur, thus making it difficult to measure performance in the traditional permeameter.

To more accurately determine the true flow through the membrane portion of this type of GCL, a water vapor transmission test was performed. An equivalent hydraulic permeability (k) was calculated using the procedure outlined by Koerner (1997). Via ASTM E96, a value of less than 5 x 10^-13cm/sec can be expected.

The coating is typically applied to the woven portion of the GCL. This coating has added another dimension to GCLs with an increase in peel values and internal shear values. The fibers that have been needled through the composite are subjected to the coating, and as the coating becomes an integral part of the GCL, the fibers are then bonded within the coating. Another added benefit of a polypropylene-coated GCL is its effectiveness as a root inhibitor (Lucas 2002).

Project overview

Construction of the 80ha (c.200 acres) composite cover soil system started in early November 2006, with an initial deployment and completion of 10,000m² on the first day of the project. Initial plans were to deploy 120,000m² during the fall prior to closing down the project for the winter. These initial plans were changed, and 800,000m² were deployed from November 2006 to February 2007. Deployment rates at times reached more than 30,000m² per day.

Although the rate of deployment of the GCL was high at times, it is restricted by the cover soil placement over the GCL. For this project, the cover soil used was a silty clay that was available locally at the mine site. Once Hazco’s truck fleet was up and running, deployment rates as mentioned reached more than 30,000m² per day (see Figure 2).

The GCL was deployed over a stable subgrade of granular fill. A 0.3m overlap was done, recorded, and supervised by Earth Tech Engineering (Figure 3).

Following the deployment of the GCL, loose bentonite was placed between panel edges. Loose granular bentonite should be placed between the panels at a rate of 2kg per lineal meter of seam if the GCL is the primary hydraulic seal. The addition of bentonite to the seam is optional when the GCL will be acting as leak isolator for an overlying membrane. Spreading of the loose bentonite is shown in Figure 4.

Following the deployment of the GCL, a 0.3m-thick silty clay layer was placed directly over the GCL, as shown in Figure 5.

Summary and recommendations

There are many types of GCLs: stitched, glued, needlepunched, different bentonite content, different geotextile weight, scrim reinforced, enhanced polymer, etc. The list is long.

Over the years and through increased competition, this area of geosynthetics engineering seems to see ever-lower cost GCLs requested for projects. This usually means thinner textiles and/or less bentonite, almost to the point of becoming less a GCL than a double-layered textile.

What is the lowest mass per unit area of bentonite that a GCL can have and still achieve, for example, the quoted manufacturer’s hydraulic conductivity? It’s getting to the point where 2.441kg/m² (0.5 lbs/ft²) seems acceptable from the standard 3.662kg/m² (0.75 lbs/ft²). The design community originally used 4.882kg/m² (1.0 lbs/ft²). There has to be a limit, but only specific project engineers will ask for those limits.

Engineers and specifiers should request an accurate breakdown of the GCL instead of simply asking for a GCL. The recommended list of data when requesting for a GCL includes:

• top geotextile Xg/m² nonwoven
• bottom geotextile Xg/m² woven or Xg/m² scrim-reinforced nonwoven
• swell index of the bentonite
• fluid loss of the bentonite
• bentonite mass per unit area at X moisture content
• grab strength of the GCL
• peel strength of the GCL
• permeability of the GCL
• index flux of the GCL
• internal shear strength of the GCL

Scrim = woven; Scrim-reinforced nonwoven = woven + nonwoven; Scrim-reinforcement = needlepunching of a woven and nonwoven geotextile together.

All needlepunched GCLs have a nonwoven top geotextile. The bottom geotextile is either a required woven on its own or a scrim-nonwoven geotextile if required for rough soil conditions or steep slope applications.

The Geosynthetic Research Institute (GRI) recommends that the bottom geotextile of a GCL contain a scrim-reinforced nonwoven geotextile. Possible failures that may occur by not using a scrim-reinforced bottom geotextile are internal erosion of the bentonite through the geotextile in hydraulic head conditions (Rowe and Orsini, 2002), and possible shrinkage of the GCL itself in the composite lining system (GRI White Paper, 2005).

As per the GRI’s White Paper of April 2005: Do not use GCLs with needlepunched nonwoven geotextiles on both sides unless one of the geotextiles is scrim-reinforced. There are numerous possibilities in this regard, but all should have a woven component embedded within, or bonded to, the nonwoven component.

The project described herein contained a bottom woven geotextile. As mentioned, a polypropylene coating was applied to the GCL used in this case study paper to decrease the permeability of the product (GCL) to the range of a geomembrane.

Another recommendation: Never use trade names. When requesting an item for a specific project, always list the testing values required from a specific product (i.e., ASTM testing values). Also, products and their names change over time, hence the suggestion to avoid using trade names. Another similar factor is to avoid having the purchasing agent and/or general contractor make a decision during the tender process.

The variety of available geosynthetic products is great, so clients seek advice from a particular distributor, supplier, or manufacturer. But are clients provided with all of the information? Suppliers would love to provide their particular premium brands, of course, but at the end of the day, the cheapest price frequently prevails. Buyers beware. Products are never equal when using trade names. Products are equal when values are provided and, hence, can be compared. Ask for them in your tender request.

Bruno Herlin is the product manager–GCLs at Terrafix Geosynthetics Inc. in Toronto. Ron Bygness, editor of Geosynthetics, contributed to this article and to “History of the Kam Kotia Mine” sidebar.

Acknowledgements

The author would like to acknowledge the contribution of a number of people and contributors to this article: Andrew Mitchell, Jeff Martin, and Christopher Hamblin of the Ministry of Northern Development and Mines for supplying the original design paper for this project, which was presented, but not published, at the 58th Canadian Geotechnical Conference; Patrick Jolicoeur and Phil Springs of Hazco Environmental & Decommissioning Services, general contractor for this composite soil cover deployment; Troy Shaw, Blu Alexander, and especially Leroy Osmond of Terrafix Environmental Inc. as the GCL installer during constant -40°C to -50°C temperature conditions during the winter.

References

ASTM D5084. “Standard Test Methods for Measurements of Hydraulic Conductivity of Saturated Porous Materials using a Flexible Wall Permeameter.”

ASTM E96. Water Vapor Transmission.

Bentofix Technologies Inc., “Fix – 412 Features of a Scrim-Reinforced GCL & Manufacturing Process,” Barrie, Ontario, Canada, 1992, (unpublished).

Geosynthetic Research Institute (GRI), Philadelphia, Pa, USA. GRI-GCL3. “Standard Specification for Test Methods, Required Properties, and Testing Frequencies of Geosynthetic Clay Liners (GCLs),” Geosynthetic Research Institute, 2005.

Koerner, K.R. and Koerner, R.M., 2005. “GRI White Paper #5 on In-Situ Separation of GCL Panels Beneath Exposed Geomembranes,” Geosynthetic Research Institute.

Lucas, S.N., 2002. “Manufacturing of and the performance of an integrally-formed, polypropylene geosynthetic clay barrier,” International Symposium on Clay Geosynthetic Barriers, Nuremburg, Germany, 2002, pp. 227-232

Maubeuge, K.P. von, 2002. “Investigation of bentonite requirements for geosynthetic clay barriers,” International Symposium on Clay Geosynthetic Barriers, Nuremburg, Germany, 2002, pp. 155-163.

Mitchell, A., Martin, J. and Hamblin, C., 2005. “Design of a Composite Soil Cover for Sulfide Tailings at the Kam Kotia Mine Site, Northern Ontario, Canada,” 58th Canadian Geotechnical Conference (presented but not published).

Rowe, R.K. and Orsini, C., 2002. “Internal erosion of GCLs placed directly over fine gravel,” International Symposium on Clay Geosynthetic Barriers, Nuremburg, Germany, 2002, pp. 199-207.