This page was printed from https://geosyntheticsmagazine.com

Wick drain installation considerations

Wick drain installation considerations at one of the largest wick drain projects carried out in North America to date.

Features | October 1, 2022 | By: Michael Etezad, Ben Riddell, Ken Bocking, Dennis Becker, Travis Pastachak, Jason Bell, Erik Wimmelbacher and Steve Boileau

Figure 1. Prefabricated vertical wick drain installation at Rainy River

Wick drains are one of the most common and cost-efficient ground improvement techniques for soft fine-grained soils. Over the years, there has been substantial development on wick drain design methods; however, construction considerations may have not been sufficiently considered during the design phase. Failure to incorporate construction considerations into the design and project execution plan may result in wick drain projects not performing as desired. 

This article provides construction considerations and recommendations related to one of the largest wick drain projects carried out in North America to date, in which about 8,530,183 linear feet (2,600,000 linear meters) of wick drains were installed. Unique features of the project include complex earthworks geometry and high confining pressures applied to the installed wick drains. Part 2 of this article provides discussions on considerations regarding the wick drain installation, including drainage blanket construction, wick drain materials, the effect of dense layer(s), wick drain installation rigs, cold weather considerations, as-built data collection, and data processing and instrumentation requirements.

Wick drain installation considerations

This section discusses important considerations that should be assessed during the wick drain installation.

Surface grading of the drainage blanket layer

The spacing of the wick drain tips near the base of foundation clay is an important consideration where potential critical failure surfaces may exist at depth. To maintain consistent spacing at depth, adjacent wick drains need to be installed along consistent parallel vectors. To facilitate parallel installation of wick drains, the drainage blanket surface must be uniformly graded with minimal concave or convex features. Where the drainage blanket surface has concave changes in grade, wick drains would diverge at depth beyond the 7-foot (2-m) design spacing.  The resulting increased wick drain spacing at depth would result in dissipation times that are greater than those used in the design. This could result in a significant decrease in the allowable rate of stockpile raising, which cannot be accommodated by the mining plan. 

While tip divergence could be corrected by carefully plumbing the rig and the mast at each wick drain location, such plumbing is tedious and would considerably reduce the productivity of the installation rigs. As a result, providing simple planar drainage blanket areas to facilitate parallel wick drain installations is an essential condition for the performance of the wick drains.

Practically, parallel installation of wick drains can be achieved either by installing all wicks vertically or by installing them at a uniform inclination along a rig heading that is parallel to the dip on the plane of the drainage blanket. 

If the second option is selected, the wick drain installation will be installed at a constant inclination from vertical and azimuth (i.e., an installation vector) perpendicular to the mean drainage blanket surface. This installation vector will need to be consistently maintained within predefined, uniformly graded areas, across time and differing installation rigs. It would still be necessary to check and set the installation vector of the mandrel for each wick drain installation and to provide construction quality control (CQC) verification of the installation details. For this project, this method was anticipated to be more challenging than simple vertical installation of the wick drains. 

The drainage blanket grading surfaces were designed based on the wick drain installation rigs’ mast adjustment range of approximately +/- 6°, so that the wick drains could all be installed vertically. As the mast of the wick drain installation rig can only be adjusted in the fore and aft direction (i.e., parallel to the rig tracks), it was necessary for the rigs to move back and forth along a constant installation direction, parallel with the dip direction on each planar section of the surface grade.   

Due to the large size of the wick drain installation area, it was not possible to provide a single uniformly sloped surface for the entire wick drain area. To reduce cut and fill regrading costs, the wick drain area was divided into a number of planar segments with differing surface slope grades and directions. A wick drain installation rig heading direction was specified for each segment parallel to dip direction so that the masts of the wick drain installation rigs could be adjusted to vertical. Figure 2 shows a typical wick drain installation orientation plan. The grade changes at boundaries between segments were normally rounded out over an area due to prior and ongoing trafficking over the drainage blanket. Where a wick rig traversed a grade break between segments, there was a zone over roughly the length of the equipment tracks where vertical installations could not be maintained, and so it was necessary to provide additional “fill-in” wick drains in these zones.

Figure 2. A typical wick drains installation orientation plan

Wick drain materials

In many locations, the confining pressures that will be applied to the installed wick drains are higher than those normally experienced in wick drain ground improvement. These high pressures result from the depth of wick drain installation (up to 131 feet [40 m]) combined with more than 131 feet (40 m) height of stockpile that will eventually be placed on top of the current ground surface. The concern was that the wick drains currently available in the market might not tolerate the high confining pressures that could affect the drainage performance of the installed wick drains. Wick drain materials available in the market were considered and compared. The selected wick drain material was tested prior to beginning the project to confirm that it would meet the design requirements. The wick drain testing carried out included grab tensile strength, puncture strength, trapezoidal tear strength, permittivity, apparent opening size, tensile strength, discharge capacity and low-temperature brittleness. 

An important design consideration was the wick drains’ discharge capacity under the normal and clogged conditions in the short and long term to ensure that the wick drains will perform as expected. Laboratory discharge capacity testing was carried out on wick drains for confining pressures up to 145 psi (1,000 kPa). Additional testing was conducted on samples obtained by construction quality assurance (CQA) and CQC during construction.

Wick drains discharge capacity may be reduced due to the foundation clay consolidation and the resulting vertical and horizontal deformations within the ground. It is important for the settlement and horizontal deformations within the clay foundation to be less than the maximum wick drain material allowable strain. In this project, the clay consolidation and deformation analyses were carried out using the Terzaghi method and finite element modeling to assess the foundation deformations during and at the end of the stockpile operation.

The effect of a dense layer 

Given the complicated surficial geology of the site, it is possible for a dense layer, cobbles or even boulders to occur within the clay till layer. Such layers could result in an early refusal of the wick drain mandrel. Where such early refusal occurs, the wick drains will not penetrate through the entire depth of the foundation clay layers and into the underlying sand till. Early refusal of a single wick drain is of little consequence; however, this issue could substantially affect the performance of wick drains if the dense layer is extensive. If a large area of foundation clay remains untreated due to the presence of an overlying dense layer, the clay without wick drains will not experience strength gain in the time required to raise the stockpile, allowing potential failure surfaces to develop at depth, passing below the zone of wick drain foundation improvement. In this project a few dense layer areas were suspected based on wick drain termination depths that did not match expectations. Boreholes were drilled in these areas to confirm the presence of a dense layer and associated early termination of wick drains, as shown in Figure 3. For the confirmed locations and areas where zones of foundation clay were not penetrated by wick drains, and thus would not have the same degree of strength gain as zones with wick drains, 2D or 3D limit equilibrium computer stability analyses were carried out to verify that the factor of safety for stockpile geotechnical stability still met the design requirement. 

Figure 3. Drilling investigation to check the presence of a dense layer

Wick drain mandrel bending 

During the second wick drain installation campaign, a significant variation in penetration depth was reported between adjacent wick drains at certain areas (e.g., in the order of 49 feet [15 m] differences in depth over the 7-foot [2-m] horizontal wick drain spacing). Boreholes were drilled at these locations to identify the correct refusal depth. The boreholes indicated that the shallower wick drain penetration depth was often correct. Based on the findings of the field program, it was inferred that the wick drain mandrels were bending and following the contact with a steeply sloped surface on the bottom sand till layer or bedrock, rather than penetrating these layers to reach refusal. Bending of the mandrel resulted in incorrect reported penetration depths of wick drains. 

As part of the CQC and CQA, the rig mandrels were checked to confirm they remained straight, and bent mandrels, if observed, were straightened or replaced.

Cold weather

For this project, water was used to lubricate the mandrel, and as a result, wick drains could not be installed when site temperature fell below 32°F (0°C). An attempt was made to install wick drains in freezing conditions and without the use of water; however, due to the high plasticity (stickiness) of clay, the rig was not able to push the wick drains to the design depth. While it may be possible to advance a mandrel through shallow frost depths, freezing air temperatures must be considered if it is necessary to use water as a lubricant to achieve the required penetration depths. 

Sufficiently covering the wick drains with stockpile material prior to ground freeze is another cold weather consideration. If the cover cannot be placed on the drainage blanket before the ground starts to freeze, fill placement on the wick drains must be delayed until the foundation soil and the drainage blanket are fully thawed the following spring. If this procedure is not followed, a multi-year frozen layer could be buried at and below the drainage blanket, preventing the drainage of water through the wick drains and drainage blanket, and resulting in wick drainage not performing as the design intended. In this project, no rockfill was allowed to be placed on top of wick drains or the drainage blanket while they were frozen.

As-built data collection and data processing

Due to the large quantity of daily wick drain installations, data management was an important part of this project. CQC installation data were collected daily using the data acquisition systems, reviewed and processed by the contractor, and then provided daily to the CQA team to review. CQA review of the CQC data included spatial data review of installation details using CAD software. A color-coded plot showing the refusal depth of each wick drain was particularly useful in spotting non-conforming installations.  

The automated data acquisition system and spatial data review techniques implemented in this project made the review of thousands of daily wick drains installations and identification of installation issues significantly more efficient than traditional manual methods. An automated system has proven to be an important addition for a large wick drain installation project.

Instrumentation requirements

As noted in Part 1 of the article, vibrating wire piezometers (VWPs), settlement plates (SPs) and slope inclinometers (Sis) were installed to monitor the performance of the stockpile. VWPs were installed at the critical locations to measure the dissipation of excess pore water pressure following loading. An important part of the VWP installation was that at each location, paired VWPs were installed close to each other but separated by a minimum two rows of wick drains. The rationale behind this is that if one or more wick drains adjacent to a VWP clogs, that VWP may indicate a slow dissipation of excess pore water pressure that may not be representative of the unclogged wick drains in the larger area. Such a condition could result in unjustified warnings and concerns on the performance of the wick drains. The installation of additional adjacent VWPs helps in identifying the source of a possible problem where elevated pore water pressure is observed. Potential causes of elevated pore water pressures could include clogging of the wick drains adjacent to the VWPs or too rapid a rate of stockpile raise.

SPs measure the consolidation settlement of the clay. This provides an independent method to confirm the excess pore water pressure dissipation measured in the VWPs and associated clay foundation strength gain. 

SIs were installed at the critical locations to measure lateral ground movement and to identify the development of a shear layer within the foundation clay.

Figure 4 shows the dissipation of excess pore water pressure following the fill placement at two VWP locations over a relatively short period. As shown, the wick drains are functioning well as expected and satisfy the design requirements.

Figure 4. Stockpile VWP results

Conclusions

This article provides details on the construction of a very large wick drain project carried out in North America. 

Part 2 of this paper discusses unique features of the project, which include considerations regarding the surface grading of the drainage blanket to provide a suitable working platform for wick drain rigs, wick drain materials and high confining ground pressure that the wick drains will experience, the effect of a dense layer, wick drain mandrel bending, cold weather considerations, automation of the wick drains installation data and data processing, and instrumentation requirements. Appropriate attention to the construction considerations provided in this article was developed throughout the project by the owner, contractor and designer, which resulted in the overall success of the project. 

Michael Etezad, P.Eng., is a principal geotechnical engineer at WSP Golder in Mississauga, Ontario, Canada.

Ben Riddell, P.Eng., is a lead geotechnical engineer at WSP Golder in Mississauga, Ontario, Canada.

Ken Bocking, P.Eng., is a fellow geotechnical engineer at WSP Golder in Mississauga, Ontario, Canada.

Dennis Becker, P.Eng., is a fellow geotechnical engineer at WSP Golder in Calgary, Alberta, Canada.

Travis Pastachak, PMP, is capital projects manager at New Gold Rainy River Mine, Ontario, Canada.

Jason Bell is projects technical supervisor at New Gold Rainy River Mine, Ontario, Canada.

Erik Wimmelbacher, P.Eng., was project coordinator at New Gold Rainy River Mine, Ontario, Canada, for the work and is now underground engineer.

Steve Boileau, P.Eng., P.Geo., is geotechnical engineer at New Gold Rainy River Mine, Ontario, Canada.

All figures courtesy of the authors.


Project Highlights

Rainy River Mine

Owner: New Gold Inc.

Location: Ontario, Canada

Engineers: WSP Golder

Geotextile Products: Titan TE-BXC30 biaxial geogrid/composite; TenCate Mirafi HP370 reinforced woven geotextile

Wick drains: Mebra-Drain MD-88

Share this Story