ABSTRACT

Tremie concrete enables the construction of deep foundations such as piles and diaphragm walls as it is self-compacting, can be placed under a support fluid and is resistant to segregation. As the urban landscape within New Zealand increases in density, deeper and more complex foundations are required. These deep foundations generate higher structural demands and often require much longer concrete pour times. Consequently, tremie concrete is increasingly becoming a highly engineered material with concrete additions and admixtures used to adapt the concrete to a specific application. These challenges have contributed to major construction issues overseas and the formation of an industry led consortium to develop guidance. This paper summarises tremie concrete guidance and discusses key considerations for tremie concrete in a project life cycle for a New Zealand context. A case study of concrete testing carried out on a recent project in Auckland is presented.

INTRODUCTION – DEEP FOUNDATIONS IN NEW ZEALAND

New Zealand has seen the rise of larger and deeper foundations associated with both transport infrastructure and large private sector developments. This trend and investment in New Zealand will continue, increasing the construction industry’s reliance on deep foundations. These deep foundations require careful consideration and introduce additional complexities for design and construction. The tremie concrete process is used to place concrete in deep foundations, sometimes underneath a support fluid, to achieve a homogeneous concrete element of the required design dimensions. Tremie concrete must have sufficient workability to be self compacting, flow around reinforcing cages and displace any water or support fluid. The concrete workability must be maintained long enough for transport and to complete the pour.

Historically a higher workability concrete was achieved by increasing the water content. However this both decreases the expected strength of the concrete and decreases the rheological stability of the concrete resulting in an increased risk that the concrete will bleed and segregate. These competing demands of workability, stability and strength have resulted in concrete additives and admixtures being developed. Admixtures, such as plasticisers and super plasticisers increase workability without adjusting water content. Retarders, can be used to delay the initial set of the concrete and maintain workability of the concrete for the duration of the pour. Additives are fine particles added to modify the behaviour of the concrete. Inert additives, e.g. limestone powder, can be used to increase the fines content which increases the stability of the concrete. Active additives, e.g. pulverised fly ash (PFA) or ground granulated blast furnace slag (GGBS) can also be used to partially replace cement, improve durability and adjust strength gain.

Support fluids, typically, either a mixture of bentonite clay and water or a polymer and water can provide support to the sides of the shaft allowing excavation. The placement of concrete must occur with minimal mixing of the concrete and the support fluid or water. When mixing occurs, the concrete segregates as the cement paste is washed away leaving poorly cemented aggregate rich zones in the foundation.

Appropriate planning and execution of tremie poured concrete foundations throughout all stages of the project life cycle is essential. The following sections summarise sources of guidance on tremie concrete and provide additional recommendations for those involved in specifying, designing and constructing deep foundations in a New Zealand based on the authors’ experience.

Tremie concrete guidance for new Zealand

Recommended guidance for tremie concrete includes:

• The EFFC/DFI Guide to Tremie Concrete for Deep Foundations, (2018). Guidance on tremie concrete, performance, mix design, testing and execution of deep foundations.
• The EFFC/DFI Guide to Support Fluids for Deep Foundations, (2019). Guidance on the use of support fluids, comprising bentonite, other clays, polymers and blended fluids.
• The ICE Specification for piling and embedded retaining walls, (2016).
• Papers by Larisch, M. (2016, 2019) on tremie concrete methods, mix considerations and waterproofing.

Tremie concrete considerations for a project lifecycle

A successful deep foundation project is influenced by appropriate consideration of the use of tremie concrete through all stages of a typical project lifecycle (Figure 1) and involves input from all project participants.

Project planning & Design

Tender & Construction planning

Execution

Close out and knowledge transfer

1. Ground investigation
2. Ground conditions appropriate for foundation selection
3. Constructability reviews and risk management
4. Additional ground investigation
5. Site layout and access
6. Concrete supply
7. Ground pre treatment
8. Difficulty of construction
9. Method of construction
10. Best practice
11. Quality controls
12. Record keeping
13. Adapting to changes
14. Review performance
15. Compare with test results
16. Share knowledge

Figure : Tremie concrete considerations for a typical project lifecycle

Project planning and design considerations

During the project and planning stage the ground conditions are investigated, which is used to inform the selection of the foundation type. Specific considerations for this stage are discussed below.

Ground conditions

A thorough understanding of the ground conditions is essential for deep foundation construction.

• Groundwater and potential inflow rates, including the presence of perched groundwater or aquifers may dictate the construction method and lengths of temporary casings.
• Loose layers that may not remain self-supporting during the drilling process, can require support fluid or casing with associated additional space and cost requirements.
• Previous filling or reclamation activities can result in highly variable deposits present beneath the site which must be understood to assess the stability of plant operations and excavation.
• Strength data and particularly presence of lenses of hard material will inform drilling rates and wear.
• If the ground is contaminated, consider if it can be stabilised, e.g. by deep soil mixing, or driven foundations. Saline groundwater may affect the performance of bentonite support fluid.

Previous foundations and underground services can be obstructions for new foundations. An early understanding should be sought from a combination of historical information and onsite location by trial pits and/or probing. In some cases, previous foundations can be incorporated into the new foundation system, provided the strength and durability can be verified.

Selection of foundation type and constructability reviews

The selection of the foundation type is best carried out with careful consideration of the ground conditions and risks in collaboration with all members of the project team. The cost and programme efficiency of a foundation solution should be weighed up against the associated risks and experience of the project team.

A robust risk management process should be adopted to manage ground risks. AS2159 (2009) provides a useful risk framework for pile design. It must be recognised that the ground is inherently variable, no matter how extensive a ground investigation may be. Early constructor involvement can focus on key risks, constructability and enable a more targeted ground investigation.

Tender and construction planning considerations

During the tender and project planning stages, the project team should consider the site-specific conditions and the constraints they place on deep foundations construction.

Site layout

The layout of a site during construction can greatly affect the productivity of foundation construction leading to extended construction timeframes for deep foundations. An extended construction time increases the risk of shaft instability and associated defects, increased filter cake thickness and debris on the base of the element. As pour times extend the workability of the tremie concrete will reduce and there is an increased chance of defects occurring during the concreting process.

Sufficient laydown space is essential for storage of reinforcing cages and casing, particularly where splicing of the reinforcement cages is required. If space is not available to splice cages together onsite, the cages must be spliced together over the foundation during installation, adding time and hazards. This can be complicated by space requirements for other trades and activities onsite, access to site, set up space and egress.

Pre treatment

Pre treatments to remove obstructions and improve weak ground can be carried out. Replacement materials should be at least as strong as the surrounding material, to ensure the design of the deep foundation is not compromised. They should not be so hard that it is difficult to subsequently excavate leading to verticality issues. Replacement materials include cement bentonite mixtures, flowable fill, and compacted fill.

Concrete supply

There is potential for concrete demand to exceed the reliable daily supply. The construction team should review the required volumes in combination with the expected time to complete drilling and caging up of foundations. Multiple cage splices can result in reinforcement placement taking several hours. A decision should be made whether concreting can realistically be completed by the end of the consented working day.

In New Zealand, many concrete plants are dry batching plants, where most of the water is added to the mix after it has been placed inside the truck. Adequate mixing is expected to be achieved during the journey from the batching plant to the work site. Wet batching plants mix both water and cement together inside the plant and dispatch it into a truck once it is mixed. Project teams should consider the type of plant supplying the site and how long it may take for trucks to reach the site. It may be necessary to delay trucks onsite to ensure the concrete has had long enough to adequately mix, or conversely if admixtures should be included in the mix design to delay the onset of curing if significant travel or wait times are expected.

Pour sequencing

Depending on the volumes of concrete required for a pour and the overall dimensions of the pour, multiple tremie’s may be required. The EEFC/DFI Guide (2018) provides guidance as to how the tremie’s should be arranged. The geological conditions should be reviewed and an assessment made as to whether the piles are likely to be able to be dry poured piles. However, there should also be a contingency for wet poured piles.

Design considerations for tremie concrete

The size of the foundation element and the congestion of the reinforcing cage must be considered early in the design to assess whether concrete can flow adequately through the reinforcing. Where laps cause additional congestion, cranking bars to run parallel inside external longitudinal bars should be considered. Particular care is required at transverse reinforcement zones, such as floor slab locations with ‘L’ bars with coupler connections protected with a box out. At these zones, the combination of the ‘L’ bars, regular longitudinal and horizontal reinforcing and transverse reinforcing ties can lead to highly congested zones, especially if butterfly type reinforcing ties are used. Other sources of congestion such as laps or splices should be positioned away from these zones. Increasing the thickness of the element to reduce the congestion should be considered. Further detailing guidance is provided in the EFFC/DFI Guide (2018).

The geotechnical shaft and base resistance of a deep foundation must consider the type of support fluid used and the construction processes. For bentonite support fluids in particular, the shaft capacity may be reduced by the presence of a filter cake of bentonite that forms at the perimeter of the shaft. Guidance is provided by various sources (Fleming and Sliwinski 1991, Lam et al 2014). There is considerable variation and the most reliable method is a load test on a foundation element that mirrors the expected construction.

Execution considerations

The execution phase is largely about following best practice and the project controls. Best practice guidance on the use of tremie concrete and support fluid is provided in the EFFC/DFI Guide (2018). The following section expands on the authors’ experience of the use of tremie concrete in New Zealand.

Typical practice in New Zealand for bored piles

Temporary casing

The temporary casing serves multiple purposes. It provides a protective barrier around the shaft for personnel. Also it supports surficial materials which are otherwise at risk of collapse into the shaft from self weight and the surcharge of the piling rig. By pushing the casing into a stiff cohesive material, e.g. clay or weathered rock, the shaft can be sealed against ingress of groundwater from the surficial soils. If the casing does not achieve a good seal, an attempt to screw the temporary casing further into the stiff deposit can be made. If the casing is too low to provide a protective barrier, a short length of oversize casing can be used.

Base cleaning and inspection

It is essential that the base of the pile is reasonably free of debris even if the foundation has no end bearing requirement. Debris on the base can be swept up in the concrete and become entrapped as inclusions. End bearing foundations require a high standard of base cleanliness. Inspection can be carried out visually if dry with a light or with CCTV or a camera lowered to the base. Where a visual inspection is not possible, a base hardness test with a weighted tape can be used (ICE 2016, Section C3.1).

Dry poured piles

Depending on the geological conditions, it may be possible for the shaft to be excavated with relatively little groundwater ingress, and the pile may be dry concreted as is common in Auckland. However, seepages from the pile shaft will cause a gradual build-up of water. A submersible pump lowered into the shaft can sometimes be used to control inflow and allow a dry pour. Otherwise the pile must be wet poured with the tremie pipes taken to the full depth of the pile.

The definition of a dry poured pile is where the depth of water at the introduction of concrete is less than 75mm and has inflow less than 25mm in 5 minutes (Brown 2018). The rate of inflow and time taken to install the cage and tremie pipes should be considered. For dry poured piles, the tremie pipe is only provided to centralise the concrete and to limit the free fall of concrete below the tremie pipe to less than 10m. Concrete falling through a reinforcing cage or against the shaft wall may segregate resulting in defective concrete at the base of the pile. For this reason, utilizing a flexible rubber pipe attached to the end of an oscillating concrete pump is not recommended.

The high energy of the free falling concrete aids compaction of the concrete around the reinforcing cage. The speed of the concrete pour is quicker, and the top of the concrete is continually covered in the most recent and workable concrete. Consequently, concreting a dry poured pile is relatively fast, has a lower reliance on concrete workability and has a lower risk of placement defects. Water seepages down the shaft of the pile can soften the material and may facilitate smearing, potentially reducing shaft friction. The extent of this should be assessed based on experience and can be guided by CCTV inspection.

Wet poured piles

Where the rate of seepage into the pile is too high for the pile to be constructed in the dry, the pile must be poured in the wet. The EFFC/DFI Guide (2018) recommends that the foundation is filled with water from an external source to provide a positive pressure over the seepage. The wet pile tremie pour technique is designed to minimise mixing of the concrete with fluid.

The initial placement of concrete is of vital importance. The EFFC/DFI Guide (2018) refers to two methods, firstly where the tremie is sealed at the base and the concrete only contacts the support fluid when the pipe is lifted off the base of the foundation and secondly where a sliding separating medium, such as vermiculite granules is placed in the tremie pipe and concrete is poured on top of this. While the former method may minimise concrete segregation, for practicality, the latter method is preferred in the EFFC/DFI Guide (2018).

An adequate separating medium is critical. If the vermiculite plug is too thin, or the concrete falls from height on to it, the concrete may punch through. The ICE (2016) recommends the height is twice the diameter of the tremie pipe. The authors have observed vermiculite placed in a plastic bag which is thought to reduce the likelihood of concrete punching through. Inflatable rubber balls, foam balls or cylinders are also referred to in the EFFC/DFI Guide (2018) which are likely to be more robust than loose vermiculite granules, but have the risk that they may become entrapped within the foundation.

There is differing opinion in New Zealand on whether the tremie pipes should be lowered to touch the base of the pile, the tremie and hopper charged full of concrete and then lifted to initiate the concrete pour, or whether the tremie should be lowered to a short distance above the base (≤200mm) and the pour initiated. Advocates of the former method consider that charging the tremie and hopper full of concrete and then lifting causes a surge of concrete that lifts up debris on the base of the pile, improving base contact. The authors are not aware of evidence of this, but accept the mechanism is logical.

The EFFC/DFI Guide (2018) recommends the latter method, because of the risk of blockage of the tremie pipe when it is left on the base. The authors have experience this in foundations over 30m deep, refer to Figure 2. A blocked tremie pipe can be disastrous as when disassembling the tremie, concrete falls into the shaft requiring a reclean of the base and the support fluid being contaminated. Blockages are due to the pressure difference between the concrete inside the tremie pipe and the support fluid resulting in water in the concrete being squeezed out leaving a dense plug of aggregate.

The concrete workability requirements for wet piles are higher than dry piles. Consequently, even if it is expected that the piles are to be constructed in the dry, it is recommended that a wet pile mix is also developed.

Figure : Concrete blocked in a tremie pipe.

Prevention of non-conformances

Non-conformances generally fall into the following categories:

1. Out of tolerance position, or verticality of the shaft or reinforcing cage.
2. Instability of the sides of the shaft leading to soil or rock becoming entrapped in the foundation.
3. Inclusions of material other than concrete becoming entrapped during the concrete pour.
4. Water tightness issues for retaining walls. Guidance is provided in Maloney (2009).

Positional and verticality tolerance issues are generally well understood and can be prevented by following best practice guidelines and careful workmanship as per EFFC/DFI (2018) and ICE (2016).

Instability of the sides of the shaft can be challenging in certain ground conditions, particularly variable low cohesion deposits such as reclaimed fill. Guidance on assessment is provided in Huder (1972). During the excavation, minor instability of the shaft may not be noticed as the material is removed from the shaft during the excavation process. In some cases, this can be detected from the spoil. Minor instability during excavation is not necessarily an issue. However, from when the base is cleaned, any further instability will result in loose spoil either being left on the base of the shaft or cast within the body of the concrete with possible strength and durability issues. It is recommended that after the cage is inserted and before the concrete is placed, the base of the shaft is checked again and the construction time minimised. If the base is found to not be in conformance with the Specification, the pile base capacity may be down rated. Or an attempt to reclean the base should be made either by air lifting or by carefully vacuuming with the tremie pipes recirculating to the desanding plant, or to remove the cage, and clean the base again.

Major instability of the shaft after the reinforcing cage is installed results in material falling against the reinforcing cage causing durability issues and may also be a safety concern for the working platform. Despite this, it can be difficult to detect. Tools to inspect the verticality and dimensions of the shaft and detect instability under support fluid include the Fugro Sonicaliper and Koden recorder. However, the most important tool remains for the site engineers and operatives to be vigilant for unexpected deviations of the depth of the shaft and concrete, and to investigate any anomalies.

Particular care must be taken when extracting temporary casings as these can conceal voids formed when installing the temporary casing or during drilling. When the temporary casing is extracted, these voids become joined with the concrete in the bore and inclusions in the foundations can result.

The third category of non-conformance occurs in wet poured foundations and results from inclusions trapped in the concrete due to the flow of the concrete in the foundation and around the reinforcing cage. This is complex and relatively poorly understood. It results in either support fluid or the interface layer becoming entrapped within the concrete, refer to Figure 3. Due to the nature of the tremie pour process, imperfections and irregularities with deep foundations are to be expected and are not necessarily defects. For example, the surface of the deep foundation will typically be roughened by the drilling process and temporary casing.

Figure : Example of inclusion defects (left) and mattressing (right) from overseas projects.

For deep foundations cast under support fluid it is important to recognise the importance of the interface layer in the formation of inclusions. The interface layer accumulates between the support fluid and the concrete, formed by material from segregated concrete and/or support fluid and spoil, e.g. Figure 4.

Figure : Interface layer of poorly cemented sand and bentonite atop a diaphragm wall (Thorp 2018).

The layer is more dense and viscous than the support fluid and rises atop the concrete. As the concrete column rises during the pour, the concrete must displace the viscous interface layer and the support fluid. Consequently the interface layer plays a key role in the formation of defects and the tremie concrete process should be carried out in a manner to minimise its formation.

Integrity testing should be considered for all deep foundations, including retaining walls. The decision whether to include testing and at what frequency should be a collaborative discussion early in the project and consider the risk profile and local experience. It is preferable to identify potential issues during the foundation construction so that corrections can be made, rather than discovering issues during excavation or loading. Methods are described in EFFC/DFI (2018) and Brown (2018). In New Zealand, low strain impact integrity testing has been commonly adopted as it is quick and low cost, however the authors note that in some cases the method is not sensitive enough to detect significant defects within a shaft. Cross hole sonic testing is also commonly used for load bearing piles. However this method can only readily detect issues within the reinforcing cage, for example a cold joint or a large collapse, and not in the concrete cover zone where most defects due to concrete placement arise. Thermal integrity profiling or TIP testing is a relatively new technology to New Zealand that has the advantages of detecting issues quickly after placement of the concrete and also providing an indication of the quality of the concrete in the cover zone of the foundation.

Case Study of Concrete Testing for a Diaphragm Wall

The EFFC/DFI Guide (2018) provides guidance regarding concrete testing regimes for deep foundations to verify the concrete behaviour follows expected norms. A recent project involved the construction of a multi-storey underground carpark. The basement construction included a 600mm thick diaphragm wall and some associated bored piles. The diaphragm wall was excavated to a depth of approximately 19m using a combination of mechanical grab and cutter head. To support the excavations during construction the foundations were constructed under a bentonite support fluid. Concreting was completed via tremie pour using a 40MPa mix, 13mm aggregate and PFA cement replacement.

To determine the final mix design, mixing trials were undertaken with the concrete supplier to assess the workability and stability performance. The performance of tremie concrete can only be assessed through a suite of tests. Testing undertaken during the trials is listed in Table 1 below. To confirm the concrete supplied from the ready-mix plant conformed to the requirements set in the mixing trials, the concrete tests were repeated for each panel during construction.

Table 1: Concrete test methods completed.

Laboratory mixing trials indicated the mix selected was sensitive to variations in water content, which was expected due to the low water content ratio. The targeted slump range during production was between 240mm and 200mm. This is slightly higher than typical mix designs, due to concerns regarding the ability to flow around higher density reinforcing cages.

Workability test results

Slump testing showed compliance with the concrete mix design requirements. Testing results show the slumps measured were often nearer the upper limit specified. Concrete suppliers generally target workability to be in the upper range of the specified requirements as this allows flexibility onsite to account for delays due to traffic or onsite issues. Workability retention tests were carried out 4hrs after the initial slump test targeting a minimum slump of 180mm at 4hrs. Typically a 5mm to 20mm loss in workability was observed. The slump retention is important during a tremie pour, as fresh concrete is placed within the foundation will begin to interact with older concrete previously placed, increasing the risk of defects if the old concrete has stiffened.

Slump

Slump retention

Figure 4: Workability tests. Left: slump and slump retention. Right: inverted cone.

Inverted cone outflow testing results show all tests were within the project requirements and EFFC/DFI (2018). Most results are clustered between 2 to 3 seconds. The inverted cone outflow test can provide information as to the viscosity of the concrete which is fundamental to the concrete rheology.

Stability test results

Bauer filtration tests were carried out on each pour. The test is designed to simulate the water retention ability of the fresh concrete under pressure such as a deep foundation. The guide limit of 15 L/m³, was achieved, indicating the concrete mix was unlikely to be susceptible to significant bleed or segregation.

Figure 5: Bauer Filtration Test Results

ASTM bleed tests were also carried out and were less than 0.1mL/min. All results fell within the parameters of project stability requirements and this was supported by the visual inspections of no significant bleeding or segregation at the tops of the panels.

Conclusions and Recommendations

The use of larger and deep foundations is increasing in New Zealand which typically have long pour times and high structural demands. This results in additional complexity and challenges for deep foundation projects. Tremie concrete is increasingly becoming a highly engineered material with concrete additions and admixtures used to adapt the concrete to a specific application. These challenges have led to major construction issues overseas and the formation of an industry led consortium to develop guidance such as the EFFC/DFI Guide to Tremie Concrete (2018) and companion Guide for Support Fluid (2019).

The success of deep foundation projects depends on appropriate consideration of tremie concrete use through all stages of the project lifecycle and input from all project participants. There are design and construction considerations for whether bored piles are poured in the dry or wet poured, either under water or a support fluid. For wet poured foundations, the importance of initiating the concrete pour to minimise mixing of the concrete with fluid is highlighted. Processes for preventing and identifying non-conformances are discussed and practitioners are encouraged to investigate emerging tools such as Thermal Integrity Profiling for integrity testing that enables an assessment of the concrete cover zone. Detailed concrete acceptance testing as described in EFFC/DFI (2018) is recommended to assess and confirm that concrete behaviour, both workability and stability, is suitable for the foundation and meets requirements.

References

AS 2159-2009 Piling – Design and installation 2009. Standards Australia.

Brown, D.A. et al 2018. Drilled Shafts: Construction Procedures and Design Methods. FHWA-NHI 18-024. U.S. Department of Transportation Federal Highways Administration (FHWA).

Guide to Tremie Concrete for Deep Foundations 2018. EFFC & DFI.

Guide to Support Fluids for Deep Foundations 2019. EFFC & DFI.

Fleming, W.K. & Sliwinski, Z.J. 1991. The use and influence of bentonite on bored pile construction. CIRIA PG3.

Huder, J. 1972. Stability of bentonite slurry trenches with some experiences in Swiss practice. Fifth European conference on soil mechanics and foundation engineering 1972. Madrid, Spain.

ICE Specification for Piling and Embedded Retaining Walls, Third edition 2016. Institute of Civil Engineers.

Lam, C. et al 2014. Effects of polymer and bentonite support fluids on concrete–sand interface shear strength. Geotechnique 64, No. 1, 28-39.

Larisch, M.D. 2016. Modern concrete technology and placement methods and their influence on waterproofing performance of diaphragm walls. The New Zealand Concrete Industry Conference 2016. Auckland.

Larisch, M.D. 2019. Concrete defects in bored piles as a result of insufficient applications of chemical admixtures. Concrete NZ Conference 2019. Dunedin.

Maloney, M. et al 2009. Reducing the Risk of Leaking Substructure, A Clients’ Guide. Institute of Civil Engineers.

Thorp, A. et al 2018. Recent experience of tremie concrete properties and testing. DFI-EFFC International Conference on Deep Foundations and Ground Improvement 2018. Rome, Italy.