Ground improvement techniques undertaken for the road embankments on soft soil for the Peka Peka to Ōtaki Expressway

Abstract

The Peka Peka to Ōtaki Expressway (PP2Ō) is a 13 km long, 4-lane expressway located on the Kāpiti Coast north of Wellington, New Zealand, serving as a replacement for the existing State Highway 1 (SH1). To the south, it connects with the MacKays to Peka Peka Expressway, and its northern extension reaches beyond the Taylors Road intersection on the current SH1. 

At both the southern and northern sections of the alignment, soft and highly compressible peat and organic silt up to about 7 m depth were encountered.  The presence of soft soils is a consequence of topographical and geological factors which has resulted in poorly drained conditions.

This paper provides an overview of the ground improvement techniques implemented to address geotechnical challenges associated with these soft soils, and the observational approach adopted during the construction phase.

1. Introduction

The Peka Peka to Ōtaki (PP2Ō) Expressway is a new expressway that has recently been completed in Wellington, New Zealand. It forms a part of the Wellington Northern Corridor, a series of road improvement projects along State Highway 1 (SH1) network between Wellington Airport and Levin (Figure 1). This development aimed to reduce congestion, improve safety, and support economic growth. 

Figure 1: Photo of the completed Peka Peka to Ōtaki Expressway.

Fletcher Construction was contracted by the Waka Kotahi New Zealand Transport Agency to deliver the Peka Peka to Ōtaki (PP2Ō) Expressway project under a Design and Build Contract. Tonkin & Taylor Limited has been engaged by the lead project designer, Beca Limited, to provide aspects of the project design including geotechnical investigations, interpretation, design, and construction verification. 

The PP2Ō Expressway is a 13 km long, 4-lane expressway on the Kāpiti Coast north of Wellington. It includes nine road bridges, rail re-alignments, shared pathways, and new local roads. This new route replaced the existing SH1. At its southern end, the PP2Ō Expressway joins with the MacKays to Peka Peka Expressway. At its northern end, the PP2Ō Expressway extends to the north of the Taylors Road intersection on the existing SH1 north of Ōtaki.

The Kāpiti Coast lies approximately 3 to 5 km to the west of the expressway alignment. To the east, lies the Tararua Ranges. The central section of the expressway covers largely flat terrace terrain and the northern and southern ends of the route cover rolling sand dunes.

2. Geology and geomorphology

The Kāpiti Coast is a narrow triangular-shaped coastal plain situated between the Tararua axial ranges to the east, and the Tasman Sea to the west. During the Last Glacial Maximum (LGM), the sea level was significantly lower than present and the coastal plain was dominated by broad greywacke gravel-dominant alluvial fans originating from major rivers that extended west from the foothills of the Tarauras. Following the LGM, the sea level gradually rose and stabilised close to the present-day level at about 6500 years BP. A coastal escarpment can be traced between Waikane and Ōtaki which marks the maximum landward migration of the coastline. Since 6500 years BP, the coastline has retreated, coinciding with the formation of widespread sand dune and inter-dune/swamp deposits across the coastal plain as presented in Figure 2 (Jones and Baker, 2005).

Figure 2: General Cross Section of the Kāpiti Coast Geology (Jones and Baker, 2005).

The PP2Ō expressway drops in elevation to the south of the alignment (near Te Hapua Road), where a steep alluvial fan is truncated by the coastal escarpment defining the eastern margin of a large inter-dune/swamp deposit (Begg et al., 2000). In this section, soft and highly compressible peat and organic silt were investigated to about 7 m depth. Large sand dunes further west of the alignment likely confine the western margin of the deposit. 

To a lesser extent, up to 4 m of highly compressible peat and organic silt deposits were observed on the northwest side of the Otaki River flood plain within the depressions formed at the interface between an alluvial terrace riser and tall sand dunes.

3. Geotechnical challenges 

The following geotechnical challenges were encountered during the design and construction of the road embankments.

• Soft to very soft peat and organic silt deposits were encountered at or near the surface up to 7 m deep. Where present, the organic deposits can be expected to cause embankment settlements that are much higher than in areas without soft soils. Furthermore, the variable thickness and highly variable compressibility of these soft soils can cause significant differential settlements over relatively short distances along the alignment. 
• The undrained shear strength of the soft soil deposits typically falls within the range of 10 to 20 kPa. The road embankment, reaching up to 4 m in height, may present a potential risk of slope instability and bearing capacity failure.
• The road embankment is situated near the existing SH1, posing challenges in both the design and construction of the ground improvement technique to ensure it does not have a negative impact on the pre-existing state highway.

4. Design solution

During the tender design process, an evaluation of geotechnical challenges was undertaken in collaboration with the construction team. The assessment of the viability of ground improvement options considered factors such as the volume of excavation and backfill materials, the potential for re-use of undercut materials, the removal and subsequent re-use of surface fill materials, and the timing of embankment surcharging.

A consensus was reached that in areas with shallow deposits of highly compressible peat and organic silt, the preferred ground improvement technique will involve undercutting and replacement. In regions where the peat and organic silt deposits were more substantial, surcharging the fill embankments emerged as the preferred method to expedite the consolidation process.

The design of the ground improvement works was informed by extensive laboratory testing and high-quality monitoring data from the nearby Mackays to Peka Peka Expressway (Coe and Alexander, 2012). The availability of this high-quality data in geologically similar materials provided confidence in our assessment of settlement durations and magnitudes. 

Estimates for undercut values, surcharge heights, and durations were provided to the construction team. This enabled the construction team to develop a viable earthworks plan that considered both the construction programme, and the volume of undercut and backfill materials.

At the northern end of the alignment, the re-alignment of the North Island Main Trunk (NIMT) rail line and the construction of bridges across the new expressway were either on or very close to the critical path. In these areas, the preferred method was undercut and replacement.

At the southern end of the project, the thickness and depth of the peat material did not allow for undercut in certain locations, necessitating the use of surcharge embankments. On the local arterial roads where timing was important, higher surcharge heights were employed to expedite settlement (typically 5 m high embankment for 3 to 4 months). On the main alignment, where there was more flexibility in the construction schedule, lower surcharge heights were utilised (typically 3 m high embankment for 6 months).

Close coordination with the stormwater and construction teams was required to manage the stormwater features during the construction of the embankments. The temporary diversion of a stream around one of the surcharge embankment eliminated the need for a temporary culvert during the surcharge period. The permanent culvert was constructed after the completion of the surcharging.

5. Undercut and replacement

In areas where shallow soft soil deposits extend up to approximately 4 m in depth, the preferred method involved undercutting these materials to reach the underlying sands or gravels. Subsequently, the excavated material was replaced with compacted site-won structural fill. This approach was favored for its efficiency in construction, mitigating the potential for future settlement issues. Moreover, it provided an opportunity for the repurposing of the excavated soft soil material for landscaping purposes. By mixing with sand, the peat was able to be used as topsoil.

To better quantify the proposed undercut and replacement volumes, simple three-dimensional (3-D) geological models were prepared and used by surveyors to assist with verifying the excavation limit. 

Given the anticipated ingress of water into the undercut excavation, measures were put in place to enable proper placement of the structural fill in successive lifts starting from the excavation base. In cases where water inflow into the excavation proved uncontrollable, a clean gravel fill layer was used as a backfill, maintaining a suitable elevation above the standing water level. Above the ground water level sand fill was placed and compacted in layers.

Figure 3: Example 3-D geological model of the proposed base of undercut (Seequent Leapfrog).

Figure 4: Undercut and replace at the southern end of the road alignment.

6. Surcharge embankment

6.1 Design 

Surcharging became necessary in locations where excavation to depths was expected to exceed 4 m depth, making it unsafe due to the inherent instability of the low-strength peat material, especially in close proximity to the existing state highway with active traffic lanes.

 Prior to embarking on the surcharge embankment design, thorough consideration was given to the following factors:

• Specifying acceptable settlement and differential settlement tolerances for the completed pavement
• Defining the boundaries and extent of the embankment
• Defining the extent of the soft soil material
• Assessing the embankment’s impact on the nearby state highway
• Determining the rate at which embankment fill material could be placed
• Evaluating the availability of on-site fill materials
• Stormwater diversions and culvert construction
• Construction staging and programme

Areas identified where surcharging was carried out are summarised in Table 1. Figure 5 presents the aerial photo of the Awatea Stream Stage 1 surcharge area.

Table 1: Surcharge areas for embankment over soft ground 

Figure 5: Awatea Stream – stage 1 surcharge area.

The timing was critical for stage 1 of the Awatea Stream surcharging, situated along the local arterial roads. To accelerate settlement, larger surcharge heights were implemented, with embankments reaching up to 5 m, and surcharge duration designed for up to 3 to 4 months. For stage 2 of the Awatea Stream surcharging along the main alignment, there was greater flexibility in the construction schedule. During this phase, shorter surcharge heights were employed, with embankments reaching around 3 m in height, and the surcharge period extended to approximately 6 months.

6.2 Peat and organic silt consolidation parameters

The settlement data obtained from the trial embankment, in conjunction with laboratory testing results conducted for the MacKays to Peka Peka Expressway, as well as the available in-situ ground investigation data for the Peka Peka to Ōtaki Expressway, were utilised to establish the consolidation parameters for the peat and organic silt material.

Coe and Alexander (2012) have previously presented comprehensive details of the trial embankment, and a summarised description is provided here. The trial embankment, constructed for the MacKays to Peka Peka Expressway, covers an area of approximately 20 m in length and 20 m in width. The depth of peat deposits within the trial site varies between 1.5 m and 2.5 m. The fill material was placed in several layers, reaching a maximum height of 2.75 m over a period of 3 months. This fill remained in place for a duration of 10 months.

The consolidation parameters for the Peka Peka to Ōtaki Expressway and the MacKays to Peka Peka Expressway projects are presented in Table 2 for comparison.

Table 2: Derived parameters for the peat/organic silt deposit 

The MacKays to Peka Peka Expressway consolidation parameters were derived by utilising the data from a trial embankment conducted for that project, as well as data from a trial embankment from the Raumati Straight Widening Project (Opus, 1999, as cited in Coe and Alexander, 2012), and construction records from the MacKays Crossing Project (Palmer, 2010, as cited in Coe and Alexander, 2012).

6.3 Embankment loading

Given the relatively low shear strength of the peat and organic silt, a progressive fill construction approach was adopted. This method involves the incremental placement of fill material in relatively thin layers, typically ranging from 0.5 to 1 m per week. This phased construction allowed time for the underlying foundation soils to consolidate and strengthen before subjecting them to the next increment of load.

The implementation of the staged construction technique necessitated close collaboration and communication among the project team including the design engineer, constructor, and supervising engineer. In addition, rigorous monitoring of ground instrumentation, including vibrating wire piezometers, settlement markers, profilometers, and inclinometers, played a pivotal role in overseeing the embankment’s performance throughout the construction process.

7. Ground instrumentation and monitoring

7.1 Monitoring plan

The surcharge embankments were monitored to measure the total settlement, settlement rate, lateral displacement, and excess pore water pressures. This monitoring employed a combination of profilometers, settlement plates, settlement stations, vibrating wire piezometers, and inclinometers to assess the surcharge embankment’s performance during construction. These data were then used to inform the necessary back-analysis of soil parameters and the fitting of monitored data curves.

Settlements were monitored along SH1 prior to any surcharging. Settlement stations at 20 m intervals along the verge of SH1 were installed. Settlement of the North Island Main Trunk (NIMT) tracks was also monitored prior to Awatea Stream stage 2 and Mary Crest Basin surcharging. Rail settlement pins were installed at 20 m intervals. Figure 6 provides a snapshot of the ground instrumentation plan for the Awatea Stream stage 2 surcharge area.

Figure 6. Snapshot of ground instrumentation plan (Awatea Stream Stage 2 surcharge area).

7.2 Monitoring of pore pressure

Vibrating wire piezometers were strategically positioned at multiple locations and depths between 3 m and 7 m below ground level. These instruments continuously recorded pore water pressures as surcharge loading stages occurred. 

7.3 Measurement of lateral movement of foundation soil

Inclinometers were installed near the base of the embankment to observe the lateral movement profile of the subsoil. These instruments not only tracked the magnitude of movement but also identified the locations of maximum movement. A representative graph of horizontal movement versus depth is presented in Figure 7 for reference.

Figure 7: Inclinometer readings at Awatea Stream – Stage 1 surcharge Area.

7.4 Measurement of settlement

Profilometers and settlement plates were installed to observe subsoil settlement within the surcharge zones. These instruments delivered valuable data concerning settlement throughout the surcharging phases, thereby offering insights into the speed and extent of settlement. This information proved instrumental for designers, enabling them to conduct back analysis and re-calibrate ground models, facilitating assessment regarding hold period completion. Figure 8 showcases the recorded settlement data for the ten settlement plates. Figure 9 features a representative graph illustrating the observed settlement data from a profilometer in the Awatea Stream stage 1 surcharge area.

Figure 8: Settlement plate readings at Awatea Stream – Stage 1 surcharge Area.

Figure 9: Profilometer readings at Awatea Stream – Stage 1 surcharge Area.

8. Geotechnical quality control plan

In order to effectively mitigate the geotechnical risks associated with embankment surcharging, a Geotechnical Quality Control Plan was developed. This plan encompassed a series of geotechnical hold points so that the surcharge removal only proceeds upon Designer’s verification that the predicted post-construction total and differential settlements meet the Principal’s Requirements. This validation process relied on back analysis and surcharge release assessments.

The outcomes of the back analysis play a crucial role in verifying the surcharge duration and give confidence in the long-term performance of the embankment. The back analysis process involved the following steps:

• Reviewing the actual soil stratigraphy, groundwater level, and the total depth of soft soil.
• Examining the actual fill height and construction history.
• Performing a settlement analysis using an updated ground model, fill height, and construction history.

In cases where the predicted settlement-versus-time curve and excess pore pressure-versus-time curve did not align with the monitoring data, the analysis model was refined. This was achieved through iterative adjustments of compression ratios (CR = Cc/1+e_o), re-compression ratios (CRR = C_s/1+e_o), rates of consolidation (C_v and C_h), and creep strain rate (C_ae = C_a/1+e_o) until the predicted curves reasonably matched the monitoring data.

Review the projected post-surcharge settlement and verify that it aligns with the Principal’s requirements.

Back analyses were conducted using the finite element software PLAXIS 2D to refine the settlement and excess pore water pressure-versus-time curves, as depicted in Figure 10. Figures 11 and 12 display a typical plot of the predicted settlement-versus-time, and excess pore pressure-versus-time, in conjunction with monitoring data.

Figure 10: Snapshot of settlement analysis using the finite element software, PLAXIS 2D for Awatea Stream – stage 1 surcharge Area.

Figure 11: Typical plot of the predicted settlement prior to surcharge removal for Awatea Stream – stage 1 surcharge area.

Figure 12: Typical plot of the predicted excess pore water pressure prior to surcharge removal for Awatea Stream – stage 1 surcharge area.

Table 4 displays a comparison between the actual surcharge durations and the design durations.

For the Awatea Stream stage 1, the detailed design projected an expected total settlement of 1390 mm when employing the surcharging method. To meet the long-term settlement criteria, it was established that the surcharging period should span 3 to 4 months. However, during the construction phase, settlement measurements indicated values of up to 1300 mm during the surcharging process. Following a back-analysis, where settlement and excess pore water pressure-versus-time curves were matched with monitoring data, and consolidation parameters were adjusted accordingly, the decision was made to prolong the surcharge duration to 4.5 months, to provide confidence that the design requirements would be met.

Table 4: Surcharge areas for embankment over soft ground

9. Conclusion

The successful implementation of the ground improvement techniques for the road embankments over the soft and highly compressible peat and organic silt at Peka Peka to Ōtaki Expressway was due to the following factors:

• Constructor’s early involvement in the tender design to help select ground improvement methods. The adopted surcharging and undercutting methods were selected based upon both the site-specific ground conditions and the construction requirements.
• Utilising extensive subsurface investigation data, laboratory testing, and good quality embankment fill monitoring data from the adjacent MacKays to Peka Peka Expressway in the design of the ground improvement works allowed confident assessment of the consolidation settlement magnitude and duration during the design stage.
• Surcharge construction and duration monitoring involved an observational approach and included review of the instrumentation data. Back analysis and curve fitting have been undertaken to assess the time for surcharge removal and estimate the post-construction settlements and differential settlements.

10. Acknowledgements

The authors would like to thank Beca Limited, Fletcher Construction, and the Waka Kotahi New Zealand Transport Agency for granting permission to publish this paper. The authors would also like to acknowledge the previous work completed by the M2PP Alliance comprising Beca Limited, Fletcher Construction, Higgins, Kāpiti Coast District Council, and the Waka Kotahi New Zealand Transport Agency for the MacKays to Peka Peka Expressway project.

References

Begg, J.G., and Johnston, M.R. (compilers) 2000. G7. “Geology of Wellington area”. Institute of Geological and Nuclear Sciences 1:250,000 geological map 10, 1 sheet + 64 p. Lower Hutt, New Zealand: Institute of Geological and Nuclear Sciences Limited.

Coe, L.J., and Alexander, G.J. (2012). “MacKays to Peka Peka Expressway: Road Embankment Construction on Peat Deposits”. Proceedings of the 11th Australia New Zealand Conference on Geomechanics, Melbourne, Australia.

Jones, A., and Baker, T. (2005). “Groundwater monitoring technical report”. Greater Wellington Regional Council, Publication No. GW/RINV-T-05/86.