The $NZ85 million project comprises a new pump station and 5 km of 1,800 mm diameter pipeline with a design capacity of 6 cubic metres per second. The pipeline comprises 2.3 km of concrete pipe installed by microtunnelling and 2.7 km of polyethylene (PE) marine pipe installed by dredge and lay.
Project background
Sewage is collected in a dedicated network that terminates at the Christchurch Wastewater Treatment Plant, where primary and secondary treatment is undertaken. After the main pathogen reduction has occurred the wastewater is currently discharged into the Avon-Heathcote estuary, where it flows out to the coastal area of Pegasus Bay.
The community had significant reservations about wastewater discharging into the Avon-Heathcote estuary, as recreational water quality in the estuary was not being met at all times. It was decided that an ocean outfall to discharge 3 km offshore would be the best solution to rectify the problem. Consent for the project was granted in November 2005, with a requirement for the project to be operational by 30 September 2009.
Pipeline contract
The contract awarded was split into the “÷land-based’ pipeline section and the marine pipeline section. The land-based section of pipeline comprised an 874 m drive beneath the Avon-Heathcote estuary to a pump station, a 604 m drive through South New Brighton Park and along Jellicoe Street to the sand dunes, and a 830 m drive from the sand dunes to a point about 600 m offshore, beyond the surf zone. The land-based pipeline will connect to the marine pipeline at this point beyond the surf zone, where the depth of water at low tide is approximately 6.5 m and the cover to the pipeline is about 3 m.
The successful tender from McConnell Dowell Constructors comprised approximately 2,300 m of 1,800 mm ID concrete jacking pipe being installed in three drives from the pump station and connecting with the marine pipeline section at the point beyond the surf zone offshore. The microtunnel boring machine (MTBM) would be recovered using a “÷wet’ marine recovery technique from a barge.
Site investigations and pipeline design
Site investigations, hydraulic analysis and pipeline detailed design were carried out by URS.
The site is located within a marginal marine setting, with the project corridor crossing estuarine, dune, beach and shallow marine environments. Investigations were undertaken to collect information required for the design of excavations along the pipeline route and at three proposed shaft locations. The onshore investigation comprised of test pits, cone penetrometer tests, piezocone penetration tests with dissipation tests, dual tube push sample holes, dynamic cone penetrometer tests, laboratory testing, water level monitoring and chemical analysis of water samples.
The subsurface investigations indicated that four distinct geological materials were present along the pipeline route. The microtunnel pipeline was expected to lie within a zone defined as dense to very dense, uniform fine beach and dune sands, with some thin medium dense fine sand or silty sand lenses. Some shell layers were also expected.
Pipeline design
Three locations were identified for microtunnel shafts: at the pump station interface, in South New Brighton Park and in the sand dunes.
The vertical alignment of the overall pipeline was governed by the minimum depth of cover requirement at the diffusers and minimum cover of the microtunnel pipeline beneath the deepest section of the estuary. Depth of cover through the majority of the microtunnel section is approximately 8 – 9 m, however a minimum cover of two pipe diameters or about 4.2 m was chosen beneath the estuary to ensure that adequate face stability and alignment control of the MTBM could be maintained should soft sediments be encountered.
Rubber ring jointed reinforced concrete (RRJRC) jacking pipes were specified. With jacking forces tending to govern the pipe structural design, the contractor was responsible for the design of the pipe and jointing system within the following limits to satisfy the hydraulic design:
- Pipe internal diameter: 1,740 mm – 1,800 mm,
- maximum working pressure: 290 kPa,
- factory test pressure: 420 kPa (subsequently reduced to 370 kPa due to larger pipe size), and
- corrosion protection: 35 mm external cover for marine conditions. Internal lining was not deemed necessary because of the high quality of the treated wastewater and lack of potential for sulphide attack
.
Settlement monitoring
Excessive settlement for the microtunnel drive from Jellicoe Street had the potential to damage existing sewer pipeline, water supply services, residential property and be a hazard to public and traffic. The cover along Jellicoe Street varied from about 9 – 11 m. The existing 150 mm diameter concrete sewer is buried about 1.5 m below the road surface directly above the microtunnel pipeline, with laterals at each property.
Given the potential to damage the existing sewer or property, the contractor was required to implement a settlement monitoring program with readings taken prior to, during and after the MTBM passes defined monitoring points along the street.
Settlement thresholds were set at 12 mm on centreline, 6 mm at 5 m off centreline and 3 mm at 10 m off centreline, which if exceeded, triggered an increase in monitoring frequency and a corrective action plan. These threshold levels were never triggered.
Construction of the microtunnel sections
The 2,308 m of tunnel was split into three drives:
- Drive 1 – 874 m from the pump station at Construction Management Area 1 (CMA1) to the park (CMA2),
- Drive 2 – 604 m from the park to the beach (CMA3) beneath Jellicoe Street, and
Drive 3 – 830 m from the beach through the surf zone and out to the connection point offshore with the dredge and lay marine pipeline.
The drives were all straight, with a down gradient of 0.3 per cent from the pump station. At the park shaft at CMA2, there is a 30 degree change in direction.
CMA2 was selected as the main site staging area, and this site was used as the launch shaft for Drives 1 and 2, necessitating the turning of the jacking rig.
Shaft construction
The shafts were all constructed as sheet piled pits, with CMA1 and CMA3 shafts being rectangular 6 m x 12 m shafts, while CMA2 was a circular shaft, 15 m in diameter. Sheet pile types AZ34 or AZ36 with a length of 24 m were used for the shafts.
The CMA1 shaft was constructed adjacent to a large sheeted pit excavated for the pump station construction, where well points were used to dewater the excavation. The shaft excavation therefore had to be dewatered and constructed in the dry to avoid any adverse effects on the adjacent works.
The CMA2 and CMA3 shafts were constructed partially “÷in the wet.’ They were initially excavated in the dry using a sump pump to lower the water table to about 8 m depth and walers fitted before being allowed to flood, with the remaining excavation being carried out with a clamshell grab. Final excavation trimming was carried out by divers using airlift equipment, the divers then coordinated the placement of prefabricated reinforcement cages and tremie poured concrete for the base slab. The shaft was then able to be pumped out with the minor leakage remaining being within the capacity of a small sump pump.
Wet recovery of the MTBM
For Drive 3, McConnell Dowell has chosen to recover the MTBM using a “÷wet’ recovery method rather than constructing a recovery shaft. To achieve this, the MTBM has been driven through to the recovery point where cover to the seabed is approximately 3 m.
Following completion of all disestablishment, cleanup and pressure testing of the concrete pipeline, which will be done with the MTBM still in place, the dredge will then excavate around the MTBM and buoyancy tanks will be fitted.
A wet recovery bulkhead door will be sealed at the back of the MTBM, which, after the flooding of the pipeline, will then be jacked free of the pipe and floated to the surface for recovery.
Once the MTBM has been removed, a specially designed mating piece will be installed by divers, to make the connection between the RRJRC jacking pipes and the PE marine pipeline.
Bentonite lubrication system and intermediate jacking stations
The slurry MTBM utilised a closed circuit slurry circulation and muck separation system located on the surface. It comprised a number of slurry tanks, which were dosed with bentonite and pumped to the MTBM cutterhead, providing face support and a means of transporting excavated material to the surface for removal. The slurry was transferred between the MTBM and separation plant through 150 mm diameter pipes and slurry pumps.
As the three tunnel drives were relatively long at 874 m, 604 m and 809 m respectively, the minimisation of jacking forces was of paramount importance.
In addition to mixing/holding tank/pumping system for bentonite production and delivery, a PLC controlled automatic injection control system regulated the cyclical injection of lubricant into injection nozzles positioned along the pipeline. During the jacking operation the bentonite fluid was injected into each nozzle for a fixed time, following which injection shifted to the next nozzle in sequence. In this way the pipeline was surrounded by lubricating fluid. Rate of injection was adjusted relative to the advance rate such that the theoretical annulus was filled.
As a contingency measure, intermediate jacking stations, comprising 14 hydraulic cylinders each capable of exerting a push force of 72 tonnes were installed at approximate 110 m spacing.
Intermediate jacking stations were installed in each drive every 115 m to provide the ability to jack separate sections of the pipeline in the event that jacking loads became too high. These did not need to be activated for Drives 1 and 2, but up to three of the intermediate stations were activated in Drive 3.
Drive 1 construction
The 874 m push across the estuary commenced on 23 May 2007 with the MTBM pushing through the tunnel eye in the CMA2 shaft. Over the first 200 to 300 m of drive some variable ground conditions were encountered that resulted in difficult driving and variable progress. The variable ground conditions encountered included shell formations, cobbles and timber. Most of the timber encountered came through the slurry system as finger-sized rounded fragments which, like the shells, probably originated from the 6,500 year old beach formation.
There were some instances where larger fragments of timber had been chewed by the cutterhead, and the cutterhead also jammed on a number of occasions on what was believed to be larger driftwood type branches. As a result of the timber the cutterhead had to be reversed at times to advance the machine and several times the slurry flow also had to reversed to clear the cutterhead ports.
Once under the estuary, a higher silt content in the face was periodically encountered that slowed down the extraction of fines in the separation plant, which in turn resulted in a slowing of MTBM penetration rate.
Tunnel invert level varied from 8.4 m below existing ground level at CMA2 to 13.5 m at CMA3. The lowest point of cover was under the main channel of the estuary where cover was about 4.2 m. The water table was within 1 m of ground surface. During the first quarter of the drive, a number of sinkholes appeared above the tunnel line.
The maximum jacking force required on Drive 1 was 600 tonnes, which was after a 72 hour stoppage to reinforce a damaged launch seal. Typical jacking loads for most of the drive were 300 to 400 tonnes, with the jacking force increasing gradually with drive length, as would be expected.
Notwithstanding the variable conditions, the 874 m long drive successfully holed through into the CMA1 shaft approximately 15 weeks after starting, at an average rate of about 7.5 m per day.
Drive 2 construction
As a result of the sinkholes occurring in Drive 1, there was considerable concern by the council over the potential for sinkholes along Drive 2 in Jellicoe Street and the effect these may have on public and traffic safety, property and buried services, including the 150 mm existing sewer line above the pipeline. It was decided that additional site investigations were necessary to determine whether variable conditions as encountered in Drive 1 would be present and an emergency contingency plan was prepared.
The site investigations comprised additional cored holes, cone petrometre tests, sampling and grading. One hole near the start of the drive indicated relatively high silt content, but apart from that, the investigations showed a more uniform sand profile than encountered in Drive 1, with only minor traces of shells and timber.
McConnell Dowell chose not to use the D-Mode feature on the MTBM for Drive 1. While it was not clear whether this would have improved the face stability and prevented the sinkholes, the decision was taken to use D-Mode on Drive 2.
The 604 m long Drive 2 commenced on 11 October 2007 and was successfully completed after approximately 7 weeks, with an average rate of 12.5 m per day. Jacking loads increased from around 80 tonnes to 200 tonnes at completion, well within the 850 tonne capacity of the main jacking station.
Importantly, the drive was completed without sinkholes or any other incident and settlement monitoring was well within the design parameters.
Drive 3 construction
Drive 3 construction commenced from the dunes shaft at CMA3 on 3 January 2008, after the separation plant and other facilities were relocated from CMA2. The original drive length was to be 830 m, however as the marine trench dredging had commenced, McConnell Dowell made the decision to stop the drive at 808 m to ensure that the MTBM and the leading pipes remained in stable ground.
The drive was completed on 4 February 2008 after an average rate of 25 m per day, on two 12 hour shifts, seven days a week. The best day was 45 m. Ground conditions were excellent, with very little silty material being encountered and the penetration rates reflected this.
Average penetration rates increased from about 70 mm/min to a maximum rate of 120 mm/min at 700 m. The penetration rate then reduced and the jacking loads increased significantly.
The jacking loads increased to over 600 tonnes at about 700 m, prompting the activation of up to three of the intermediate jacking stations. A gelling agent was used in the bentonite to counter the flocculating effect of saltwater but the injection rate of the bentonite along the pipeline was not increased with the increasing penetration rate. The higher jacking loads were therefore attributed to reduced lubrication along the pipeline.
D-Mode was not used in Drive 3 because ground conditions were expected to be uniform, cover was high and there were no consequences of sinkholes or settlement.
Conclusions
Microtunnelling proved to be the most economical option for the Christchurch City Council Ocean Outfall project, as well as the option with the least environmental and community impact. The project was the first time microtunnelling has been used on this scale and over this length of drive in New Zealand.
Once into construction the microtunnel operation presented the contractor with a number of difficult challenges, including variable ground conditions, a drive beneath an estuary with minimum cover, noise effects on the community, managing risks along Jellicoe Street and the final drive beneath the surf zone to the wet recovery of the MTBM.
The three drives were all successfully completed without any significant incident.
The project has demonstrated that pipe jacking of drives of this length is well within the limits of current technology.
The ocean outfall project has also confirmed the many benefits of choosing the microtunnel option over conventional dig and lay construction as well as clearly demonstrating that in a situation such as this, it is a cost competitive solution.
This article is a summary of the paper by Ron Fleming, John Moore and Gwyn Jones entitled Microtunnelling the Ocean Outfall Pipeline, Christchurch, New Zealand. The paper was presented at the Australian Tunnelling Society conference in May this year.