A memorable micro-tunneling project experience worth re-telling

New Chelsea RIver Siphon

Abstract:

Concurrently with the Boston Central Artery project, the need to deepen the channel depth of the Chelsea River at the North Point Fort area of East Boston caused the MWRA to replace a pressurized water syphon under the river by a deeper one. It was to be installed by means of a 1,100 feet micro-tunnel under the river, with a 72inch diameter drive allowing the insertion of a 48 inch diameter cast steel pressure water line.

Diaphragm walls were used for the construction of both launch and receiving shafts with an invert 85 feet down from ground surface. In view of occasional boulder obstacles in the till, a new Soltau RVS 800 machine powered by three planetary motors was selected to allow access to the face through the bulkhead to eventually break down the obstacle under compressed air. A hyperbaric chamber was available on site but was never used. A 72inch diameter, one inch thick wall Permalok pipe was used as the jacked pipe, the largest diameter available at the time. Although an exit seal was present at the launch shaft, no re-entry seal was planned at the receiving shaft given the uncertainty about the accuracy of reaching the target. The geological profile along the alignment was through glacial till with a coarse alluvium (240 Feet) passage in the middle corresponding to the Chelsea River bed original tahlweg. The ground cover between the bore and the river bed was approximately 35 Feet.

Once underway, jacking forces became a concern even before reaching the alluvium, following machine interruptions caused by breakdowns due mainly to the abundance of boulders; high alluvium permeability required lubrication improvements that were implemented successfully; passing over mechanical problems of the Soltau machine, the question on how to arrive in the receiving shaft in the dry with a possible 80 feet of water head on the annulus became a preoccupation for all involved. The option of receiving the machine in the wet was not acceptable due to subsequent machine scheduling on another project. The implementation of an annulus controlled dynamic backfilling with specially formulated self-hardening slurry provided the solution to an exit in the dry into the receiving shaft.

Key Words: syphon, shaft, micro-tunneling, Permalok pipe, lubrication, alluvium, talweg, self-hardening slurry.

Project site location

Access Shafts Construction:

The general contractor at the time was Modern Continental, a local utility contractor that grew exponentially with the advent of the Central Artery Project (alias Big Dig). Modern Continental undertook specialty construction disciplines such as deep foundations, inclusive of diaphragm walls, or micro-tunneling (assisted by Westcon) on top of regular heavy construction activities. These disciplines were combined for the construction of the MWRA new pressurized water syphon under the Chelsea river between North Point and East Boston as an all in-house construction performance.

Circular slurry wall is a classic method for deep shafts sinking in soils; these were installed using conventional clamshells and chisels under bentonite slurry. On the North Point side, old deep sea-shore obstructions as well as boulders on both sides created trenching difficulties that were overcome but resulted in substantial concrete over-pours and joint leakage mitigation techniques being required. Very hard till at subgrade allowed an excavation in the dry without requiring a tremie seal despite the hydrostatic head. Finally, the shafts performed well as intended.

Micro-tunneling:

Although the Soltau RVS 800 was a new slurry shield machine, Modern Continental had gained substantial experience with micro-tunneling around Boston, albeit on shorter and shallower drives. Permalok jointed jacking pipes represented a technological progress at the time and 72” was the largest diameter available. Most micro-tunneling machines are of the slurry shield type where water is used to convey the excavation spoil to the surface where a solids separation plant removes the cuttings and recirculates the water. In very pervious ground, bentonite and or polymers are added to the water to reduce outbound slurry losses and improve face stability.

Like with all new machines, breakdowns were inevitable, especially here with many more boulders than anticipated, the triple motor configuration being a local first lead eventually to the replacement of some motors and gearing while immobilizing the machine at crucial times. The till contained many more boulders than anticipated by the DBR. This caused slow progress and hardship for the machine. Lubrication with bentonite slurry was attempted during interruptions; however, with an 80 Ft of external head pressure, high lubrication counter pressures caused some lubricant leakage at the entrance seal allowing ultimately brackish water to appear. A doubling of the entry seal eliminated the leakage of brackish water and allowed the reconditioning of the annulus.

Lubrication:

High lubricant pressure causes the formation of thick mud cakes by filtration at the interface of the bore with the ground and reducing the annular space doing so and potentially increasing future jacking forces. Delays and brackish ground water would have a negative impact on bentonite filtration performance and could cause even thicker mud cakes than in a fresh water environment.

The Permalok pipe sections were 20 Ft long. Lubrication ports were initially provided every other pipe corresponding to a 40 Ft gap between lubrication ports: an optimist approach to lubrication that proved workable until mechanical breakdowns interrupted the drive; further on, getting into the pervious gravelly alluvium of the Chelsea river’s tahlweg, the loss of lubrication efficiency increased jacking forces to a level compromising the ultimate goal. An intermediate jacking station was built and installed in a record time thus making plausible the drive completion. High alluvium permeability allowed the bentonite slurry to escape through the gravel, possibly breaching the river bottom and preventing the maintenance of sufficient counter pressure to maintain the annulus’ integrity: jacking pressures increased significantly in view of being only at mid-point of the drive using more than two third of the available jacking force.

A mystery crab appeared on the shaker screen of the solids separation plant, the presence of which could either be a nice prank or a sign of a break-through to the river that could be realistic under the circumstances; the mystery was never solved but induced serious efforts in controlling the annulus: increasing the number to lubrication ports to each pipe section, formulating a much thicker lubricating slurry with very low filtration using more bentonite, polymers and other gelling additives. This was implemented all along the 240 linear feet of alluvium on the alignment before reverting to a straight bentonite/polymer lubrication at each pipe once regaining control of the annulus past the river tahlweg.

In addition, it was decided that the lubrication around the machine should be achieved independently of the jacked pipes. A direct line to the machine shield’s ports was installed and dubbed the Bentonite Express. These interventions successfully did reduce the increase of jacking forces and ultimately removed the cloud about the ability to complete the drive. With the addition of the intermediate jacking station, the boring progressed slowly but regularly and the outcome of the drive was not in doubt anymore. The friction forces of the section forward of the intermediate jacking station were maintained at a normal level thanks in part to the Bentonite Express saturating the annulus around the trailing pipe; the section behind the intermediate jacking station flirted with maximum jacking frame capability available from the launch shaft, but did crawled forward.

Exit:

It is not known whether the accuracy of the guidance system or the contractor’s lack of confidence, with such a long drive for the times, removed the possibility of reaching the receiving shaft on target through a receiving seal. The typical option in such a situation is to exit in a flooded shaft and retrieve the machine with the help of divers; this was not possible because the machine was committed to another project, this project being well behind schedule and with no available time for the flooded machine reconditioning.

Solution:

The problem was how to exit in the dry with an annulus under 80 feet of head with a potential communication with the Chelsea river; something to keep anyone awake at night.

An imaginative solution had to be devised without delaying the progress of the machine.

Since the bore was back in the glacial till, one could consider the ground as essentially impervious albeit erodible. The challenge was then to impede any ground water eventual flow forward from the Chelsea River alluvium bed through the annulus into the receiving shaft. With consideration of not increasing the jacking forces in any significant way, limited options were available to control the annulus. The final plan was to create an active annulus sealing zone 40 ft long 60 Ft away from the receiving shaft. The reasoning was that filling the annulus with a slow setting self-hardening slurry-grout that would pack the annulus with a mud cake capable of resisting the hydrostatic head while remaining plastic enough not to impact the jacking force noticeably, until the machine eventual break through, a number of days away. Injection pressures were maintained at all times, even during jacking pipe addition breaks. Injection ports were created for the train of pipes that would travel through the treatment zone, and connections were made accordingly to keep the process dynamic until breakout and machine extraction.

The fact that the injected formulation was nevertheless cementitious required a close coordination between all trades, time being of the essence, with the activities taking place over 24 /7. Based on rates of progress, the active sealing zone would start setting progressively after the 4th day. Breakthrough did occur after 3 days. The gamble that the last 60 ft of the bore forward would be consistent glacial till with no saturated seam leaving the bore unprotected was a risk Modern was willing to take. And the gamble paid off: the exit through the diaphragm wall occurred in a perfectly dry condition. The active zone was maintained until the trailing tube was extracted and the Permalok pipe penetrated through the diaphragm wall and conventional sealing and waterproofing was performed from inside the shaft.

The installation of the 1,100 LF of 48” cast steel bell and spigot pressure pipe riding on radiant spacers sliding on plastic pad was a new challenge in itself and did not go without incidents given the fragility of the spacers having to pull out on a few occasions to replace damaged spacers, 3 spacer rings per 20 ft pipe section.

Annular space backfill: The final challenge following the successful pressure testing of the service pipe was to grout the annulus between the 72” and 48” pipes. Installation of grout lines concurrently with the installation of the service pipe would have been difficult and the quality control would have limited, and impossible after insertion of the pipe.

A novel approach proposed by Modern and approved by the MWRA consisted on gravity grouting the entire length of the 1,100 LF drive by bulkheading both ends of the annulus at the shafts, providing an entry port at the invert on the launch shaft and a vent port at the spring line at the receiving shaft, both connected to a riser to the surface. The slow setting proprietary non-shrink self-hardening slurry was mixed on the surface and dropped to the bottom of the shaft through a 3” riser connected to the entry port. By flooding the annulus from one end to the other, the annulus backfilling was completed with no possibility of trapping air; once the slurry-grout emerged through the exit riser at the surface with the same density as that of the one going in, the backfilling was terminated and kept under the vertical column pressure until complete set of about one week. Although the backfilling of 500 CY could have taken place over a single long shift, given the slow setting nature of the formulation, the backfill was completed in two short shifts. The simplicity of the process carries inherently the assurance of a 100% homogenous annulus backfill.

Conclusions: The attraction of underground construction is the amount of challenges facing contractors having to respond to such challenges, be they geological, mechanical or technological. More than in other fields of construction, contractors have the leading role in deciding on the course of events that they alone can control. On the New Chelsea River Siphon, a number of first were performed with success:

first tri-motor cutting head drive micro-tunneller slurry shield in the US,
first over 1,000 LF 72” Permalok jacked pipe drive,
first dry receiving shaft entry with 80 Ft of head and dynamic annulus grouting
first annulus backfilling by gravity flooding with slow setting slurry-grout.

This project has contributed substancially to the progress of micro-tunneling technology that is of great use in today’s ever-increasing conquest of the underground space. The project was completed during the summer of 2000.

Geologist tunnel expert P. Tarkoy snapping the 72” SOLTAU RVS 800 exit in the receiving shaft.

About the author: Gilbert Tallard is a civil engineer whose career has been linked to many aspects of underground construction and environmental remediation. His specialty is drilling and support fluids, grouts and self-hardening slurries. At the time of this project, he was providing Modern with lubrication and self-hardening slurries advice and formulations.

Ref. Peter J. Tarkoy, pH.D. 2001 paper on geological and mechanical aspects of the project: Challenges and Successes in Micro-Tunneling in the Chelsea River Crossing. Bauma 2001 5th micro-tunneling symposium 5/6 April 2001.