Tokyo Bay highway engages eight mega TBMs 01 Aug 1994

Fig 1. Plan and section of the 15km long tunnel/bridge highway

As a civil engineering project, the Trans-Tokyo Bay Highway is an enormous undertaking. The 15km long, estimated ¥1,450 billion (approximately ¥100 = US$1 in July 1994) (about US$15 billion) project comprises a combined 10km long bored tunnel and a 5km long high- and low-level bridge under and over the bay with one man-made island providing the tunnel/ bridge transition and another in the middle of the bored tunnel to provide a ventilation station (Fig 1). When complete, the link will shorten the current 2h 45min journey around the bay (depending on the traffic congestion) to just 45min. It will take some 15min to drive through the 10km long tunnel.

The original plan was for two 5km bridge sections either side of a central 5km long immersed tube tunnel but this impinged severely on Tokyo Bay's busy international shipping lanes. A bored tunnel reduced this impact significantly. It also overcame bridge clearance problems imposed by nearby Haneda, Tokyo's second international airport on the west side of the bay.

Fig 2. Trans-Tokyo Bay is part of a larger highway strategy

Tunnelling all the way proved non cost-effective. The less technically demanding combined high level/low level bridge over the less heavily trafficked waters of the east channel is less expensive than the tunnel alternative. The bored tunnel was said to be estimated at almost the same cost/m as the immersed tube and both were estimated at more than double the cost/m of the east channel bridge.

Work on the projected 8 year and 8 month (104 month) construction programme started in July 1987 with an official opening date set for March 1996. Trans-Tokyo Bay is unusual in that it will be the first public infrastructure project in Japan to rely substantially on private finance. In October 1986, the Trans-Tokyo Bay Highway Corporation (TTB) was established as a privately financed corporation to work in conjunction with the Japan Highway Public Corporation (JHPC) to raise the necessary capital and build the project.

JHPC is using part of its government funding to finance and build the necessary junctions of the TTB with existing road networks at each landfall. But at ¥214.5 billion, this is a small part of the overall funding requirement (Table 1). The remaining ¥1,224 billion is to be raised by TTB.

Of the TTB ¥90 billion working capital, about a third is invested by the 366 TTB company shareholders which include most of Japan's private insurance, steel, manufacturing and construction companies as well as the private railway company. Most of Japan's largest manufacturing and construction companies are also directly involved in the project.

Another third of the working capital is contributed by the JHPC with the final third provided by the seven prefectures and local and city government bodies, which will benefit directly from the link. Another ¥583.7 billion is raised through government bonds while ¥175.2 billion is to be raised as official loans with a 50/50 mix of government and private capital and a low interest rate of about 3%. The remaining ¥375 billion is to be raised by TTB as higher interest private loans and loans from the Japan Development Bank whose interest rates vary from about 5-8%.

Table 1. Project budget in billions of Yen

Table 2. Financing arrangements

All loans will be serviced and investments paid back through toll revenues which will be set by the JHPC. With a traffic forecast of 33,000 vehicles per day in the first year and rising to a projected 64,000 per day after 20 years, the pay back period of the project at current toll charges is calculated at about 30 years.

Originally, it was believed that all the necessary funding, private and public, could be raised in Japan. However, several overseas investors have come forward to take up TTB bonds since they became available. To early August 1994, TTB had secured funding for some 40% of the total ¥1,224 billion construction cost.

Unfortunately, after celebrating start of construction in May 1989, progress slipped behind programme and the original March 1996 opening date was rescheduled for more than a year later in early 1997. Work however has progressed rapidly since then and by early August, some 50% of the construction was complete with the eastern channel bridge ready for use by the end of 1994 and all eight TBMs expected to be boring by early 1995.

Fig 3. Kawasaki man-made island supported by 119m deep slurry diaphragm wall panels

With a project of the scope of TTB, the Japanese have drawn on their renowned wealth of research and development, pooling the experiences and expertise of manufacturers, contractors, civil designers and clients alike into various technological associations to collectively meet present challenges and grow toward facing the next. Realisation of the project will be a collaboration rather than a competition among the largest construction companies in Japan, most of whom are also shareholders in the privately financed TTB corporation.

Each element of the project holds distinction, but it is the bored tunnel section which pushes most at the edges of known technology. At 10km long and 14.14m o.d., Trans-Tokyo Bay will be the largest and longest undersea road tunnel in the world to date. In addition to the demands this represents for the design, ventilation and operation of the completed facility, its excavation and construction is extremely challenging.

The two, 2-lane tunnels pass under the sea in very soft saturated soils. They will be subjected to high water pressures both during and after construction and there is the constant high probability of having to endure strong earthquakes.

TBM tunnelling

To accommodate two traffic lanes each, the two parallel tunnels have an internal diameter of 11.9m. They are lined with a primary 650mm thick bolted and water sealed precast concrete segmental lining and finished with an inner 350mm thick in-situ concrete lining with a waterproofing membrane in between. The primary lining is designed to withstand all loads while the secondary lining improves water-tightness against the high projected water pressures, supplements the weight of the tunnel to resist buoyancy, and will protect the primary lining in the event of an in-tunnel fire. This double shell lining gives the tunnels an outer diameter of 13.9m and the TBMs an enormous 14.14m o.d. The primary segmental lining comprises 1.5m wide rings of 11 identical segments and a large wedged key segment. The 650mm thick segment is reinforced with 200-280kg of steel/m³ and weighs about 10 tonne.

One of the eight enormous 14.14m o.d. slurry TBMs that will excavate the twin-tube 10km long undersea tunnel

The geology under the bay comprises layer upon layer of soft wet soils ranging from very soft and weak alluvium clays at the top down to harder clays, clays containing more sand, and layers of diluvium. There are no rocks or boulders in the soft soils and bedrock is up to 700-800m deep. The permeability of the material ranges from the most impermeable in the soft alluvium clay at the top to very permeable in the lower diluvium deposits.

To stay within the most favourable tunnelling material, the tunnel lies just one tunnel diameter (about 15m) below the bed of the bay. For these reasons the slurry tunnelling technique was chosen for all eight TBMs. The large diameter and the minimum cover were said to be the main reasons against using the EPB technique. There is also a longer track record in Japan of large diameter slurry rather than EPB TBMs.

Of the eight TBMs, three each are supplied by Kawasaki and Mitsubishi with IHI and Hitachi supplying the other two. Another three companies, Komatsu, Hitachi and Mitsui, are supplying the ancillary equipment including four large slurry separation plants to support two TBMs each. Each TBM will be operated by a different three-partner joint venture. The principals of these 24 companies include Kajima, Taisei. Nishimatsu, Tobishima, Obayashi. Maede, Kumagai and Shimizu.

Fig 4. Each 11.9m i.d. two lane tunnel has a waterproofing membrane and a final in-situ concrete lining

According to the project schedule, the tight three year tunnelling programme is based on an expected TBM penetration rate of about 30mm/min and an advance rate of 6 m/day working four tunnnelling crews on a 7 day/week, 24h/day schedule. To maintain advance rates and increase safety in the working environment, the massive precast concrete segments will be erected by segment erector robots. The totally automated systems will not only pick up and place each segment but will also insert and tighten the 110 straight bolts in each ring. These corrosion-protected bolts will remain in place as part of the permanent works. It is expected to take between 2 and 4 hours to build each ring.

Japan's location in a highly seismically active zone presents a major design consideration for the project. Segmentally lined tunnels are known to withstand the shock waves of earthquakes remarkably well but, nevertheless, on these large road tunnels, the Japanese have incorporated a flexible joint in the segmental lining. This will help the tunnels with their more rigid steel reinforced in-situ concrete secondary lining to accommodate the movement. The flexible joints are special 1.5m wide rings which will allow up to 59mm of shear deformation. There are eight flexible joint sets in the tunnels, four in each tube. These are located close to the vertical ventilation shaft junctions where differential movement is most critical. The flexible joints and a rubber shock absorber fitted to each segment will also accommodate settlement movement of the tunnels in the soft material.

Core excavation of the Kawasaki man-made island

All eight TBMs will operate some 60m below sea level (20-25m of water and another 15m of overburden). To cope with the potential 6 bar pressure, the 13.5m long shields are fitted with four rows of wire brush tail seals. Each machine is fitted with an enormous main bearing. The Hitachi Zozen TBM is fitted with a 45tonne, 7.2m diameter single piece slewing bearing supplied by RKS SA-SKF of France. The others are fitted with 8.3m diameter, 40 tonne bearings from Roesch Rothe Erde of Germany.

Each machine is powered by 20 electro/hydraulic motors of 75kW each and has a total thrust capacity of 24,000t exerted via 48 x 2550mm stroke thrust rams. The 14.14m o.d. cutterheads rotate very slowly at 0.39rev/min maximum (0.1-0.2rev/min normally) and the maximum torque is 3250 tonne metre. The slot gates on the drum type cutterheads can be controlled from completely closed to fully open. The tools on the cutterheads have been especially treated in efforts to extend their life and limit the need to execute man-entry inspections and replacements. Any man-entry would require ground freezing.

The four head-on undersea junctions will be accomplished using ground freezing. The 1m thick collar of frozen ground will extend about 5m along the junctioned shields reducing to about 3m wide at the outer perimeter. Each TBM is designed and manufactured with the freezing facility equipment built in. There are 40-50 ports in the shields' bulkheads and compressed air can also be applied in the excavation chamber.

Prior to launch of the TBMs, construction of the two man-made islands and the earth works at each landfall has progressed. All four sections required extensive consolidation of the soft soils, many kilometres of complex diaphragm wall construction, and the placing of vast amounts of fill material. Perhaps the most impressive structure is the 189m diameter Kawasaki island (Fig 3).

Eight 1.5m wide flexible joint rings are inserted into the primary segmental lining to help the accommodate the settlement and seismic activity

Situated in about 28m of water, the island is built like a large doughnut within which the 98m diameter core has been excavated to about 42m below the sea bed. First, the soft sea bed soils were stabilised by a deep mixing method of grout injection and installation of sand compaction piles. Prefabricated steel jacket structures were then installed to form a combined retaining wall, mole structure and work platform. A 2.8m thick x 119m deep slurry wall was then built between the two jacket structures. Once installed, the inner jacket was removed, the seawater was pumped out, and excavation of the core began. Up to 11 large excavators and an enormous barge-mounted 45m³ clam shell crane were seen working in the doughnut pit during a visit to the island in early 1993. A system of deep well dewatering reduced the uplifting pressure of the saturated soil during excavation of the core.

Unfortunately, as the concrete floor at the bottom of the core was being installed at the end of excavation in November 1993, symptoms of potentially serious water seepage through the foundations were detected.

As a precautionary measure, the core was reflooded to about 20m.

The exact cause of the seepage became the subject of investigations, but the situation was recovered and construction resumed in April 1994.

At the time of writing, ground freezing stabilisation of the surrounding soil was being prepared for launch of the island's four TBMs.

A similar need to reflood the Kirarazu manmade island occurred in July 1993. This island is of a different construction with the transition ramp to the bridge created by material backfill within a strong cellular retaining wall structure and without the need for seabed excavation. The reflooding was required when water seepage was detected at the joints between the steel caisson and steel cell revetments. Two months was lost on the Kirarazu Island programme while it took four months to recover from the Kawasaki incident.

Traffic projections

Segments of the primary lining are erected by automatic robot erectors

The areas adjacent to the landfalls of the bay link are heavy industrial zones. Some 20% of Trans-Tokyo Bay traffic is therefore expected to be heavy goods vehicles. The link is designed for highway traffic speeds of 80km/h maximum on maximum gradient slopes of 4%. The tunnels will have a longitudinal ventilation system with 9 electro precipitators (four or five in each tube) providing air filtration. The tunnels will also be equipped with automatic side-wall cleaning systems which will run continuously. The Kirarazu transition island between the tunnel and east channel-bridge will be furnished with motorway services where drivers can break their journey if necessary.

The capacity of the highway is limited initially by the two, 2-lane tunnels. When traffic flows reach 50,000 vehicles/day or so, a third parallel tunnel will be excavated to meet increasing demand. This third tube will operate a contra-flow to cope with peak period demands. The third tube is not expected to be needed until at least 10 years after the initial highway opens. All provisional work for the third tube, however, is included in the current scope of the islands and tunnel landfall works.

Trans-Tokyo Bay represents the ambition and vision that has come to be expected of the Japanese. When opened it will rank among the world's great civil tunnelling achievements along with the Seikan Tunnel, the Storebaelt link and the Channel Tunnel. Like them, the endeavour is wished every success.