The Hastings Bathing Water Improvement Scheme tunnel resulted from the Urban Wastewater Treatment Directive which requires authorities to collect storm water and sewage flows and prevent pollution going into the seas. Overflows which occurred at every storm will now go via Coombes Pumping Station to a new wastewater treatment works; spill rate is reduced to one year in every five.

Southern Water had considered various options, involving large tunnels, individual storage tanks, and combined storm overflows, connected by transfer mains. Capital value assessed against operational costs led to a large tunnel choice. Feasibility studies over four years, costing 10% of capital value, had examined tunnel integrity, the effect of sewage and sewer gas on concrete; ventilation and odour control; flow data; and solids retention inside the tunnel. David Butler at Imperial College advised on this.

The client’s original contract strategy was to control the risk, and the original proposals went out under the IChemE Green Book as reimbursable contract with target cost. A novel approach to the procurement was taken to ensure that the external costs the contractors were going to absorb on the scheme were minimised. The control of risk was very important for both parties, and they clearly wanted to understand what risks they were undertaking. Southern Water, in a highly regulated business, wanted surety of costs, and a saving of £1M or £2M was important.

Work began in 1995 with contract award in July 1997 as a £42.8M design-and-build with contractor Miller taking on key risks for ground conditions, settlement and artificial hard obstructions.

They felt sufficiently confident in methods and understanding of the ground to offer a “Red Book”, all-risk, lump-sum, fixed price. It was a success: despite great difficulty the main contract was less than 10% over budget and completed just two months late on a two-year project. All the statutory obligations laid down by the Environment Agency were met.

Two contractors from four were pre-qualified on the basis of innovation and experience in large diameter tunnelling and these were awarded £150,000 to develop the initial design from the feasibility study. Southern Water was not expert in tunnelling and consultant Sir Charles Haswell & Partners shadowed the project all the way through.

An issue was planning approval. It was decided to work up a scheme in full and complete all the site investigations before looking at potential worksites.

The only available sites were the greenfield or park areas to be reinstated after use which was a challenge; It was not popular. Owners were approached with outline proposals to persuade them to accept the use of their land. At Coombes Pumping Station there were about 400 properties very close to the site and environmental health factors were also important to make sure that we had adequate odour control.

Impact was minimised by tight controls, such as a requirement that muck went by rail not road. Southern Water let Miller get on, by using its experience to deal with the local council and other planning authorities, the Environment Agency, environmental health and the DETR. Southern Water also worked with Railtrack for putting structures over the railways. A local liaison committee was formed with Miller Civil Engineering and talks were held directly with the local people to explain what was happening so that they were not unnecessarily concerned about it.

Tunnel ground conditions were extremely variable: predominantly weak sedimentary rock of mudstones with a clay fraction up to 50%, siltstones, and limestones, with bands in excess of 200Mpa as well as weakly cemented, or uncemented sandstone and sand. Water pressure was up to three bar and there were a number of known faults. Settlement control was paramount and a significant factor in the choice of machine, a Mixshield.

Other options were examined, but an open-faced shield with compressed air did not sufficiently minimise exposure to compressed air and had potential for excessive face loss without treatment of faulted areas Settlement control where air pressure was lost could be poor. A bulkhead shield minimised compressed air exposure but meant a small volume of compressed air in the bulkhead could less easily deal with rapid changes in ground. Concerns were also held concerning the material lock, or rotary feeder mechanism to transfer the material from full to atmospheric pressure.

An earth pressure balance machine was considered, but extremely variable ground with possibly 200MPa rock overlying uncemented sand was a worry and foam and polymer technology available at that time was considered embryonic. There was also insufficient confidence in achieving adequate face control.

A slurry system would work even with a high-clay content rock because material would generally break down; difficulties with sticking in the head, and transportation, would not be significant. But traditional slurry shields cannot cope with rapidly changing face pressure because they regulate pressure only by the speed of the slurry pumps. Loss of support is possible. However, the Mixshield uses an air bubble above the slurry to hold a constant face pressure.

A second-hand machine was immediately available, giving a cost saving and manufacturer Herrenknecht was willing to share risk with Miller. The machine was modified to deal with the 200MPa rock, clay, and sand with 3bar water pressure in it. The head, designed by Herrenknecht with input by Miller, was a compromise, using both rollers and picks for soft ground and extremely hard rock conditions.

The partnership with Herrenknecht had two important features. Firstly, key Herrenknecht fitting and electrical operatives were integrated into the team. Secondly, although Miller were responsible for driving the machine, a pain/gain target was devised with Herrenknecht that enabled them to demonstrate their confidence in their machine.

Detailed face pressure calculations were not considered valid or useful in the conditions and the philosophy was to maintain a modest excess pressure, in the event only 2-3m above hydrostatic.

Spoil processing, particularly with a high clay fraction, was extremely important and from the beginning a full-time specialist spoil processing engineer was introduced into the tender team.

The main concern was to separate slurry bentonite from the high ground clay fraction.

Simply tankering away the combined slurry was completely unacceptable environmentally and commercially.

With up to 100 t/hr of fine material it was necessary to maintain density, grading and viscosity of the bentonite in order to ensure that the machine’s face support system operated correctly.

There were three contenders to create a flexible system, all starting with primary screening, and then hydrocyclone treatment Tertiary treatment for the fine fraction, which could reach 100% of material, was crucial, with all of it passing the effective cut-point of half a millimetre at the screens.

Options were centrifuges, plate presses and belt presses, extensively trialled before work started with material from the cores. belt pressing trials were at Allied Colloids in the UK, and plate and frame trials with MS in France.

But although performance of plate-and-frame and belt presses was within 1%, the former had limited capacity. Site restriction and cost made it impossible to introduce sufficient plate-and-frame presses.

The tunnel route was chosen to mitigate settlement under often the poor-condition properties in Hastings. But Southern Water and Haswell decided to take a direct line between the Alexander Park drive shaft and the Warrior Square shaft, typically under three- or four-storey buildings, well over a hundred years old, many of them with no foundations.

A structural survey of 700 properties was undertaken, as well as a condition survey of 1,600 properties to assess possible damage reimbursement. Some one hundred properties had pre-tunnelling works carried out at a cost of about £75,000 as part of a risk-sharing strategy under which Southern Water prepared any property with pre-existing faults, to allow the TBM to pass with some confidence. Contractor Miller took on the risk of the damages which were either inevitable or which were caused by out-of-specification settlement.

On past performance of similar machines, a face loss percentage of about 0.75% was expected from the Mixshield. Anticipated damage was in the region of minor to negligible. However nine houses at Braybrook Terrace needed a significant amount of work. After discussing the poor state of repair of these houses, already showing signs of subsidence, it was decided with Miller and Charles Haswell, to demolish them prior to the TBM launch.

The tunnel lining was configured as a tapered trapezoidal lining with a self-cleansing profile. It was designed as a soft-ground ring to cater for full overburden which, given the soft rock conditions, was considered conservative. Only one type of ring was designed rather than try to differentiate between the low cover sections, or the different ground conditions along the total drive length of the 1.6 km main tunnel. The ring incorporated discrete male-female shear pads between the rings to force good build. Grouting was all done via the tail skin and the ring had no grout hole. If secondary grouting was required, this was carried out by drilling through the back of the lifting point.

The ring incorporated an EPDM gasket. Southern Water had decided that the storage tank required a special profile for self-cleansing purposes, together with a low-flow channel and a walkway. taking this profile it was decided to produce a non-circular ring. With this shape the benching was significantly reduced, which saved considerable time on the programme. The ordinary plates are 300mm thick and the Ôlamb chopsÕ, as they were affectionately known, were about 450mm with a trapezoidal relief in the invert for the drainage channel.

Shaft Construction

Six shafts were all caisson-jacked down because of concern that some of the sandstones would behave like running sand. Although there was the potential for a lot of water it was considered that all of the shafts could be jacked down as dry caissons with pressure relief wells. A new shaft lining developed together with Charcon had a twin tie-bar system to give a uniform loading and stability on the EPDM gasket, an integral bentonite hole, and a shear pad from ring to ring, again to force good build. The ring also had a scarf joint and could be underpinned should the caisson get stuck. Construction went very well in the soft ground and in mudstones where sinking rates achieved 4-5m a week. But in limestone rates were down to about 1-1.5m a week.

Tunnel Construction

The first challenge was to get the 7.5m diameter machine to site. It could enter a port in London and then transfer down the A21, a single carriage road; this was refused by the Department of Transport, worried about complaints from complete road closures. A second option was that it could go to the port of Newhaven, along the coast from Hastings. But this required going across country and cutting down many lighting columns, traffic lights, and, more worryingly, people’s front hedges and trees. A third option was selected of bringing it on to the beach at Hastings which went remarkably well and caused a lot of public interest as the machine moved up the town.

The machine was lowered with an 800 t crane.

As part of the planning agreement, working hours at Alexandra Park, where the main drive started, were from 7am in the morning to 11pm. From 11pm to 7am was used for maintenance in the cutterhead and any other general maintenance, allowing a daily 16-hour driving, and eight-hour maintenance cycle.

The Mixshield operated with a compressed air bubble system, which gave an early indication if pressures were wrong, and if pressure in the head was being lost. The system required the material to be kept in suspension as it was pumped to the surface separation plant. It went then by conveyor belt over the adjacent road and railway and transferred into a long thin muck bin. From there it was transported approximately 50 miles (80km) away to a tip near Maidstone.

This arrangement was not ideal, but Hastings Council would not allow tipper wagons to run in and out every ten minutes. Instead the railway was used with one train taking 1000t, the spoil from five rings. Two trains ran a day when good progress was being achieved.

For grouting the annulus behind the ring a sanded grout mix was pumped in through the tail-skin. The sand was needed to give it a bit more body to the grout and to give early strength for the gantries behind the TBM.

Early hold-up

Soon after starting we found we were not getting a seal at the face. In addition massive settlement occurred, totally above all predictions, of the order of 100-120mm. Two holes also appeared on the surface. After studying the settlement figures it was obvious that a very narrow trough was forming, a chimney effect, and it was decided to go into the head to see what was happening. If the pressure was slightly high or low the face made or lost water, which was an indication of porous ground.

It was found the mudstone was more mud than stone and was settling out in the excavation chamber, clogging up the wheels and the cutterhead. It was impossible to keep the slurry in suspension. Progress rate had been down to between 2-5mm a minute, three or four hours cutting for a ring.

Modifications were discussed at length with Herrenkencht. Paddles were fitted on the back of the cutting wheel, so as it swept round the excavation chamber it stirred the slurry and slurry flow capacity was increased from 1000m3/h to 1400m3/h by a second pump which simply circulated slurry around the cutterhead.

Centre rollers in the cutting-wheel were replaced. and the pick lengths on the cutter-wheel were altered to give a larger chip size rather than turning the mudstone into a fine pulp.

This improved matters but progress was still only steady in the mudstones and clays. In granular materials, Tilgate stones and sandstones, the machine performed better because granular material stayed in suspension better. Progress improved to 20-30mm/min, but there was still a problem getting into the head to examine what was happening.

The air was turned off temporarily to access the face to have a look after carefully selecting a location; there was a concern that if the ground did come in it would cause high settlement. When access was again made it was discovered that the narrow band of blue Tilgate Stone was highly fissured rock and all the air was passing straight through it.

A super-hydrophilic polymer was used to seal the rock and compressors were increased from three to six giving over 4000m3/min of air to stabilise the face. This was successful.

Once the face could be viewed properly the high wear that had occurred was visible. All the cutters, and virtually all the picks has to be changed and a number of the pick boxes that the picks are mounted in were also damaged beyond repair. Rollers and picks could be changed from behind the cutting wheel. But to change the pick boxes it was necessary to go in the front of the machine which meant tunneling in front, doubly difficult because only allow three hours work could be allowed at pressure.

After these launch problems, it was felt necessary to re-examine the risks in Braybrook Valley to try and pinpoint a fault that ran through the valley. Further site investigation was carried out which also gave more information on the rock-soil interface and on hydrogeology. A more detailed risk assessment also looked at the hazards that could be discerned from the launch and at their effect across the valley. As a result the capacity to use 24 hour, 7-day working, should the conditions require it was decided on. Monitoring rate was increased as the more difficult ground was approached and provision was made for quick assembly of an expert review panel if settlements reached certain trigger levels.

In addition emergency preparation was instigated with a full briefing of the work-force and, in particular, the establishment of good communications between the TBM drive and the separation plant.

The Braybrook Valley saw the best two weeks progress on the whole contract. Settlement was less than expected and the best week was achieved, 60 rings or 90 m in five 16-hour shifts.

On the second half of the contract, progress was significantly improved. A schedule was devised to measure the roller and cutter wear and determine changes. The gauge and profile cutters were changed a lot earlier with only 5mm and 10mm wear whilst the ones on the face were changed at 15, 20 and 25mm wear.

But restrictions on muck-away were a problem. Rail respects good environmental principles sounded good at planning but in practice was not practical solution. Only one tip in the region had a rail-head and spoil had to go from Hastings by train, almost into London, and all the way out to Maidstone. A local tip for trucks was 3 miles (4.8 km) away.

Two trains a day had to fit in with commuter traffic which was very inflexible, especially as they were too few to be of interest to a rail operator. Trains were delayed or shunted into sidings, not keeping up with production and causing problems on site.

Environmental legislation caused other problems. For example work stopped for five weeks to make a minor modification to the tip licence allowing spoil to be spread out to dry in the sunshine. This seemed unnecessary as the spoil was fresh and mixed only with tap water.

Interpretation of legislation was extremely narrow and only limited discretion was allowed. An apparently impractical and intransigent stance was adopted by the Environment Agency. Convoluted statutory protocols were involved in what seemed to be minor changes, and in this respect the Environment Agency’s hands were tied by the existing legislation, or at least the UK interpretation of them. There did not seem to be an integrated overall strategy, or even a co-ordinated approach among regional Environment Agency officers to deal with the question of spoil disposal on a major project which was in itself a result of the Environment Agency need to respond to a European Directive.