Contesting Brunelís unwelcome bequest
December Issue 2009
As we all know, Victorian engineers created a railway infrastructure that has far surpassed its original design life and for that we are eternally grateful. However we occasionally come up against engineering challenges left by our forefathers, possibly designed to keep us on our toes! They certainly remind us that we must never take anything for granted. Recent work carried out by Network Rail in Sapperton Long Tunnel is a good example of this. Moreover, it acts as a timely reminder as to why we need the CDM regulations today.
Sapperton Tunnel is situated under the Cotswold Edge between Kemble and Stroud, on the Golden Valley Line that links Swindon and Gloucester. The Cheltenham and Great Western Union Railway started work on it in 1839 but, four years later, it was taken over by the Great Western. There are three tunnels in the vicinity - a 3490m canal tunnel and two accommodating the railway. The longer of these is 1705m long and separated from the shorter 323m bore by a deep cutting.
Below the invert
Work was started on the Long Tunnel in 1841. Mark Brunel arranged for ten shafts to be excavated - four of 3m diameter and six smaller ones. These were designed to allow engineers access into the hillside, providing a better understanding of the geology and underground conditions as well as enabling tunnelling to begin. Whilst there are many shafts around the country that are used today for ventilation and access, it is unusual to find shafts that descend below the tunnel’s invert. That is the situation at Sapperton.
Brunel had already determined the gradient of the railway at this point and the depth at which the bore should be constructed. However, the shafts started to expose an unstable layer of Fullers Earth at this level, running the length of the proposed tunnel. As a consequence, Brunel decided to construct it at a higher level where ground conditions were more stable. Although this meant that the track gradient was steeper than planned, the anticipated length of the tunnel was shortened and costs were therefore reduced. Government approval was given and work started in 1847.
This was a perfectly reasonable decision to make at the time and everyone was happy. It has though left future generations with an interesting legacy - ten shafts extending up to 6m below the completed tunnel’s invert. They had been capped off with timber, covered with ballast and two tracks installed. It was not known whether the shafts themselves had been filled in.
In 1950, the driver of a train travelling through the tunnel noticed a void under one of the tracks. Fortunately, there wasn’t an accident but our Victorian inheritance was now becoming apparent. Something had to be done, and quickly.
The solution adopted at the time was to expose all ten shafts below track level and install concrete beams reinforced with bullhead rail and shear links. It was not known whether backfill had been included in the process. Fortunately, more beams were cast than were actually required and one was left in a nearby yard.
Tested to destruction
In 2001, Network Rail carried out further work, designed by Tony Gee and Partners, on the six smaller shafts but left the four larger ones alone. Subsequently, a decision was made to carry out a destructive test on the redundant beam to establish whether it was strong enough to withstand modern RA8-level loading required for the route. This determined that the concrete in the beam had deteriorated, concluding that the maximum loading for the beams was equivalent to RA5. The four shafts - numbered 4, 6, 8 and 10 - needed to be stabilised without too much delay. Shafts 6 and 8 were considered to be the least stable because of surrounding formation conditions - they straddled either side of the tunnel’s centre.
A seven-day possession was negotiated with First Great Western to coincide with the recent autumn half term. Under the overall control of Dan Tipper, Network Rail’s Project Manager, a scheme was developed by Carillion, the principal contractor under the Western Framework Contract. Carillion appointed Tony Gee and Partners for the design work whilst Network Rail assigned Deborah Elliott as the Scheme Project Manager. Adrian Martin took on the role of Site Manager.
Two sites established
More than 100 people were involved on site, working alternate 12-hour shifts on the two sites created within the tunnel - one for shafts 4 and 6, accessed from the Hailey Farm end, and the second for shafts 8 and 10 which was reached via the Frampton Mansell portal. The project was planned so that all plant and materials for each site had to come in from the designated end. As Deborah Elliott explained, this approach removed any need for plant or people to cross over from one site to the other, eliminating a considerable degree of potential risk to the work.
Many of the workers were not able to work 12 hours and then travel home to rest so mobile sleeping accommodation was provided on site. Bunkabin supplied 12 units using purpose-built modules which allow two people to sleep in considerable comfort. This proved to be very successful and is becoming quite a common feature on Network Rail sites.
The budget - around £2 million - allowed for all four shafts to be stabilised but, until work started, no-one could be sure how much time would be needed to complete many aspects of the work. Two road-rail machines on each site, provided by Ready Power, showed their true versatility. Working in tandem in rail mode, they removed the track from one road; the ballast was excavated and then the concrete beams supporting the other track were lifted out, having first been cut into three. At this stage, they could see that the shafts had been filled with compacted rubble; prior to this, all workers in the vicinity had to wear a harness tied back to the tunnel wall in case there was a 6m drop.
With the old beams removed, two new concrete beams - delivered to site in three longitudinal sections - were used to span the shaft. These were designed by Tony Gee and Partners to a strict weight limit, enabling the use of road-railers performing tandem lifts. Each beam was designed to support one rail so additional spacer beams, designed to allow for construction loading, were placed to fill the gap between them. Once this was completed, the road-rail machines moved onto the new beams and, working in road mode, repeated the process for the other line. We sometimes criticise road-rail vehicles but what other form of plant could do this in such a confined space?
Cutting concrete beams
Cutting each beam into three sections before removal was not as straightforward as it might sound. This part of the process caused the team the greatest concern and this proved to be justified. Various, one-metre diameter diamond-tipped disc saws were tried but they all struggled to cope - when the saw cut through the shear links within the beam, the heat generated tended to weld the disc to the reinforcement. At one stage, it was taking almost six times longer than anticipated. However, they persevered and eventually the process was refined and a Hilti diamond disc was selected as the most suitable cutting machine.
Once the two critical shafts 6 and 8 had been stabilised, a decision had to be made about shafts 4 and 10. Adrian Martin explained that he gathered all the key people together - including the plant controllers and operators - who he said were doing a superb job. After a review of all the options, with valuable input from many of the site workers and designers, they decided that they would attempt to stabilise the Up line of each shaft, leaving the Down line for another day. This they did and the work was completed without incident - the possession being handed back four hours before the deadline. Clearly this was a good decision, highlighting the value of contingency planning and effective dynamic risk assessment on site involving all the key people.
Effective p-way support
Within the possession, the team was able to stabilise a troublesome embankment close to Kemble Station and the local Gloucester track maintenance team, led by its Section Manager Paul Graham, managed to carry out a considerable amount of preparatory work, easing clips in the tunnel in preparation for rerailing next summer. Their flexibility was much appreciated by Adrian Martin and the project team. As is always the case in such circumstances, you want the track maintenance team to be there just when you need them and for them to disappear when you don’t. The Gloucester team did just that. Many a project manager will look on with envy!
Over 12,000 hours were worked throughout the possession. No accidents or incidents were reported. All the essential work was completed within budget. The lines were opened on time at the planned speed of 30/50 mph. Further tamping enabled linespeed of 90mph to be introduced one week later, also as planned. Initial plans are now being developed to complete the Down line stability work to shafts 4 and 10 next Easter.
It’s clearly a job well done by all concerned and, coupled with this, they were able to successfully trial a new form of high visibility clothing involving a fluorescent luminous gel called FHOSS Technology.
A can-do attitude
A group of soldiers from The Royal Engineers, who recently returned from Iraq, visited the site as part of a shared learning and best practice initiative. They were impressed with the design life of the new concrete beams which is 120 years. Also, these beams have been designed so that Network Rail’s successor can remove them with considerable ease - CDM at work! The Royal Engineers were very impressed with this approach to longevity given that, in their environment, engineering solutions are much more immediate and usually required for a very limited time. They also pondered how they would get their Chinook into the tunnel if they were to do the work!
The most significant comment they made was to say that they thought that the military had a monopoly on “professionalism and a can-do attitude”; that is until they visited this site. Now that is a real compliment.