Sara, Emily and Jean Salt; Fred and Clara Potter. Ring any bells with you? Why would they? Except these five souls were taken by the railway in the most appalling circumstances imaginable.

They weren’t on board a train; neither were they trackside. At 05:35 on the morning of Tuesday 28th April 1953, the houses they were sleeping in disappeared into a crater, consumed in an instant by the ground they stood on.

Temple Drive in Swinton, part of Manchester’s homogeneous suburbia, offers little to set it apart from thousands of other streets. You have to look deeper to see what’s different here. Back in 1909, 22 and 24 Temple Drive – a pair of semi-detached houses – were unwittingly erected above a buried shaft, sunk to expedite the driving of Clifton Hall Tunnel 60 feet below them 60 years earlier.

Physical evidence of the tunnel’s temporary construction shafts – eight in number – was lost behind the lining as bricklayers pushed forward with their work. The ground above was reinstated; memories then faded, plans were filed away or lost. To all intents and purposes, the shafts were gone. But the timber frame embedded within the brickwork underneath No.3 shaft – supporting the 200- ton column of wet sand that filled its former void – was decaying. Gradually the load was being transferred onto the lining. When it could no longer withstand this burden, the contents of the shaft burst through into the tunnel, catastrophically undermining the dwellings above.

A police officer living across the road was awoken by a sharp cracking noise. He rushed to the window in time to see the houses folding into the earth; another had its end wall torn off. A violent rush of air propelled debris skywards. All around was devastation. The Salts and the Potters didn’t stand a chance.

Knowledge gap

The tragedy on Temple Drive left a legacy that still commands the attention of our engineers. Its immediate aftermath spawned a flurry of investigation as the railway sought to quantify the extent of its liability nationwide. There was nothing unique about Clifton Hall. Most tunnels of any length were progressed by sinking shafts and, job done, it was cheaper to dispense with them than make them permanent, except where ventilation needs demanded otherwise.

BurdaleShaft [online]

A depression resulting from the collapse of a hidden shaft.

Typical was the now-disused Queensbury Tunnel to the north of Halifax, with its backstory of construction difficulties. As work got underway there, the intention of sinking eight shafts was revisited such that only seven were ever started. Two never made it to tunnel level. The position of the northernmost shaft didn’t change with the replanning and was thus known locally as No.8 shaft throughout its operational life. This apparent discrepancy proved untenable given the events in Swinton. Staff were despatched to walk over the tunnel with a critical eye and an exchange of letters followed. On 3rd June 1953, the District Engineer wrote to the Chief Civil Engineer in York informing him that “the other filled-in shaft, although from local knowledge is known to exist, cannot be located on site, and there is no information on the original plans indicating its position.” He was wrong but didn’t know it.

The same uncertain story repeated itself many times over as records were dusted off and studied to mitigate what the Ministry of Transport inquiry report later described as “the danger which arises when vital knowledge is not readily available.” Its author, Brigadier C A Langley, recommended that “all tunnel records be reviewed and that any special features be brought to the notice of the maintenance and examining staff. It is also desirable that the position of disused shafts should be permanently marked in the tunnels themselves, so that these places can be particularly watched.”

Fast-forward through 60 years of sporadic hidden shaft projects and we discover that Network Rail’s 693 bores host 315 of them. Confirmed, located, evaluated. But it’s indicative of just how tricky it is to unpick history with confidence that there are another 223 suspected hidden shafts waiting to be found, based on the results of desktop studies. This is a mission that could keep consultants busy long into the future. Network Rail is determined it won’t.

Drawing a line

So what is a shaft? What probably comes to mind is a large hole in the roof of the tunnel extending upwards, with a chimney-like structure on the surface – known as a protection wall – topped by a grille to prevent anything substantial falling into it. But there’s more. Network Rail actually records four categories:

    • an open shaft as described above, used for ventilation or aerodynamic purposes
    • a blind shaft, discernible within the tunnel but with one or more caps preventing access
    • a proven hidden shaft, invisible within the tunnel and at ground level but established beyond reasonable doubt through historical records etc and on-site investigations
    • a suspected hidden shaft, indicated by documentary or local sources etc but not proven.

Brigadier Langley’s underlying  intention was for the fourth group to disappear, consumed into the third as enquiries confirmed their existence. Considerable progress has been made over the past 15 years but the new Control Period (CP5) will bring a further acceleration in activity. There is real corporate and regulatory appetite to get the thing finished within the next five years or so.

Needle in a haystack

Network Rail’s asset management standards demand that every one of its tunnels has a Tunnel Management Strategy (TMS), comprising three parts: a desktop study, a risk assessment and an action plan to mitigate any intolerable risks. The first of these elements involves an exhaustive trawl of all known information sources, including those beyond railway ownership like the National Archives and Bodlean Library. This comes together to offer the deepest possible insight into a tunnel’s construction and subsequent history.

HiddenShafts-231 [online]

Engineer Laura Pollard from Donaldson Associates measures the length of cored brickwork before confirming the presence of voids behind the lining.

It can though be hit and miss in terms of shafts so a rule of thumb is applied whereby a hidden shaft is presumed to exist if no evidence of one is found for more than 300 metres, a figure that reflects general experience with spacings elsewhere. There is of course a recognition that some tunnels have more frequent shafts whilst others have none at all, perhaps due to their depth, landowner stipulations or the availability of machinery that negated a need for them.

The hidden shaft programme sits alongside the TMS, initially building on the desktop study. Depending on the confidence of an inferred shaft location, a second phase can involve opening the geophysical toolbox to help narrow down the search. A range of techniques – ground penetrating radar, conductivity, resistivity, thermography, 3D tomographics – have all been tried in the past but the results have been patchy, often providing evidence of anomalies but only rarely pinpointing shafts. Aerial photography has also been used to look for depressions or wet patches.

The tendency now is to go directly to a third stage: low-cost drilling within the tunnel. This has delivered the best returns by far, even when the baseline information has been a little vague. Small holes are cored through the crown at perhaps 1.5m centres, allowing an engineer to see behind the lining with an endoscope. If anything is observed that requires further investigation, larger holes (upwards of 50mm) are opened to gain better visibility as part of a second phase, allowing the use of devices such as telescopic cameras or laser scanners. The results can be surprisingly clear and detailed.

The aim of all this is to obtain two corroborative but independent pieces of information which both point to the same place. This marks the tipping point where a suspected shaft becomes confirmed. With luck, archive drawings and written commentary – even newspaper articles describing the works – can be sufficiently detailed to provide this evidence without heading out to site, but at some point there is usually a need to venture into or over the tunnel.

Which way forward?

Hidden shafts come in many forms, from fully brick-lined – as if permanent – at one end of the scale to a rough hole through competent rock at the other. But by far the majority are in “altered ground”, featuring remnants of their temporary timber lining, voids and collapses of the loose/granular material surrounding the shaft. Many incorporate intermediate-level timbers, beams or brick domes that divide them into chambers; around half are backfilled, in whole or in part, perhaps with trees or assorted debris.

This make-up, along with its stability and the local geology, contributes to the risk assessment from which a shaft’s future management strategy evolves. The process drives an understanding of whether movement of the shaft is likely and what the potential impact might be at ground level and in the tunnel, any remedial action being prioritised on the basis of the route’s linespeed, traffic type and service frequency, as well as land use above.

Where a shaft is located on moorland, a reasonable course of action might simply be to inform the landowner, fence off the area and control the risk through inspection. Not only are most tunnels subjected to an annual detailed exam whereby every part of its structure can be touched, a walk over the top is also undertaken to check for ground movement.

Urbanisation changes the landscape literally and metaphorically. Not without reason, developers seek to extract maximum value from their land assets, sometimes resulting in houses being put up within feet of known shafts. Try following the line of Glenfield’s mile-long tunnel under west Leicester – with its 13 shafts – to see that played out. But what if there was no sign of a shaft when the builders moved in? What if 22 and 24 Temple Drive were not the only dwellings to be plonked unsuspectingly where they really shouldn’t?

Mitigation measures

Only about one in ten hidden shafts require remediation, usually in the form of ground engineering although the precise methodology varies according to circumstance. Each design is bespoke. With potential for the works to impose additional load on the lining, an initial stage can involve the provision of reinforcement, perhaps by improving the brickwork’s stiffness through cross-stitching and grouting, or spraying a concrete arch. An alternative approach has seen grout injected to form a saddle around the extrados of the arch (its upper/outer surface). Uniquely in Drumlanrig Tunnel, gauge constraints drove  the installation of steel frames, suspended at the crown below its hidden shafts and secured back into the rock using high-capacity cable bolts.

HiddenShafts-135 [online]

Work to investigate a hidden shaft in Cowburn Tunnel involved holes being cored at the crown of the arch to see what’s behind the lining.

The shaft void itself can then be treated, resulting in it becoming a stiffer anomaly through a phased introduction of grout or lightweight foam concrete. There have been some extreme cases, notably at Corby where a shaft was dug out completely and a larger one sunk around it with segmental rings, the intention being to replace something that couldn’t be managed with something that could.

Reconstructive work is often required at the surface. When a shallow circular depression appeared in parkland above Blackheath Tunnel, a geotextile mattress – secured with gabion anchors – was installed at a depth of 4 metres to ensure any repetition could only prompt a gentle settling of the ground. Lengthy investigations found no sign of a shaft except water ingress in the tunnel, but logic determined there must have been one at some point.

A situation has not yet arisen where the available engineering options were deemed insufficient to deal with an emerging risk. The coming-together of circumstances that led to the Temple Drive collapse has not been repeated. But it could be, that’s accepted. Householders have been evacuated on one occasion – quite speedily – following ground movement associated with a hidden shaft, resulting in them spending a few days in temporary accommodation. Elsewhere, an almost-integral garage had to be demolished for remedial works, so close was one shaft to the adjacent house. And that’s what really strikes you about this: the pot-luck nature of it. The railway might have sunk these holes and chosen to erase them, but it has had no influence over subsequent development, either close to them or on top: hospitals, roads, schools. That kind of risk is uncomfortable to manage.

Calling time

There is an obvious operational imperative to filling the knowledge gap created by hidden shafts. A substantial budget – into the tens of millions – has been committed to doing so which should establish a considerable ongoing workload for the consultancies and contractors asked to take part.

Sixty years have elapsed since, out of nowhere, events in Swinton forced the railway’s civil engineers to confront this liability, and yet we’re still talking about it. But not for much longer perhaps. “The urgency for wanting to close this out is accelerating”, insists Colin Sims, Network Rail’s Principal Engineer, Asset Management Technical Services. “Not because of any incidents; we are just of the opinion that it’s been hanging around for too long. We have an intention to find them all and we won’t close the file until we do.”

Photos courtesy of Four by Three