After the summer break, on 18 September 2025 a large audience returned to the Telford Theatre at the ICE, in London, for an insight into the recently completed Exploratory Works for the Coire Glas hydropower project under development in Scotland. The night was also notable as the first time, of hopefully many, when a joint lecture has been held with (mostly) BTSYM speakers to the full membership.
For a relatively small project, it would have been a wonder to some why four speakers were necessary, but as it would become clear, Coire Glas was a model of collaboration, and all the key players in the venture were represented.
The night was an honest retelling of the full story of a project, from inception to completion, although in this case it’s only the first chapter in what will be the biggest scope in underground works of its kind in the UK in a generation.
The presentation kicked off with an overview of the full Coire Glas scheme, an ambitious pumped storage hydroelectric power station, capable of up to 1,300 MW generation, and 30 GWh’s long duration storage. Sean Murchie, Engineering Geologist for the project developer, SSE Renewables, started off by taking us through the landscape of this part of the Great Glen, with the plant’s lower works on the banks of Loch Lochy, and the upper reservoir located in a convenient horseshoe of the Munro ‘Sron na Coire Garbh’. In between the two, sits Sean Mheall, the Southern ridge of the Munro, and it is inside of this mountain that the workings of Coire Glas will be situated: turbines; pumps; watercourses; surge shafts; and, kilometres more of access tunnels.
The concept of the Exploratory Works came from SSE’s commitment to obtain detailed Ground Investigation (GI), to de-risk their future works, especially given the size and scale of the planned excavations. It was stressed that although the broader Great Glen has been studied extensively, relatively little site specific information was available and this was identified as a project risk early on.
Those that are aware of the geological history of the Great Glen, the longest fault line in the UK, will understand that it was likely to be an interesting study. The published geology for the site included two main domains: the Great Glen Fault Zone itself, including heavily sheared and brecciated fault rocks, these cataclastic rocks having a predominantly quartzite or psammite protolith, with subsidiary amounts of semipelite, amphibolite and metacarbonate; then, to the North, the Tarvie Psammite Formation, consisting of more competent meta-sediments – predominantly psammite (metamorphosed sandstone) with subordinate bands of semipelite and/or pelite.
Sean continued with a rundown of SSE’s strategy for the Exploratory Works, involving a two pronged approach: a campaign by Fugro of conventional surface GI (in some unconventional locations, such as a borehole from the top of the mountain and others from the floor of the coire (glacial hollow); and, the underground works by Strabag, accessing areas that surface drilling simply couldn’t reach. This would be made possible with the construction of the Exploratory Adit – a relatively small heading, originally planned at 900m in length, and then to be followed by a campaign of underground GI, to target the locations of key structures even deeper underground.
The design for the Exploratory Adit had to be undertaken with only the geological information already available, supplemented by some limited surface outcrops, and a healthy amount of engineering judgement. Tom Taplin, Engineering Geologist in the design joint venture of Stantec-Cowi, the client’s Designer, explained how the factors at play were run through the Q-system, taking into account rock quality, jointing, and stress influences, for a range of anticipated conditions. This analysis resulted in a suite of conservative designs, or support classes, from A to D, that could be installed depending on the actual conditions encountered underground.
For areas of better ground, like Class-A/B, ground support consisted of a thin sprayed concrete lining (SCL), with either spot bolting, or systematic bolting respectively. If conditions deteriorated, Class-C support included lattice girders and overhead spiling. Class-D was for the worst of conditions, anticipated within the Great Glen Fault Zone, and would employ H-Beams, spiles, face bolts, and a closed invert. The designs also included the construction of wider passing bays and turning niches, to facilitate construction, and the drilling gallery at the Adit’s inbye end.
Strabag got access to the site in December 2022, not exactly earth works season in the Highlands of Scotland, noted Joe Coxson, Tunnel Agent with Strabag. There were other complexities to the mobilisation, owing to the site’s Loch-side plot and the single track lane running to it. These problems were overcome with an initial delivery of plant using a barge via the Caledonian Canal, and then some targeted improvements and strengthening of bridges along the Kilfinnan Road.
The works on site over the first winter involved felling of the commercial woodland for the required compounds, then preparations of all the support functions required for tunnelling; offices, welfare, batching plant, water treatment plant, explosives magazine and workshops. The work compounds themselves were completely off-grid, with no access to power, water, waste disposal, or any telecoms connection – all had to be set up from scratch to supply the tunnelling, and deal with any discharge.
Another early consideration was the tunnel spoil, which wouldn’t be leaving the site, given it wasn’t viable to transport the quantities involved along the Kilfinnan Road – not to mention this material is likely to be incorporated into future works, so it isn’t really waste at all. Instead, the most suitable area was selected to hold the material for the duration of the works, and beyond, with a design for this Spoil Storage Area undertaken by Arup.
Work on the Adit itself started in March 2023, with the first blast of what would eventually be slightly more than 500 advances on the project. A pre-split successfully created a clean rock face for the portal, with soil nails supporting the slope above, then rock dowels and netting below. Credit is due to RJ McLeod and Albion Drilling, whose local workforce were indispensable during this stage of the mobilisation.
The tunnelling crew’s first activity was creating a false portal at the tunnel eye, consisting of a lattice girder and a sprayed concrete arch. This allowed a full test of the tunnelling systems before excavation commenced, and valuable familiarisation with the equipment itself. The traditional blessing of our Santa Barbara was held in mid-May, thanks to Father Danny from the local parish, and tunnelling got underway.
Joe continued with an overview of the Adit’s long section, with an initially steep decline of 7%, then a shallower 4% after the first hundred metres. Passing bays were planned roughly every 250m, each with a turning niche. The majority of the Adit was on a slight right-hand curve, towards the planned underground powerhouse location.
The full drill and blast cycle was explained, with an overview of the equipment used, and engineering tools at the team’s disposal, such as Epiroc’s Underground Manager software, for creating and implementing drill plans, and Amberg Tunnel software for checking profiles and recording as-builts.
An interesting feature of the Exploratory Works was the effort devoted to face mapping. Two engineering geologists had a full hour to map, if required (which would have enraged most over-eager Pit Bosses) as, for the support class system to work, it was vital that each advance was properly captured. The agreed upon Q-value determined by the geologists dictated the ground support to be installed. The geologists were also empowered to instruct any ‘toolbox measures’ – pre-authorised in a Required Excavation Support Sheet (RESS) – if they deemed any features required additional support.
The Works were procured on NEC3 Option B contract, priced with a Bill of Quantities, enabling scaling to match the installation of each support class.
Also explained was a rundown of the blasting methodology in use, of interest to those who’ve not yet tried their hand at hard rock tunnelling. Given the tunnel’s length, it was more economical to use cartridge emulsion explosives, which were initiated with a conventional non-electric system.
A standard 3m tunnel advance had 66no. charged holes, and 4no. void holes around a ‘burn cut’, using roughly 180kg of product. The powder factor here was on the higher side, but given the Shaeff tunnel excavator in use had an apron and conveyor system, good fragmentation was key. Detonators ranged in delays from 0ms to 7000ms, dictated by the blast pattern in use, and all were initiated simultaneously using bunch connectors.
More interesting techniques were also required for some of the different tunnel profiles. A wedge cut was successfully trialled for enlarging an area for a passing bay, with two wider perpendicular advances, followed by a section of slashing. Smooth blasting was also required for some of the in-situ stress testing later in the works – not easy in the highly deformed geology. The larger profiles of the drilling gallery required 140no. charged holes, and upwards of 300kg of product per blast.
Tom T took over here to give an overview of the importance of not just the face mapping, but all of the geological deliverables that were an output of the Adit’s construction. Given the scope of the Exploratory Works, he pointed out, the data collected along the way was as important as the construction itself.
Along with the mapping records, completed by both the Strabag and the Stantec-Cowi geologists, there was also a detailed LiDAR and photogrammetry record of each advance, which provides an unaltered record of the exposed ground, regardless of the support later installed. Forward probing was maintained ahead of the tunnel face, to give early warning of any weak zones, or more importantly, water ingress. A pre-excavation grouting rig was on hand in case of any severe water strikes, but thankfully never used.
The Epiroc jumbo in use also provided ‘measure while drilling’ (MWD) data for every probe hole, blast hole and rock bolt drilled, which when collated, formed a very visual representation of the geological conditions. Tunnelling continued steadily through to November 2023, clocking to reach beyond Ch 700m, almost into the final stretch to the drilling gallery. It was at this point that a feature was encountered consisting of very poor, highly deformed rocks, with ground conditions that fell outwith the existing support classes.
By the next advance this feature was present in almost half the face, and was evidently much poorer ground than accounted for at this depth. In the hope this was a finite feature, an additional short advance was undertaken, but the poor rock continued up to the crown yet again. Additional support was instructed in this area, and tunnelling brought to a safe stop. Sean explained how SSE tackled the issue on all fronts, and was keen to highlight this period being noteworthy for the collaboration between all of the teams on site.
Given tunnelling was on hold, waiting for detailed GI before starting the design wasn’t an option. From the information available and numerical modelling, Stantec-Cowi proceeded with conservative lining design for the location, the assumptions for which could be verified later on. This design, which became known as Class-SQ, consisted of a fully circular H-beam, excavated as a staggered top heading and bench, with further temporary support from face bolts and overhead spiles installed in every advance.
This design was turned around within weeks of intercepting the fault, and allowed for the H-beams and additional support materials to be ordered before the Christmas shutdown.
Initially, probing was the only investigation method available, which successfully located the fault plane dipping down above the tunnel crown, but was unable to estimate the farthest extent ahead of the face. Tom Robinson, Engineering and Technical Manager with Strabag, detailed how more detailed information was obtained. Strabag mobilised one of sister company Zublin’s drilling crews (who were conveniently already lined up for the later GI works) to do an exploratory borehole at the tunnel face.
This provided a full long section of the fault zone that could be logged and its characteristics understood, and also samples that could be sent for lab testing ahead of construction. Amberg also assisted at this stage, utilising their Tunnel Seismic Prediction surveying, its first use in the UK, to generate a 3D interpretation of the fault zone.
Despite the safe stop, Joe assured us that Strabag’s tunnelling team were not inactive at this time, and preparations were required underground before works could restart. A section of tunnel behind the fault zone was enlarged for a passing bay and turning niche – needed for more working room. Additional support was installed at the tunnel face, consisting of reinforced ‘Norwegian Ribs’ and additional bolting. Tunnelling workshops were held regularly during this period, allowing for optimisations in the construction of the Class-SQ lining to be discussed, and roadblocks overcome. The Adit’s alignment was shifted to allow for the initial excavation to be protected by horizontal spiling, preferable to drilling them inclined each advance. An alternative excavator was brought in, better suited for the new excavation profile, and injection hoses were proposed instead of steel packing behind the H-beams. Everyone’s objective was to get through the fault zone as quickly, and safely, as possible.
Tunnelling restarted in early March 2024, the team utilising a road header and soft ground support more common in London than in the Highlands. As conditions improved, blasting resumed, and it became suitable to transition to a larger horseshoe top heading and a double length invert excavation. In all, 28m of Class-SQ was installed, before transitioning back into a horseshoe shaped profile and use of regular support. Despite the reduced advance length and time consuming support installation, the team had managed to average more than 1m/day progress through the fault zone; commendable under the circumstances.
On most projects, it would be difficult to be positive about a three month stoppage due to poor ground conditions, but Sean reiterated that one of the key objectives of the Exploratory Adit was to characterise the ground conditions. Therefore, having a better understanding of the location and nature of this fault zone prior to Main Works construction will benefit the project in the long run. In fact, SSE were so happy with the data being obtained from the Exploratory Works that, in May, they took the decision to extend the Adit by 300m, and instructed an additional two drilling galleries. The data from the Adit was being reviewed in real-time to verify assumptions against the Reference Design for the Main Works, and the extension provided an opportunity to obtain quality tunnelling GI all the way to the depth of the proposed turbine caverns.
There was also a cost-time benefit with the upcoming underground GI works. An extra couple of months of tunnelling actually reduced the number of long exploratory holes, and opening up more work areas allowed for simultaneous drilling. It also provided a much better spread of locations for in-situ stress testing, and closer to the caverns themselves.
While the extension to the Exploratory Adit came as somewhat of a surprise to the team, it would be just another challenge to overcome, and quickly. Of the 176m of tunnel remaining before the first drilling gallery, 153m of that was constructed in that April alone, the best month on record. Building on from the team dynamic of the fault zone, everyone had their responsibilities, and had to understand each other’s constraints. A new alignment was required, which Stantec-Cowi had to thread through the area of the future main work structures, to avoid later clashes. Strabag had to secure labour, support materials, explosives, and sufficient plant and equipment to facilitate the extra tunnelling. SSE had to coordinate the effort, ensuring they were getting value for money from the extra work.
The design ended up including seven different profiles, up to 8m x 6m, and included a much tighter radius bend to site the last gallery in the optimum position. No matter, just three months later, on 1st August 2024, advance 502 was blasted, the last on the Exploratory Works.
Tom T took this opportunity to run through a summary of all the tunnelling GI that was obtained from the Adit’s construction. The vast majority of the tunnel, almost three quarters, was supported as Class-B (or a variation thereof) with a Q-value of 1.0 to 3.0. Water inflow included strikes up to 30 l/min, but thankfully all but half a dozen were finite, and petered out within an hour.
The stitched together LiDAR scans provided an excellent fly-through of the entire tunnel, in its unsupported state, and can be used for analysis of joint orientations, along with assisting the construction team with space proofing. All the MWD data has also been collated, which shows the relative hardness of the ground throughout the adit.
It was Tom R’s responsibility to organise and oversee the second phase of the works, the Underground GI. This looked very different from its original scope after the extension, but at least the constraints of the drilling equipment had been properly understood while the extension design was evolving. Strabag has the benefit of Zublin, also operating in the UK, and their first rig commenced drilling within weeks of the tunnel’s completion. The scope now included almost 2km of wireline and conventional exploratory core drilling, along with a host of in-hole geophysical testing, all to target the planned location of key underground structures.
Strabag’s geologists continued to be busy underground, as all the core samples were logged and photographed to exacting standards, before being transported to surface, to minimise disturbance. A large contingent of subcontractors assisted with the specialist geophysics. There are too many to mention all but Robertson Geo deserve special recognition for their collaboration with Zublin to undertake tests in non-conventional shallow and even uphill inclinations. Plate jack tests were also undertaken, by Sol-Experts, in specifically constructed niches in Gallery-B and Gallery-C. These tests involved putting huge forces into the tunnel walls, through a large hydraulic ram, and measuring displacement within the rock mass.
With the last of a suite of overcoring tests completed in early December, in barely four months one of the most comprehensive underground GI campaigns of its kind in recent years was concluded. That said, it was another couple of months before the real deliverable, a 5,500 page Factual Report, was completed, a collation of every relevant record of the GI works, which should keep Stantec-Cowi busy until the Main Works commence.
Joe took over again for a recap of the project in numbers, before drawing attention to some of the project’s achievements. Tackling what was initially one of the biggest hurdles, a UK ‘hard rock skills gap’, ended up being one of the main successes, with experienced drillers, newly qualified shotfirers, and engineering geologists upskilled for face mapping and core logging, all now available for future works.
The project’s reporting provides a powerful insight into the progress data, to inform on works planning and mitigate the main sources of delays. The rate of tunnelling increased steadily over the course of the project, another healthy sign for the upcoming works.
Another success was the relationship with the local Scottish Fire and Rescue Service (SFRS), whose commitment to support the project was invaluable.
Regular contact, including desktop exercises, underground inductions, and evacuation drills, ensured the SFRS knew what works were happening on site, and what resources were available in the event of an incident.
There were also significant technical innovations, not least of which was the use of tunnelling for GI itself.
The British Geological Survey (BGS) has been involved in the project from early on, and they’ve received further valuable insight into an area of significant geological interest. On the construction side, the methodologies that have been developed during the course of the works, such as smooth blasting, could have a huge saving in time, cost and wastage during the Main Works. The Underground GI also included some significant ‘firsts’ in hard rock techniques.
Successfully managing the ground risk was the most important safety consideration for the works, but the standard of more conventional H&S practice on site is also deserving of mention. The project has clocked more than 500,000 man-hours and is now 2+ years without a Lost Time Incident (LTI).
Given the site’s location, managing environmental hazards has also been a big consideration. The Scottish Environment Protection Agency (SEPA) granted abstraction and discharge licences to the site in 2023, and during the course of the Works more than 160,000m3 of tunnel effluent was successfully uplifted, treated and returned to Loch Lochy, within strict water quality criteria. Seventeen Planning Monitoring Office inspections over the course of the Works have been made without a single non-compliance.
The project has welcomed visits by SEPA, the Highland Council, Nature Scot, and other local stakeholders, giving them insights into what tunnelling involves, which should assist them when working with us in future.
It would be remiss not to also mention the project’s recent accolades, including Project of the Year at the 2024 NCE Tunnelling Awards, and GI Project of the Year, two years running, at the Ground Engineering Awards.
Finally, deserving of one last mention, the collaborative spirit between Client, Designer and Contractor at Coire Glas was perhaps the greatest success, and allowed the project to overcome significant hurdles, and everyone came out the other side still on good terms!
Sean brought the presentation to a close with another note of thanks to everyone involved, but not before answering the question on everyone’s lips. What’s next? Coire Glas is awaiting its final investment decision from SSE, which is heavily dependent on the outcome of the Government’s Long Duration Energy Storage (LDES) ‘cap and floor’ payment mechanism. Engagement with the Designer, along with civil and M&E contractors, to finalise the scheme is ongoing. The plan is to commence Main Works construction once funding is secure, with the scheme to be completed in the early 2030’s.
Q&A
The speakers took questions following their presentation.
Q: Were you able to target all of the important underground elements required for the Main Works?
Sean: The Exploratory Works targeted structures based on a preliminary Reference Design, which will be refined for the Detailed Design. We have a good body of information for the Adit alignment, but also have the findings from the surface and underground GI. We’re confident we’ve carried out the most detailed GI possible at this stage.
Q: Is it anticipated that the fault zone encountered could intersect with any underground structures?
Sean: Review of the ground modelling is still ongoing, but with the fault discovered and its character understood, the position of structures can be reviewed to avoid interacting with it, or appropriate support designed.
Q: Are you doing long-term convergence monitoring? In the SQ section, was immediate support required?
Joe: Yes, quarterly monitoring is ongoing. During construction a regime of daily, weekly and monthly monitoring was completed, based upon support class. We continue to have had little movement since the monitoring arrays were installed. In the fault zone, we saw 2mm-3mm of convergence, especially at the base of the horseshoe, before the invert was fully closed, implying immediate support was required.
Q: Were there concerns about tunnel collapse during construction? How did you manage the risk?
Joe: Up to the fault zone, there was a lot of confidence that the ground support was suitably robust, given the support class system in use. The fault zone was a genuine concern, given the poor rock, and even with robust initial support installed, the Designer required additional measures for long term stability.
Q: What does Class-SQ stand for? What will the Exploratory Adit be used for long term?
Sean: SQ was a code without a specific meaning that stuck. Given the tunnel’s position, the current idea is it will be useful for access, but exactly how is in planning.
Q: Why was a reduced diameter borehole used for the exploratory drilling?
Tom R: Borehole diameter was dictated by the size of the rigs in use. The diameters used in vertical drilling on surface weren’t possible with rigs suitable for underground use. We considered HQ (63.5mm-diameter core size) to be sensible, given the equipment, and the distances of drilling required almost entirely in hard rock. This was agreed early on with the Designer, and did not affect the lab testing.
Q: Did surface GI findings affect decisions? Was it possible to interpret the fault zone thickness?
Sean: This was a separate package of works, but the Designer received data from both the surface and tunnelling GI, which was being assessed in real-time, so was able to influence their decisions. The fault zone wasn’t encountered from the surface boreholes, given the location that it intercepted the steep hillside above.
Tom R: Some of the underground boreholes were used to target the fault zone, to ascertain its position, with some success.
Q: What type of rock bolts have been installed? If it becomes a permanent adit, will it have to be re-lined?
Joe: Almost all rock bolts were fully grouted galvanised steel dowels, ranging from 25mm-32mm diameter, and 3m-3.6m in length. Some self-drilling hollow bars were utilised in the softer rock of the fault zone. GFRP bars were used in some instances, where longer bolts were required, given their reduced weight.
Sean: There is no plan for the Adit to be used as a permanent hydraulic tunnel. If it were to be used long-term as an access tunnel, a design review would be carried out.
Q: Have investigations been carried out on the stockpiled material, for re-use?
Sean: Yes, we have intentions to re-use the stockpiled material. We have conducted laboratory testing to understand its properties.
Q: Why do you think there was so little groundwater inflow during probing? Is the inflow from surface water?
Tom T: The most persistent inflows early on are possibly connected to streams on surface. All the other water encountered would have been from localised pockets, as it dried up after a short time.
Joe: Water strikes occurred during the underground GI. Two of the longest boreholes remain open to record inflow. Geological monitoring is ongoing, e.g., with piezometers, extensometers.
Q: What is the expected design life of the Exploratory Adit?
Sean: Given the scope of the works, the current Adit design life is short-term.
Q: How far from the Adit was drilling for overcoring tests? Stress regime?
Tom R: Overcoring too close to the Adit was avoided, to ensure valid results, which as a rule was taken to be a horizontal depth greater than the excavated span in that location. The testing also had to be targeted to areas that were unfractured, and the shallower the testing the easier the method. Depths were agreed with the Designer on site. The K-ratio and overall orientation were difficult to determine, given the geological setting. Results varied, and need to be reviewed in conjunction with the other underground geophysics, which is currently ongoing by the Designer.
Q: Were boreholes drilled from surface to tunnel level?
Sean: Surface GI included a steeply inclined borehole to 650m depth, in the planned location of the surge shaft, drilled from a compound on the top of the mountain, only accessible by helicopter.
