Neither Scotland nor the rest of the UK has seen the likes of what is happening in the Highlands for many years: new hydro construction with big, hard rock tunnelling.

As the largest renewable energy generator in the UK with 1.1GW of conventional hydro capacity, with the 300MW Foyers pumped storage scheme and total of 68 stations built in the 33 years to 1963, Scottish and Southern Electricity (SSE) is steeped in power assets. The utility’s asset base boasts 77 reservoirs, 94 dams, 300km of tunnel, 200km of aqueduct, 75km of pipeline and delivers 3200GWh of electricity annually.

The prime hydro sites have been developed. Therefore, aside from the benefit of having extra fast responding generation capacity at a time of generally higher energy costs, the additional incentives that helped bring about the Glendoe project included extra income under the state’s Renewables Obligation for adding extra ‘green’ capacity. Features of the site also offered a suitable reservoir location, high head, limited recreational use, the closeness of the transmission grid and that no designated sites requiring protection measures would be affected.

The Project

Glendoe is a 100MW hydro project being developed above Fort Augustus to use a single turbine-generator to exploit the runoff in a catchment of 75km² with an average catchment flow of 4m³/s. The low flows require a large catchment to capture enough water to exploit the potential of the high-head site.

Therefore, the project will have 16 intakes to feed a pipeline and 6.9km long aqueduct tunnel that lead to the reservoir, which has yet to be impounded behind a 920m long, 35m high rockfill dam. The water will be conveyed down a 6.2km long headrace tunnel under a gross head of 608m to the powerhouse some 300m underground.

In the powerhouse cavern, a six-jet, vertical axis-mounted Pelton turbine driven by 18.62m³/s of water under a net head of 605m, should help the client realise its annual output estimate of 180GWh. The fast responsiveness of the plant will mean that, already spinning, the turbine-generator will have a loading rate of 0-100MW in a minute. From standstill the turbine and generator would need to be synchronised, reaching full load in four minutes. The water will then be discharged from the turbine and flow via the 1.78km long tailrace tunnel to Loch Ness.

The total budget to capture that potential and give SSE the most efficient hydro station in the UK is US$296M (£150M) once commissioned in early 2009, said the client’s project manager Neil Sandilands. The scheme has been developed with extensive environmental consultation and the end product will be largely unnoticed by tourists who visit the area, as most of the infrastructure is being constructed underground.

Procurement

The design build contract was awarded in a two stage tender process beginning with a reference design (Stage 1) before moving on to the contractor’s proposed design (Stage 2). Jacobs has been the client’s adviser in the process, having itself competed for the engineering services contract.

Stage 1 established a reference design against which the interested parties could bid. The key parameters, from feasibility studies, were an underground powerhouse, access tunnel, single turbine, headrace, dam, aqueduct tunnel and pipelines.

There had been no real, intrusive site investigation at the time the reference design was being pulled together. Therefore, reference ground conditions were based on a desk study and a walk over the site which helped to create a “level playing field” for contractors, said David Hobson of Jacobs, tunnels adviser to the engineering unit within the project management team.

The client received 14 expressions of interest from its entry in The European Journal, four of which were to translate to submissions under the reference design, having submitted requests on particular work sought to be done in the site investigation. Two bidders were chosen for the last stage.

The two contractors then went head to head, preparing designs and revised cost estimates. With a bid of US$257M (£130M), the Hochtief Glendoe JV (Hochtief with Pöyry Energy, VA Tech Hydro and Andritz) won the bid competition in December 2005. The design and construct contract is under the NEC, 2nd Ed, Option A, with measurement by activity schedule, bonus and damages provisions, and is a fixed price arrangement except for volatile commodities, such as steel and fuel oil.

Project and risk management

SSE is the project manager for the Glendoe scheme, operating with its engineering adviser Jacobs in a single team with three sections: engineering, resident engineer and contracts. The adviser is paid on time worked and there is an independent site supervision team. Working alongside the contractor there are monthly contract review meetings, weekly site meetings and the aim of open communications.

A further aspect of the client’s close involvement in the project is SSE taking the major geotechnical risk on the job. It has three geologists on site. The Barton Q system was used for the reference design and is the basis for evaluating compensation events, and there are also automatic extensions of time. Costs are so far “mostly neutral” though Sandilands said they were slightly in the client’s favour due to “good ground”.

With respect to programme risk, in the reference design the powerhouse cavern was taken offline from the power tunnel (headrace + tailrace) to keep them from both being on the critical path of the construction schedule. Hobson noted for the stage 2 detailed design submissions, where the contractors could re-conceive the project layout, they came up with a fairly similar arrangement. Contractors also had to note, though, that due to the almost 12% incline of the headrace (unlike the more common drop tunnel between shallow runs) certain kinds of mucking out would be ruled out, e.g. rail.

The investigation for the power tunnel and the powerhouse was limited due to the access restrictions and depth of cover. For the contractor submissions, site investigation work was done by Fugro, there was ground based geophysics used and aerial geophysics, by Fugro Airborne. Seven boreholes were sunk, both vertical and inclined, ranging in depth from 25m-350m to target suspected locations of faults and fracture zones. “The boreholes were among some of the deepest in civils in the UK,” said Hobson.

A variety of tests were performed to establish characteristics such as UCS and abrasivity, and hydrofracture tests were performed at the tunnel horizon and cavern area to avoid water pressure “jacking” open the rock. The aim of the investigation was to determine a set of ground reference conditions which could be agreed by all parties. With all the extra data and contractor requests, the client modified the reference design to suit the respective tenderers.

Ground classification system

The Glendoe project area is south east of the Great Glen Fault zone and the rock consists of metamorphosed sedimentary rock of the Dalradian Supergroup. The rocks are folded and sheared with a sequence of interbedded quartzites, quartz schist and quartz mica schists and little groundwater.

Rock strengths range from 30MPa-40MPa up to 100MPa and, more commonly, 130MPa-150MPa. The range of cover to the tunnels is 250m-350m.

The classification adopted was a contractual requirement and was largely hazard based. A geological model was established at the design stage with rock mass properties identified for each tunnel, including any suspected faults or fissures, zones of limited cover and areas of unfavourable joint sets. Mitigation support measures were defined for the contract on the basis of four classes of rock (there having been five at the outset, in the reference design).

Hazards are re-assessed on site and the support criteria adjusted, if necessary, to suit local conditions. Contractor payment, though, is based on agreed rock class and not the support required to be installed.

Drill and blast excavation

Apart from the powerhouse, there are five main sections of tunnel in the Glendoe project to be excavated by drill and blast – the access tunnel to the cavern, a 150m long looping connection tunnel from the access to the top of the tailrace, the first 340m of the lower end of the tailrace tunnel and its abutting TBM adit tunnel, and lastly the aqueduct tunnel (figure 1). There is also a short cable tunnel close to the cavern.

Tunnel excavation began in May 2006 at the 1159m long main access tunnel, and this principal drill and blast drive was completed in January, having averaged 42m per week. Lining was steel fibre reinforced shotcrete plus rockbolts in the horseshoe-shaped tunnel section, 7m by 7m (face area of 42m2).

According to the contactor’s Johann Schöndube, who leads Hochtief’s global drill and blast activities, shotcreting saw some challenges both in terms of weather and local experience in the early days. He also noted that there were limits on explosives storage in the early months. Other difficulties included very heavy plant wear and breakdown at the outset due to worn clutches, brakes, drive shafts and so on due to the inclined tunnel.

More positively, to minimise manual work for health and safety reasons some tunnel areas were widened out for access from 2.2m-3.2m to take Jumbos and excavators. The contractor also saw low drill steel wear and with wider tunnels there would be reduced number of turning niches for plant.

Overall performance on drill and blast work so far at Glendoe saw 1.9kg/m3 of explosives used, 2.13 drillmetre per m3 excavated, the average pull per round was 4.8m, about 2.8m3 per metre of shotcrete was required, and 4.4 rockbolts per metre. Equipment used included a Tamrock Axera T-08 boomer 2+1, a CAT 950G wheel loader, three MB dump trucks (3332/3335), a CAT 313 wheel excavator, a Putzmeister WKM 133 shotcrete pump and two MB mix trucks, and a MAI 500 NT grout pump.

To excavate the aqueduct tunnel, which has a horseshoe-shaped cross section 4.8m wide by 4.5m high (face area of 19.5m2), the contractor plans to open up to five faces. The distance progressed is still in high, single-percent digits. The geology is mostly like elsewhere on the project as is known so far, though there could be localised alteration of rock due to contact metamorphism with acid igneous intrusive rocks. The support lining is steel fibre reinforced shotcrete with mesh as required, plus rockbolts.

For the 47m long by 33m high by 19m wide powerhouse, the top heading of the cavern has been opened, similar lining support being used plus sets as required. The contractor has reached crane-beam level with part of the first of four benches excavated. The dig started in January and is scheduled to be finished in June.

The predicted, or estimated, geology along the tunnel lengths to be excavated by drill and blast were 10% in the best (Class 1) and 75%, 13% and 2%, respectively, in Classes 2-4.

The excavation activities, and the discussions with the client’s geologists to settle classification, resulted in the same approximate split in aggregate between Classes 1-2 to Classes 3-4, or 86% and 14%, respectively – except they fell as Class 2 and Class 3; there was no Class 1 or Class 4.

However, Schöndube said he felt that about half of the resulting Class 3 could have been reasonably put as Class 4 – “but we’ll live with the cost” he added.

TBM excavation

The rock has been good so far for the TBM ‘Eliza Jane’ excavating upstream with a 5.022m cut diameter for the majority of the tailrace and then the headrace. The tailrace tunnel has been completed, a 108m long bridging tunnel to headrace tube is also completed, and more than 1000m of the headrace has now been bored, with completion due next January.

The proportion of tunnel in the different rock classes (1-4) were 60%, 25%, 10% and 5%, respectively. About 90% of the drive so far has been in Class 1, which is much better than expected, said Christian Zimmerman, TBM manager. The better class of rock helps improve advance rate.

The average daily advance along the tailrace was almost 17m, including the learning curve. By the time the unshielded TBM reached the headrace the advance rate had improved significantly, increasing to about 32m per day – well ahead of the required average of 24m per day. On average, by mid-March advance rates along the power tunnel were 21m per day. Best day and week were almost 74m and 273m, respectively.

Support is provided by rockbolts, mesh reinforced shotcrete and rockbolts, up to a maximum support of a full circle steel rib with 200mm thickness of applied shotcrete, should it prove necessary. However, so far, with Class 1 being the predominant rock the minimum support is 50mm shotcrete, if needed, and rockbolts are placed with Atlas Copco 1132 drills. For probing ahead an Atlas Copco Cop 1532 drill is used.

Muck away is achieved via a suspended conveyor type system with a belt cassette. Material supply is via a Deutz diesel powered low profile vehicle.

Zimmerman explained to the meeting about the features of the Robbins hard rock gripper machine, which was refurbished by and hired from Herrenknecht. The machine has installed power of 2900kW, has seven 315kW drive motors, a tightening torque of 2669kNm and breakaway torque of 3203kNm, and total thrust from four cylinders of 13,400kN. It weights 600 tonnes and is 200m long, being able to take horizontal and vertical curves of 500m and 750m radii, respectively. The cutterhead has 26 single discs, four double disc cutters and the diameters are 17”.

In terms of speed, Zimmerman said the procurement route took 2.5 months to make good the TBM, as compared to about eight months for a completely new machine. Refurbished parts included the 52 tonne cutterhead, shield, the 70 tonne, main beam, gripper, thrust cylinders and hydraulics. New parts included electric motors/gearboxes, main bearing, backup system, electrics, conveyor system and multi-service vehicles.

He added that the TBM had been chosen based on its availability, cost, and track record.

The contractor consulted early with the UK Health & Safety Executive to help satisfy regulations when transport and assembly of the parts got underway. Zimmerman said the approach resulted in full compliance with BS 6164 from the outset and it has helped to achieve a very low accident rate and a very good working relation with the UK’s Health & Safety Executive officials.

On site, the TBM excavated uphill from the relatively mild 0.5% of the tailrace and is getting underway with the steeper headrace climb of 11.7% gradient to reach the top, at the power tunnel inlet.

The tunnel is in three legs – 2585m, 3880m and 1450m – separated by 1000m radius horizontal curves. In total, about 7750m of the power tunnel will be bored by the TBM. Adding the 340m stretch of drill and blast in the tailrace, the excavated power tunnel will be 8090m long.

Figures given on TBM production so far put 24% of the time to boring, 11% to rock support, 13% to cutters, 9% to M&E and 5% to re-gripping. Conveyor work took 20%.

The BTS meeting closed with questions from the floor and the chairmen thanking the presenters for a most interesting and detailed presentation of the project.


Fig 1 – Planb of the project showing tunnel lengths and construction method Fig 1 – Planb of the project showing tunnel lengths … Fig 2 – The drill and blast pattern being used at Glendoe Fig 2 – The drill and blast pattern being used … Fig 3 – Longitudinal section of the TBM tunnel with the rock classes and support requirements added Fig 3 – Longitudinal section of the TBM tunnel with … Questions & Answers

Bob Ibell (London Bridge Associates) asked what measures the contractor had taken for training his staff in tunnel rescue procedures.
The contractor responded that a professional team had been engaged and had trained on site personnel. Regular follow up training was carried out. Representatives of the client and contractor were on the rescue team. It was noted that even though emergency services were aware of the works that were ongoing the on site rescue team provided the first response to any incident.
David Powell (Mott MacDonald) asked what the velocity of the water was within the headrace tunnel and whether a rock trap was provided upstream of the powerhouse.
The presenters responded that the velocity was in the region of 1.5m/s and that a large rock trap was provided for. With the low velocity the speakers thought that this provision was a conservative feature.
Colin McKenzie (retired) commented that in his experience rock adjacent to the Great Glenn fault often exhibited micro cracking in a plane parallel to the fault. He asked if the contractor had experienced such features.
In response it was stated that maintaining the profile was difficult in some locations. The roughness of the drill and blast tunnels did not have a major impact upon the hydraulics and overbreak was generally less than the ‘lookout’ resulting from the excavation of a 5m advance length.
Martin Knights (Jacobs) asked about the number of nationalities involved on the project.
At one point more than 25 could be counted.
Eddie Woods (CLRL) asked how well the geophysics related to the reality.
Of the 6% of the aqueduct that has been completed the correlation was good. Additional SI post contract award had allowed an improved correlation to be determined. The geophysics was regarded as a very useful design tool.
John Sawyer (Jacobs) asked if any disputes had arisen from the classification system used and if the presenters could comment on its implementation.
To date the geologists had been able to reach agreement on the classification system of the support installed. Sub classes of 2a, 2b, etc. were developed through agreement. Additional spiling was required in class 3.7, which was very close to a class 4 support, but an equitable agreement had been reached on all support classifications.
Rapporteur – Chris Bambridge
Edited & expanded – Patrick Reynolds